AkanaWiki - User contributions [en]

Merneha

2009-12-02T01:20:54Z

Nebulawindphone: /* Phase */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməgəmeh-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməgəmeh-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə ceci-ci]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Singulatives in -reh not -rə.

'''Cəmə''' ''n. I'' "Body"

'''Cəməjəmə''' ''n. IV'' "Everyone"

'''Cər-u''' ''stat.'' "One"

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cec-u''' ''v.i.'' "Defecate," redup. '''Cecec-u'''.

'''Cigic''' ''quant'' "Some"

'''Ciri, -rjən''' ''n. I'' "Cat," pl. '''Ciciri,''' pauc. '''Cirirə''' "Litter of cats."

'''Curireh-u''' ''v.t.'' "Drink," redup. '''Cucurireh-u'''.

'''Curicic-u''' ''v.i.'' "Quench one's thirst,"
* applic. '''Wədəcuriciceh-u''' "Enjoy, benefit from."

'''Cur-u''' ''v.i.'' "Fly," redup. '''Təcur-u''' "Fly around."

'''Ǝlum-u''' ''v.i.'' "Get up; go; set off on foot" redup. '''Ǝləlum-u''' "Go around, go here and there."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''ərah, -rarən''' ''n. I'' "Lover." (< PP '''*nriax''' "spouse")
* pauc. '''ərarə''' "Courting couple."

'''əramjə, -ramjan''' ''n. I'' "Bride."

'''ərahwə, -rahwan''' ''n. I'' "Groom."

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Halas, -lazən''' ''n. I'' "Chief," pl. '''Haralas'''.

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Hədər-u''' ''stat.'' "Ten," redup. '''Hihədər-u''' "Ten by ten."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh, -nərən''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Həreh''' ''pron.'' "I."
* '''Hərə''' "You and I."
* '''Hər''' "We."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed," redup. '''Hərəsur-u''' "Bleed out,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnur-u''' ''v.i.'' "Make camp," redup. '''Herẽnur-u'''.

'''Hẽnurwərəj, -rəjən''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hilucer-u''' ''stat.'' "Nine."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming," redup. '''Hĩhĩmət-u'''.

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Ira''' ''pron.'' "He, she,"
* '''Irar''' "They."

'''Ireh''' ''pron.'' "It,"
* '''Ir''' "They (inanimate)."

'''Irug-u''' ''v.i.'' "Go," redup. '''Irirug-u'''.

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Irukuj-u''' ''v.i.'' "Have gone, be gone."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kaməl''' ''n. I'' "Pan," pl. "Kəkaməl."

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kasə''' ''n. IV'' "Wind."

'''Katərə-w''' ''v.i.'' "Go hunting," redup. '''Kəgatərə-w'''.

'''Katərẽceh-u''' ''v.t.'' "Catch, hunt successfully," redup. '''Kəgatərẽceh-u'''.

'''Katəreh-u''' ''v.t.'' "Hunt for," patient trigger '''Katərec-u''', redup. '''Kəgatəreh-u'''.

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc, kəjən''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak, discuss," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədərət''' ''n. I'' "Message."

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərah-u''' ''stat.'' "Dark," redup. '''Kəgərah-u'''.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Kəstin, -tinən''' ''n. I'' "Knife," pl. '''Kəgəstin'''.
* pauc. '''Kəstĩrə''' "Plough."

'''Kũh-u''' ''stat.'' "Fat," redup '''Kugũh-u'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Ləlah-u''' ''stat.'' "Eight."

'''Lərut, -rudən''' ''n. I'' "Ocean," pl. '''Lələrut'''.

'''Ligwə, -gwan''' ''n. I'' "Parent," pauc. '''Ligwərə''' "Mother and son; father and daughter," pl. '''Liligwə'''.

'''Ligwəligwə''' ''n. IV'' "The Ancestors."

'''Ligwə-w''' ''stat.'' ''Who is (someone's) parent."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Lwəgwə''' ''n. I'' "River, creek."

'''Məgəmeh-u''' ''v.t.'' "Break."
* redup. '''Məməgəmeh-u''' "Defeat."

'''Məgis-u''' ''stat.'' "Holding (something), carrying (something)."

'''Məmir-u''' ''stat.'' "Six."

'''Mənə''' ''n. I'' "Bay."

'''Məpeh-u''' ''v.t.'' "Eat," patient trigger '''Məcic-u'''.

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Mər-u''' ''stat.'' "Three."

'''Medalə, -lan''' ''n. I'' "Opponent," pauc. '''Medalərə''' "One another's opponents," pl. '''Memedalə'''.

'''Milmeh-u''' ''v.t.'' "Tell (a story)," redup. '''Mimilmeh-u'''.

'''Minə''' ''n. I'' "Woman," pl. '''Miminə'''.

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Naheh''' ''pron.'' "You,"
* '''Nahə''' "You two,"
* '''Nəra''' "You all."

'''Nətər-u''' ''stat.'' "Seven."

'''Nigət, -gətən''' ''n. III'' "Berries."

'''Nizəj-u''' ''v.t.'' "Trades or barters with."

'''Nuheh-u''' ''v.t.'' "Pour out," redup. '''Nunuheh-u'''.

'''Nuh-u''' ''stat.'' "Whistling."

'''Nugim, -imən''' ''n. I'', pauc. '''Nugĩrə''', pl. '''Nunugim'''.

'''Pahu''' ''det.'' "What sort?"

'''Pəcur-u''' ''stat.'' "Two."

'''Pəka''' ''det.'' "Which?"

'''Pəkmjeh-u''' ''v.t.'' "Hunt," patient trigger '''Pəkmic-u'''.
* caus. '''Pəkmirazeh-u''' "Banish, outlaw, cause to be hunted."

'''Pəmə''' ''n. I.'' "Father," pl. '''Pəbəmə''',
* pauc. '''Pəmərə''' "Father and son."

'''Pəmir-u''' ''v.i.'' "Call out, shout," redup. '''Pəbəmir-u'''.

'''Pətu''' ''quant.'' "How many?"

'''Piri, -rjən''' ''n. I'' "Day," pl. '''Pibiri'''.

'''Pusmeh-u''' ''v.t.'' "Swallow," redup. '''Pupusmeh-u'''.

'''Pũt, -dən''' ''n. I'' "Bridge," pl. '''Pũpũdən'''.

'''Sirəm, -əmən''' ''n. I'', pl. '''Sizirəm'''.

'''Sism-u,''' ''v.i.'' "Sink," redup. '''Sizism-u'''.

'''Sumeh''' ''quant.'' "Every, all."

'''Sunəj, -nəjən''' ''n. I'' "River," pl. '''Suzunəj'''.

'''Swəłc-u''' ''stat.'' "Sick," redup. '''Swəzwəłc-u'''.

'''Swəłcwəłcu''' ''stat.'' "Crippled."

'''Təg-u''' ''stat.'' "Blue," redup. '''Tədəg-u'''.

'''Təjək, -jəgu''' ''n. I'' "Plant; herb, esp. medicinal."

'''Təkəbeh-u''' ''v.t.'' "Capture," redup. '''Tədəkəbeh-u'''.

'''Tərun, -runən''' ''n. II'' "Leaves of a tree."
* sg. '''Tərũreh''' "Leaf; page," pl. '''Tədərun''' "Leaves; pages."

'''Təməs, -məzən''' ''n. I'' "Village, pl. '''Tədəməs'''.

'''Tərãp, -rãbu''' ''n. I'' "Traveler," pl. '''Tədərãp'''.

'''Təs-u''', ''stat.'' "Five," redup. '''Tətəs-u''' "Five by five."

'''Tilə, -lan''' ''n. I'' "Dagger," pl. '''Tedilə'''.

'''Tunəw, -nəwən''' ''n. II'' "Horns." As a relational noun, "Peak, point." pl. '''Tudunəw'''.
* sg. '''Tunəweh''' "Horn; ferocity."

'''Tuswət, -wədən''' ''n. II'' "Someone's eyes," sg. '''Tuswəceh''' "An eye," pl. '''Tuduswət''' "Many eyes."

'''Ujək-u''' ''stat.'' "Covering," redup. '''Ulujək-u'''.

'''Ujəgĩcwir-u''' ''stat.'' "Covered in," redup. '''Ulujəgĩcwir-u'''.

'''Ulwəs, -wəsən''' ''n. I'' "Worm," pl. '''Ululwəs'''.

'''Ũrəweh-u''' ''v.t.'' "Cook," redup. '''Umũrəweh-u'''.

'''Uspəneh-u''' ''v.t.'' "Weave," redup. '''Uzuspəneh-u'''.

'''Wəgudeh-u''' ''v.t.'' "Give (something)," takes recipient as an optional oblique argument.
* redup. '''Wəgũkudeh-u''' "Give (something) up."
* applic. '''Nĩkudeh-u''' "Give (to someone)," takes thing given as an optional oblique argument.

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərer-u''' ''stat.'' "Four."

'''Wərĩhəj, -həjən''' ''n. I'' "House."

'''Wərjeh-u''' ''v.t.'' "Live in, dwell in," redup. '''Wərĩrjeh-u'''

'''Wərineh-u''' ''v.t.'' "Have dwelled in, hail from."

'''Wərn-u''' ''stat.'' "Bright," redup. "Wərwərn-u."

'''Wiłk, wiłkən''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T06:04:27Z

Nebulawindphone: /* Fish story */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməgəmeh-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməgəmeh-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə ceci-ci]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Singulatives in -reh not -rə.

'''Cəmə''' ''n. I'' "Body"

'''Cəməjəmə''' ''n. IV'' "Everyone"

'''Cər-u''' ''stat.'' "One"

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cec-u''' ''v.i.'' "Defecate," redup. '''Cecec-u'''.

'''Cigic''' ''quant'' "Some"

'''Ciri, -rjən''' ''n. I'' "Cat," pl. '''Ciciri,''' pauc. '''Cirirə''' "Litter of cats."

'''Curireh-u''' ''v.t.'' "Drink," redup. '''Cucurireh-u'''.

'''Curicic-u''' ''v.i.'' "Quench one's thirst,"
* applic. '''Wədəcuriciceh-u''' "Enjoy, benefit from."

'''Cur-u''' ''v.i.'' "Fly," redup. '''Təcur-u''' "Fly around."

'''Ǝlum-u''' ''v.i.'' "Get up; go; set off on foot" redup. '''Ǝləlum-u''' "Go around, go here and there."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''ərah, -rarən''' ''n. I'' "Lover." (< PP '''*nriax''' "spouse")
* pauc. '''ərarə''' "Courting couple."

'''əramjə, -ramjan''' ''n. I'' "Bride."

'''ərahwə, -rahwan''' ''n. I'' "Groom."

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Halas, -lazən''' ''n. I'' "Chief," pl. '''Haralas'''.

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Hədər-u''' ''stat.'' "Ten," redup. '''Hihədər-u''' "Ten by ten."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh, -nərən''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Həreh''' ''pron.'' "I."
* '''Hərə''' "You and I."
* '''Hər''' "We."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed," redup. '''Hərəsur-u''' "Bleed out,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnur-u''' ''v.i.'' "Make camp," redup. '''Herẽnur-u'''.

'''Hẽnurwərəj, -rəjən''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hilucer-u''' ''stat.'' "Nine."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming," redup. '''Hĩhĩmət-u'''.

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Ira''' ''pron.'' "He, she,"
* '''Irar''' "They."

'''Ireh''' ''pron.'' "It,"
* '''Ir''' "They (inanimate)."

'''Irug-u''' ''v.i.'' "Go," redup. '''Irirug-u'''.

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Irukuj-u''' ''v.i.'' "Have gone, be gone."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kaməl''' ''n. I'' "Pan," pl. "Kəkaməl."

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kasə''' ''n. IV'' "Wind."

'''Katərə-w''' ''v.i.'' "Go hunting," redup. '''Kəgatərə-w'''.

'''Katərẽceh-u''' ''v.t.'' "Catch, hunt successfully," redup. '''Kəgatərẽceh-u'''.

'''Katəreh-u''' ''v.t.'' "Hunt for," patient trigger '''Katərec-u''', redup. '''Kəgatəreh-u'''.

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc, kəjən''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak, discuss," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədərət''' ''n. I'' "Message."

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərah-u''' ''stat.'' "Dark," redup. '''Kəgərah-u'''.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Kəstin, -tinən''' ''n. I'' "Knife," pl. '''Kəgəstin'''.
* pauc. '''Kəstĩrə''' "Plough."

'''Kũh-u''' ''stat.'' "Fat," redup '''Kugũh-u'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Ləlah-u''' ''stat.'' "Eight."

'''Lərut, -rudən''' ''n. I'' "Ocean," pl. '''Lələrut'''.

'''Ligwə, -gwan''' ''n. I'' "Parent," pauc. '''Ligwərə''' "Mother and son; father and daughter," pl. '''Liligwə'''.

'''Ligwəligwə''' ''n. IV'' "The Ancestors."

'''Ligwə-w''' ''stat.'' ''Who is (someone's) parent."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Lwəgwə''' ''n. I'' "River, creek."

'''Məgəmeh-u''' ''v.t.'' "Break."
* redup. '''Məməgəmeh-u''' "Defeat."

'''Məgis-u''' ''stat.'' "Holding (something), carrying (something)."

'''Məmir-u''' ''stat.'' "Six."

'''Mənə''' ''n. I'' "Bay."

'''Məpeh-u''' ''v.t.'' "Eat," patient trigger '''Məcic-u'''.

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Mər-u''' ''stat.'' "Three."

'''Medalə, -lan''' ''n. I'' "Opponent," pauc. '''Medalərə''' "One another's opponents," pl. '''Memedalə'''.

'''Milmeh-u''' ''v.t.'' "Tell (a story)," redup. '''Mimilmeh-u'''.

'''Minə''' ''n. I'' "Woman," pl. '''Miminə'''.

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Naheh''' ''pron.'' "You,"
* '''Nahə''' "You two,"
* '''Nəra''' "You all."

'''Nətər-u''' ''stat.'' "Seven."

'''Nigət, -gətən''' ''n. III'' "Berries."

'''Nizəj-u''' ''v.t.'' "Trades or barters with."

'''Nuheh-u''' ''v.t.'' "Pour out," redup. '''Nunuheh-u'''.

'''Nuh-u''' ''stat.'' "Whistling."

'''Nugim, -imən''' ''n. I'', pauc. '''Nugĩrə''', pl. '''Nunugim'''.

'''Pahu''' ''det.'' "What sort?"

'''Pəcur-u''' ''stat.'' "Two."

'''Pəka''' ''det.'' "Which?"

'''Pəkmjeh-u''' ''v.t.'' "Hunt," patient trigger '''Pəkmic-u'''.
* caus. '''Pəkmirazeh-u''' "Banish, outlaw, cause to be hunted."

'''Pəmə''' ''n. I.'' "Father," pl. '''Pəbəmə''',
* pauc. '''Pəmərə''' "Father and son."

'''Pəmir-u''' ''v.i.'' "Call out, shout," redup. '''Pəbəmir-u'''.

'''Pətu''' ''quant.'' "How many?"

'''Piri, -rjən''' ''n. I'' "Day," pl. '''Pibiri'''.

'''Pusmeh-u''' ''v.t.'' "Swallow," redup. '''Pupusmeh-u'''.

'''Pũt, -dən''' ''n. I'' "Bridge," pl. '''Pũpũdən'''.

'''Sirəm, -əmən''' ''n. I'', pl. '''Sizirəm'''.

'''Sism-u,''' ''v.i.'' "Sink," redup. '''Sizism-u'''.

'''Sumeh''' ''quant.'' "Every, all."

'''Sunəj, -nəjən''' ''n. I'' "River," pl. '''Suzunəj'''.

'''Swəłc-u''' ''stat.'' "Sick," redup. '''Swəzwəłc-u'''.

'''Swəłcwəłcu''' ''stat.'' "Crippled."

'''Təg-u''' ''stat.'' "Blue," redup. '''Tədəg-u'''.

'''Təjək, -jəgu''' ''n. I'' "Plant; herb, esp. medicinal."

'''Təkəbeh-u''' ''v.t.'' "Capture," redup. '''Tədəkəbeh-u'''.

'''Tərun, -runən''' ''n. II'' "Leaves of a tree."
* sg. '''Tərũreh''' "Leaf; page," pl. '''Tədərun''' "Leaves; pages."

'''Təməs, -məzən''' ''n. I'' "Village, pl. '''Tədəməs'''.

'''Tərãp, -rãbu''' ''n. I'' "Traveler," pl. '''Tədərãp'''.

'''Təs-u''', ''stat.'' "Five," redup. '''Tətəs-u''' "Five by five."

'''Tilə, -lan''' ''n. I'' "Dagger," pl. '''Tedilə'''.

'''Tunəw, -nəwən''' ''n. II'' "Horns." As a relational noun, "Peak, point." pl. '''Tudunəw'''.
* sg. '''Tunəweh''' "Horn; ferocity."

'''Tuswət, -wədən''' ''n. II'' "Someone's eyes," sg. '''Tuswəceh''' "An eye," pl. '''Tuduswət''' "Many eyes."

'''Ujək-u''' ''stat.'' "Covering," redup. '''Ulujək-u'''.

'''Ujəgĩcwir-u''' ''stat.'' "Covered in," redup. '''Ulujəgĩcwir-u'''.

'''Ulwəs, -wəsən''' ''n. I'' "Worm," pl. '''Ululwəs'''.

'''Ũrəweh-u''' ''v.t.'' "Cook," redup. '''Umũrəweh-u'''.

'''Uspəneh-u''' ''v.t.'' "Weave," redup. '''Uzuspəneh-u'''.

'''Wəgudeh-u''' ''v.t.'' "Give (something)," takes recipient as an optional oblique argument.
* redup. '''Wəgũkudeh-u''' "Give (something) up."
* applic. '''Nĩkudeh-u''' "Give (to someone)," takes thing given as an optional oblique argument.

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərer-u''' ''stat.'' "Four."

'''Wərĩhəj, -həjən''' ''n. I'' "House."

'''Wərjeh-u''' ''v.t.'' "Live in, dwell in," redup. '''Wərĩrjeh-u'''

'''Wərineh-u''' ''v.t.'' "Have dwelled in, hail from."

'''Wərn-u''' ''stat.'' "Bright," redup. "Wərwərn-u."

'''Wiłk, wiłkən''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T06:01:58Z

Nebulawindphone: /* The copula */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməgəmeh-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməgəmeh-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə ceci-ci]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Singulatives in -reh not -rə.

'''Cəmə''' ''n. I'' "Body"

'''Cəməjəmə''' ''n. IV'' "Everyone"

'''Cər-u''' ''stat.'' "One"

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cec-u''' ''v.i.'' "Defecate," redup. '''Cecec-u'''.

'''Cigic''' ''quant'' "Some"

'''Ciri, -rjən''' ''n. I'' "Cat," pl. '''Ciciri,''' pauc. '''Cirirə''' "Litter of cats."

'''Curireh-u''' ''v.t.'' "Drink," redup. '''Cucurireh-u'''.

'''Curicic-u''' ''v.i.'' "Quench one's thirst,"
* applic. '''Wədəcuriciceh-u''' "Enjoy, benefit from."

'''Cur-u''' ''v.i.'' "Fly," redup. '''Təcur-u''' "Fly around."

'''Ǝlum-u''' ''v.i.'' "Get up; go; set off on foot" redup. '''Ǝləlum-u''' "Go around, go here and there."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''ərah, -rarən''' ''n. I'' "Lover." (< PP '''*nriax''' "spouse")
* pauc. '''ərarə''' "Courting couple."

'''əramjə, -ramjan''' ''n. I'' "Bride."

'''ərahwə, -rahwan''' ''n. I'' "Groom."

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Halas, -lazən''' ''n. I'' "Chief," pl. '''Haralas'''.

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Hədər-u''' ''stat.'' "Ten," redup. '''Hihədər-u''' "Ten by ten."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh, -nərən''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Həreh''' ''pron.'' "I."
* '''Hərə''' "You and I."
* '''Hər''' "We."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed," redup. '''Hərəsur-u''' "Bleed out,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnur-u''' ''v.i.'' "Make camp," redup. '''Herẽnur-u'''.

'''Hẽnurwərəj, -rəjən''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hilucer-u''' ''stat.'' "Nine."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming," redup. '''Hĩhĩmət-u'''.

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Ira''' ''pron.'' "He, she,"
* '''Irar''' "They."

'''Ireh''' ''pron.'' "It,"
* '''Ir''' "They (inanimate)."

'''Irug-u''' ''v.i.'' "Go," redup. '''Irirug-u'''.

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Irukuj-u''' ''v.i.'' "Have gone, be gone."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kaməl''' ''n. I'' "Pan," pl. "Kəkaməl."

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kasə''' ''n. IV'' "Wind."

'''Katərə-w''' ''v.i.'' "Go hunting," redup. '''Kəgatərə-w'''.

'''Katərẽceh-u''' ''v.t.'' "Catch, hunt successfully," redup. '''Kəgatərẽceh-u'''.

'''Katəreh-u''' ''v.t.'' "Hunt for," patient trigger '''Katərec-u''', redup. '''Kəgatəreh-u'''.

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc, kəjən''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak, discuss," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədərət''' ''n. I'' "Message."

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərah-u''' ''stat.'' "Dark," redup. '''Kəgərah-u'''.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Kəstin, -tinən''' ''n. I'' "Knife," pl. '''Kəgəstin'''.
* pauc. '''Kəstĩrə''' "Plough."

'''Kũh-u''' ''stat.'' "Fat," redup '''Kugũh-u'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Ləlah-u''' ''stat.'' "Eight."

'''Lərut, -rudən''' ''n. I'' "Ocean," pl. '''Lələrut'''.

'''Ligwə, -gwan''' ''n. I'' "Parent," pauc. '''Ligwərə''' "Mother and son; father and daughter," pl. '''Liligwə'''.

'''Ligwəligwə''' ''n. IV'' "The Ancestors."

'''Ligwə-w''' ''stat.'' ''Who is (someone's) parent."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Lwəgwə''' ''n. I'' "River, creek."

'''Məgəmeh-u''' ''v.t.'' "Break."
* redup. '''Məməgəmeh-u''' "Defeat."

'''Məgis-u''' ''stat.'' "Holding (something), carrying (something)."

'''Məmir-u''' ''stat.'' "Six."

'''Mənə''' ''n. I'' "Bay."

'''Məpeh-u''' ''v.t.'' "Eat," patient trigger '''Məcic-u'''.

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Mər-u''' ''stat.'' "Three."

'''Medalə, -lan''' ''n. I'' "Opponent," pauc. '''Medalərə''' "One another's opponents," pl. '''Memedalə'''.

'''Milmeh-u''' ''v.t.'' "Tell (a story)," redup. '''Mimilmeh-u'''.

'''Minə''' ''n. I'' "Woman," pl. '''Miminə'''.

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Naheh''' ''pron.'' "You,"
* '''Nahə''' "You two,"
* '''Nəra''' "You all."

'''Nətər-u''' ''stat.'' "Seven."

'''Nigət, -gətən''' ''n. III'' "Berries."

'''Nizəj-u''' ''v.t.'' "Trades or barters with."

'''Nuheh-u''' ''v.t.'' "Pour out," redup. '''Nunuheh-u'''.

'''Nuh-u''' ''stat.'' "Whistling."

'''Nugim, -imən''' ''n. I'', pauc. '''Nugĩrə''', pl. '''Nunugim'''.

'''Pahu''' ''det.'' "What sort?"

'''Pəcur-u''' ''stat.'' "Two."

'''Pəka''' ''det.'' "Which?"

'''Pəkmjeh-u''' ''v.t.'' "Hunt," patient trigger '''Pəkmic-u'''.
* caus. '''Pəkmirazeh-u''' "Banish, outlaw, cause to be hunted."

'''Pəmə''' ''n. I.'' "Father," pl. '''Pəbəmə''',
* pauc. '''Pəmərə''' "Father and son."

'''Pəmir-u''' ''v.i.'' "Call out, shout," redup. '''Pəbəmir-u'''.

'''Pətu''' ''quant.'' "How many?"

'''Piri, -rjən''' ''n. I'' "Day," pl. '''Pibiri'''.

'''Pusmeh-u''' ''v.t.'' "Swallow," redup. '''Pupusmeh-u'''.

'''Pũt, -dən''' ''n. I'' "Bridge," pl. '''Pũpũdən'''.

'''Sirəm, -əmən''' ''n. I'', pl. '''Sizirəm'''.

'''Sism-u,''' ''v.i.'' "Sink," redup. '''Sizism-u'''.

'''Sumeh''' ''quant.'' "Every, all."

'''Sunəj, -nəjən''' ''n. I'' "River," pl. '''Suzunəj'''.

'''Swəłc-u''' ''stat.'' "Sick," redup. '''Swəzwəłc-u'''.

'''Swəłcwəłcu''' ''stat.'' "Crippled."

'''Təg-u''' ''stat.'' "Blue," redup. '''Tədəg-u'''.

'''Təjək, -jəgu''' ''n. I'' "Plant; herb, esp. medicinal."

'''Təkəbeh-u''' ''v.t.'' "Capture," redup. '''Tədəkəbeh-u'''.

'''Tərun, -runən''' ''n. II'' "Leaves of a tree."
* sg. '''Tərũreh''' "Leaf; page," pl. '''Tədərun''' "Leaves; pages."

'''Təməs, -məzən''' ''n. I'' "Village, pl. '''Tədəməs'''.

'''Tərãp, -rãbu''' ''n. I'' "Traveler," pl. '''Tədərãp'''.

'''Təs-u''', ''stat.'' "Five," redup. '''Tətəs-u''' "Five by five."

'''Tilə, -lan''' ''n. I'' "Dagger," pl. '''Tedilə'''.

'''Tunəw, -nəwən''' ''n. II'' "Horns." As a relational noun, "Peak, point." pl. '''Tudunəw'''.
* sg. '''Tunəweh''' "Horn; ferocity."

'''Tuswət, -wədən''' ''n. II'' "Someone's eyes," sg. '''Tuswəceh''' "An eye," pl. '''Tuduswət''' "Many eyes."

'''Ujək-u''' ''stat.'' "Covering," redup. '''Ulujək-u'''.

'''Ujəgĩcwir-u''' ''stat.'' "Covered in," redup. '''Ulujəgĩcwir-u'''.

'''Ulwəs, -wəsən''' ''n. I'' "Worm," pl. '''Ululwəs'''.

'''Ũrəweh-u''' ''v.t.'' "Cook," redup. '''Umũrəweh-u'''.

'''Uspəneh-u''' ''v.t.'' "Weave," redup. '''Uzuspəneh-u'''.

'''Wəgudeh-u''' ''v.t.'' "Give (something)," takes recipient as an optional oblique argument.
* redup. '''Wəgũkudeh-u''' "Give (something) up."
* applic. '''Nĩkudeh-u''' "Give (to someone)," takes thing given as an optional oblique argument.

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərer-u''' ''stat.'' "Four."

'''Wərĩhəj, -həjən''' ''n. I'' "House."

'''Wərjeh-u''' ''v.t.'' "Live in, dwell in," redup. '''Wərĩrjeh-u'''

'''Wərineh-u''' ''v.t.'' "Have dwelled in, hail from."

'''Wərn-u''' ''stat.'' "Bright," redup. "Wərwərn-u."

'''Wiłk, wiłkən''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T06:00:36Z

Nebulawindphone: /* The copula */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməgəmeh-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməgəmeh-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Singulatives in -reh not -rə.

'''Cəmə''' ''n. I'' "Body"

'''Cəməjəmə''' ''n. IV'' "Everyone"

'''Cər-u''' ''stat.'' "One"

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cec-u''' ''v.i.'' "Defecate," redup. '''Cecec-u'''.

'''Cigic''' ''quant'' "Some"

'''Ciri, -rjən''' ''n. I'' "Cat," pl. '''Ciciri,''' pauc. '''Cirirə''' "Litter of cats."

'''Curireh-u''' ''v.t.'' "Drink," redup. '''Cucurireh-u'''.

'''Curicic-u''' ''v.i.'' "Quench one's thirst,"
* applic. '''Wədəcuriciceh-u''' "Enjoy, benefit from."

'''Cur-u''' ''v.i.'' "Fly," redup. '''Təcur-u''' "Fly around."

'''Ǝlum-u''' ''v.i.'' "Get up; go; set off on foot" redup. '''Ǝləlum-u''' "Go around, go here and there."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''ərah, -rarən''' ''n. I'' "Lover." (< PP '''*nriax''' "spouse")
* pauc. '''ərarə''' "Courting couple."

'''əramjə, -ramjan''' ''n. I'' "Bride."

'''ərahwə, -rahwan''' ''n. I'' "Groom."

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Halas, -lazən''' ''n. I'' "Chief," pl. '''Haralas'''.

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Hədər-u''' ''stat.'' "Ten," redup. '''Hihədər-u''' "Ten by ten."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh, -nərən''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Həreh''' ''pron.'' "I."
* '''Hərə''' "You and I."
* '''Hər''' "We."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed," redup. '''Hərəsur-u''' "Bleed out,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnur-u''' ''v.i.'' "Make camp," redup. '''Herẽnur-u'''.

'''Hẽnurwərəj, -rəjən''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hilucer-u''' ''stat.'' "Nine."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming," redup. '''Hĩhĩmət-u'''.

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Ira''' ''pron.'' "He, she,"
* '''Irar''' "They."

'''Ireh''' ''pron.'' "It,"
* '''Ir''' "They (inanimate)."

'''Irug-u''' ''v.i.'' "Go," redup. '''Irirug-u'''.

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Irukuj-u''' ''v.i.'' "Have gone, be gone."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kaməl''' ''n. I'' "Pan," pl. "Kəkaməl."

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kasə''' ''n. IV'' "Wind."

'''Katərə-w''' ''v.i.'' "Go hunting," redup. '''Kəgatərə-w'''.

'''Katərẽceh-u''' ''v.t.'' "Catch, hunt successfully," redup. '''Kəgatərẽceh-u'''.

'''Katəreh-u''' ''v.t.'' "Hunt for," patient trigger '''Katərec-u''', redup. '''Kəgatəreh-u'''.

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc, kəjən''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak, discuss," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədərət''' ''n. I'' "Message."

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərah-u''' ''stat.'' "Dark," redup. '''Kəgərah-u'''.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Kəstin, -tinən''' ''n. I'' "Knife," pl. '''Kəgəstin'''.
* pauc. '''Kəstĩrə''' "Plough."

'''Kũh-u''' ''stat.'' "Fat," redup '''Kugũh-u'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Ləlah-u''' ''stat.'' "Eight."

'''Lərut, -rudən''' ''n. I'' "Ocean," pl. '''Lələrut'''.

'''Ligwə, -gwan''' ''n. I'' "Parent," pauc. '''Ligwərə''' "Mother and son; father and daughter," pl. '''Liligwə'''.

'''Ligwəligwə''' ''n. IV'' "The Ancestors."

'''Ligwə-w''' ''stat.'' ''Who is (someone's) parent."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Lwəgwə''' ''n. I'' "River, creek."

'''Məgəmeh-u''' ''v.t.'' "Break."
* redup. '''Məməgəmeh-u''' "Defeat."

'''Məgis-u''' ''stat.'' "Holding (something), carrying (something)."

'''Məmir-u''' ''stat.'' "Six."

'''Mənə''' ''n. I'' "Bay."

'''Məpeh-u''' ''v.t.'' "Eat," patient trigger '''Məcic-u'''.

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Mər-u''' ''stat.'' "Three."

'''Medalə, -lan''' ''n. I'' "Opponent," pauc. '''Medalərə''' "One another's opponents," pl. '''Memedalə'''.

'''Milmeh-u''' ''v.t.'' "Tell (a story)," redup. '''Mimilmeh-u'''.

'''Minə''' ''n. I'' "Woman," pl. '''Miminə'''.

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Naheh''' ''pron.'' "You,"
* '''Nahə''' "You two,"
* '''Nəra''' "You all."

'''Nətər-u''' ''stat.'' "Seven."

'''Nigət, -gətən''' ''n. III'' "Berries."

'''Nizəj-u''' ''v.t.'' "Trades or barters with."

'''Nuheh-u''' ''v.t.'' "Pour out," redup. '''Nunuheh-u'''.

'''Nuh-u''' ''stat.'' "Whistling."

'''Nugim, -imən''' ''n. I'', pauc. '''Nugĩrə''', pl. '''Nunugim'''.

'''Pahu''' ''det.'' "What sort?"

'''Pəcur-u''' ''stat.'' "Two."

'''Pəka''' ''det.'' "Which?"

'''Pəkmjeh-u''' ''v.t.'' "Hunt," patient trigger '''Pəkmic-u'''.
* caus. '''Pəkmirazeh-u''' "Banish, outlaw, cause to be hunted."

'''Pəmə''' ''n. I.'' "Father," pl. '''Pəbəmə''',
* pauc. '''Pəmərə''' "Father and son."

'''Pəmir-u''' ''v.i.'' "Call out, shout," redup. '''Pəbəmir-u'''.

'''Pətu''' ''quant.'' "How many?"

'''Piri, -rjən''' ''n. I'' "Day," pl. '''Pibiri'''.

'''Pusmeh-u''' ''v.t.'' "Swallow," redup. '''Pupusmeh-u'''.

'''Pũt, -dən''' ''n. I'' "Bridge," pl. '''Pũpũdən'''.

'''Sirəm, -əmən''' ''n. I'', pl. '''Sizirəm'''.

'''Sism-u,''' ''v.i.'' "Sink," redup. '''Sizism-u'''.

'''Sumeh''' ''quant.'' "Every, all."

'''Sunəj, -nəjən''' ''n. I'' "River," pl. '''Suzunəj'''.

'''Swəłc-u''' ''stat.'' "Sick," redup. '''Swəzwəłc-u'''.

'''Swəłcwəłcu''' ''stat.'' "Crippled."

'''Təg-u''' ''stat.'' "Blue," redup. '''Tədəg-u'''.

'''Təjək, -jəgu''' ''n. I'' "Plant; herb, esp. medicinal."

'''Təkəbeh-u''' ''v.t.'' "Capture," redup. '''Tədəkəbeh-u'''.

'''Tərun, -runən''' ''n. II'' "Leaves of a tree."
* sg. '''Tərũreh''' "Leaf; page," pl. '''Tədərun''' "Leaves; pages."

'''Təməs, -məzən''' ''n. I'' "Village, pl. '''Tədəməs'''.

'''Tərãp, -rãbu''' ''n. I'' "Traveler," pl. '''Tədərãp'''.

'''Təs-u''', ''stat.'' "Five," redup. '''Tətəs-u''' "Five by five."

'''Tilə, -lan''' ''n. I'' "Dagger," pl. '''Tedilə'''.

'''Tunəw, -nəwən''' ''n. II'' "Horns." As a relational noun, "Peak, point." pl. '''Tudunəw'''.
* sg. '''Tunəweh''' "Horn; ferocity."

'''Tuswət, -wədən''' ''n. II'' "Someone's eyes," sg. '''Tuswəceh''' "An eye," pl. '''Tuduswət''' "Many eyes."

'''Ujək-u''' ''stat.'' "Covering," redup. '''Ulujək-u'''.

'''Ujəgĩcwir-u''' ''stat.'' "Covered in," redup. '''Ulujəgĩcwir-u'''.

'''Ulwəs, -wəsən''' ''n. I'' "Worm," pl. '''Ululwəs'''.

'''Ũrəweh-u''' ''v.t.'' "Cook," redup. '''Umũrəweh-u'''.

'''Uspəneh-u''' ''v.t.'' "Weave," redup. '''Uzuspəneh-u'''.

'''Wəgudeh-u''' ''v.t.'' "Give (something)," takes recipient as an optional oblique argument.
* redup. '''Wəgũkudeh-u''' "Give (something) up."
* applic. '''Nĩkudeh-u''' "Give (to someone)," takes thing given as an optional oblique argument.

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərer-u''' ''stat.'' "Four."

'''Wərĩhəj, -həjən''' ''n. I'' "House."

'''Wərjeh-u''' ''v.t.'' "Live in, dwell in," redup. '''Wərĩrjeh-u'''

'''Wərineh-u''' ''v.t.'' "Have dwelled in, hail from."

'''Wərn-u''' ''stat.'' "Bright," redup. "Wərwərn-u."

'''Wiłk, wiłkən''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T05:59:52Z

Nebulawindphone: /* Lexicon */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməgəmeh-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməgəmeh-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Singulatives in -reh not -rə.

'''Cəmə''' ''n. I'' "Body"

'''Cəməjəmə''' ''n. IV'' "Everyone"

'''Cər-u''' ''stat.'' "One"

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cec-u''' ''v.i.'' "Defecate," redup. '''Cecec-u'''.

'''Cigic''' ''quant'' "Some"

'''Ciri, -rjən''' ''n. I'' "Cat," pl. '''Ciciri,''' pauc. '''Cirirə''' "Litter of cats."

'''Curireh-u''' ''v.t.'' "Drink," redup. '''Cucurireh-u'''.

'''Curicic-u''' ''v.i.'' "Quench one's thirst,"
* applic. '''Wədəcuriciceh-u''' "Enjoy, benefit from."

'''Cur-u''' ''v.i.'' "Fly," redup. '''Təcur-u''' "Fly around."

'''Ǝlum-u''' ''v.i.'' "Get up; go; set off on foot" redup. '''Ǝləlum-u''' "Go around, go here and there."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''ərah, -rarən''' ''n. I'' "Lover." (< PP '''*nriax''' "spouse")
* pauc. '''ərarə''' "Courting couple."

'''əramjə, -ramjan''' ''n. I'' "Bride."

'''ərahwə, -rahwan''' ''n. I'' "Groom."

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Halas, -lazən''' ''n. I'' "Chief," pl. '''Haralas'''.

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Hədər-u''' ''stat.'' "Ten," redup. '''Hihədər-u''' "Ten by ten."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh, -nərən''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Həreh''' ''pron.'' "I."
* '''Hərə''' "You and I."
* '''Hər''' "We."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed," redup. '''Hərəsur-u''' "Bleed out,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnur-u''' ''v.i.'' "Make camp," redup. '''Herẽnur-u'''.

'''Hẽnurwərəj, -rəjən''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hilucer-u''' ''stat.'' "Nine."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming," redup. '''Hĩhĩmət-u'''.

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Ira''' ''pron.'' "He, she,"
* '''Irar''' "They."

'''Ireh''' ''pron.'' "It,"
* '''Ir''' "They (inanimate)."

'''Irug-u''' ''v.i.'' "Go," redup. '''Irirug-u'''.

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Irukuj-u''' ''v.i.'' "Have gone, be gone."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kaməl''' ''n. I'' "Pan," pl. "Kəkaməl."

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kasə''' ''n. IV'' "Wind."

'''Katərə-w''' ''v.i.'' "Go hunting," redup. '''Kəgatərə-w'''.

'''Katərẽceh-u''' ''v.t.'' "Catch, hunt successfully," redup. '''Kəgatərẽceh-u'''.

'''Katəreh-u''' ''v.t.'' "Hunt for," patient trigger '''Katərec-u''', redup. '''Kəgatəreh-u'''.

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc, kəjən''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak, discuss," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədərət''' ''n. I'' "Message."

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərah-u''' ''stat.'' "Dark," redup. '''Kəgərah-u'''.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Kəstin, -tinən''' ''n. I'' "Knife," pl. '''Kəgəstin'''.
* pauc. '''Kəstĩrə''' "Plough."

'''Kũh-u''' ''stat.'' "Fat," redup '''Kugũh-u'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Ləlah-u''' ''stat.'' "Eight."

'''Lərut, -rudən''' ''n. I'' "Ocean," pl. '''Lələrut'''.

'''Ligwə, -gwan''' ''n. I'' "Parent," pauc. '''Ligwərə''' "Mother and son; father and daughter," pl. '''Liligwə'''.

'''Ligwəligwə''' ''n. IV'' "The Ancestors."

'''Ligwə-w''' ''stat.'' ''Who is (someone's) parent."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Lwəgwə''' ''n. I'' "River, creek."

'''Məgəmeh-u''' ''v.t.'' "Break."
* redup. '''Məməgəmeh-u''' "Defeat."

'''Məgis-u''' ''stat.'' "Holding (something), carrying (something)."

'''Məmir-u''' ''stat.'' "Six."

'''Mənə''' ''n. I'' "Bay."

'''Məpeh-u''' ''v.t.'' "Eat," patient trigger '''Məcic-u'''.

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Mər-u''' ''stat.'' "Three."

'''Medalə, -lan''' ''n. I'' "Opponent," pauc. '''Medalərə''' "One another's opponents," pl. '''Memedalə'''.

'''Milmeh-u''' ''v.t.'' "Tell (a story)," redup. '''Mimilmeh-u'''.

'''Minə''' ''n. I'' "Woman," pl. '''Miminə'''.

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Naheh''' ''pron.'' "You,"
* '''Nahə''' "You two,"
* '''Nəra''' "You all."

'''Nətər-u''' ''stat.'' "Seven."

'''Nigət, -gətən''' ''n. III'' "Berries."

'''Nizəj-u''' ''v.t.'' "Trades or barters with."

'''Nuheh-u''' ''v.t.'' "Pour out," redup. '''Nunuheh-u'''.

'''Nuh-u''' ''stat.'' "Whistling."

'''Nugim, -imən''' ''n. I'', pauc. '''Nugĩrə''', pl. '''Nunugim'''.

'''Pahu''' ''det.'' "What sort?"

'''Pəcur-u''' ''stat.'' "Two."

'''Pəka''' ''det.'' "Which?"

'''Pəkmjeh-u''' ''v.t.'' "Hunt," patient trigger '''Pəkmic-u'''.
* caus. '''Pəkmirazeh-u''' "Banish, outlaw, cause to be hunted."

'''Pəmə''' ''n. I.'' "Father," pl. '''Pəbəmə''',
* pauc. '''Pəmərə''' "Father and son."

'''Pəmir-u''' ''v.i.'' "Call out, shout," redup. '''Pəbəmir-u'''.

'''Pətu''' ''quant.'' "How many?"

'''Piri, -rjən''' ''n. I'' "Day," pl. '''Pibiri'''.

'''Pusmeh-u''' ''v.t.'' "Swallow," redup. '''Pupusmeh-u'''.

'''Pũt, -dən''' ''n. I'' "Bridge," pl. '''Pũpũdən'''.

'''Sirəm, -əmən''' ''n. I'', pl. '''Sizirəm'''.

'''Sism-u,''' ''v.i.'' "Sink," redup. '''Sizism-u'''.

'''Sumeh''' ''quant.'' "Every, all."

'''Sunəj, -nəjən''' ''n. I'' "River," pl. '''Suzunəj'''.

'''Swəłc-u''' ''stat.'' "Sick," redup. '''Swəzwəłc-u'''.

'''Swəłcwəłcu''' ''stat.'' "Crippled."

'''Təg-u''' ''stat.'' "Blue," redup. '''Tədəg-u'''.

'''Təjək, -jəgu''' ''n. I'' "Plant; herb, esp. medicinal."

'''Təkəbeh-u''' ''v.t.'' "Capture," redup. '''Tədəkəbeh-u'''.

'''Tərun, -runən''' ''n. II'' "Leaves of a tree."
* sg. '''Tərũreh''' "Leaf; page," pl. '''Tədərun''' "Leaves; pages."

'''Təməs, -məzən''' ''n. I'' "Village, pl. '''Tədəməs'''.

'''Tərãp, -rãbu''' ''n. I'' "Traveler," pl. '''Tədərãp'''.

'''Təs-u''', ''stat.'' "Five," redup. '''Tətəs-u''' "Five by five."

'''Tilə, -lan''' ''n. I'' "Dagger," pl. '''Tedilə'''.

'''Tunəw, -nəwən''' ''n. II'' "Horns." As a relational noun, "Peak, point." pl. '''Tudunəw'''.
* sg. '''Tunəweh''' "Horn; ferocity."

'''Tuswət, -wədən''' ''n. II'' "Someone's eyes," sg. '''Tuswəceh''' "An eye," pl. '''Tuduswət''' "Many eyes."

'''Ujək-u''' ''stat.'' "Covering," redup. '''Ulujək-u'''.

'''Ujəgĩcwir-u''' ''stat.'' "Covered in," redup. '''Ulujəgĩcwir-u'''.

'''Ulwəs, -wəsən''' ''n. I'' "Worm," pl. '''Ululwəs'''.

'''Ũrəweh-u''' ''v.t.'' "Cook," redup. '''Umũrəweh-u'''.

'''Uspəneh-u''' ''v.t.'' "Weave," redup. '''Uzuspəneh-u'''.

'''Wəgudeh-u''' ''v.t.'' "Give (something)," takes recipient as an optional oblique argument.
* redup. '''Wəgũkudeh-u''' "Give (something) up."
* applic. '''Nĩkudeh-u''' "Give (to someone)," takes thing given as an optional oblique argument.

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərer-u''' ''stat.'' "Four."

'''Wərĩhəj, -həjən''' ''n. I'' "House."

'''Wərjeh-u''' ''v.t.'' "Live in, dwell in," redup. '''Wərĩrjeh-u'''

'''Wərineh-u''' ''v.t.'' "Have dwelled in, hail from."

'''Wərn-u''' ''stat.'' "Bright," redup. "Wərwərn-u."

'''Wiłk, wiłkən''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T05:59:02Z

Nebulawindphone: /* Lexicon */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməgəmeh-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməgəmeh-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Singulatives in -reh not -rə.

'''Cəmə''' ''n. I'' "Body"

'''Cəməjəmə''' ''n. IV'' "Everyone"

'''Cər-u''' ''stat.'' "One"

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cec-u''' ''v.i.'' "Defecate," redup. '''Cecec-u'''.

'''Cigic''' ''quant'' "Some"

'''Ciri, -rjən''' ''n. I'' "Cat," pl. '''Ciciri,''' pauc. '''Cirirə''' "Litter of cats."

'''Curireh-u''' ''v.t.'' "Drink," redup. '''Cucurireh-u'''.

'''Curicic-u''' ''v.i.'' "Quench one's thirst,"
* applic. '''Wədəcuriciceh-u''' "Enjoy, benefit from."

'''Cur-u''' ''v.i.'' "Fly," redup. '''Təcur-u''' "Fly around."

'''Ǝlum-u''' ''v.i.'' "Get up; go; set off on foot" redup. '''Ǝləlum-u''' "Go around, go here and there."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''ərah, -rarən''' ''n. I'' "Lover." (< PP '''*nriax''' "spouse")
* pauc. '''ərarə''' "Courting couple."

'''əramjə, -ramjan''' ''n. I'' "Bride."

'''ərahwə, -rahwan''' ''n. I'' "Groom."

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Halas, -lazən''' ''n. I'' "Chief," pl. '''Haralas'''.

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Hədər-u''' ''stat.'' "Ten," redup. '''Hihədər-u''' "Ten by ten."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh, -nərən''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Həreh''' ''pron.'' "I."
* '''Hərə''' "You and I."
* '''Hər''' "We."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed," redup. '''Hərəsur-u''' "Bleed out,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnur-u''' ''v.i.'' "Make camp," redup. '''Herẽnur-u'''.

'''Hẽnurwərəj, -rəjən''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hilucer-u''' ''stat.'' "Nine."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming," redup. '''Hĩhĩmət-u'''.

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Ira''' ''pron.'' "He, she,"
* '''Irar''' "They."

'''Ireh''' ''pron.'' "It,"
* '''Ir''' "They (inanimate)."

'''Irug-u''' ''v.i.'' "Go," redup. '''Irirug-u'''.

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Irukuj-u''' ''v.i.'' "Have gone, be gone."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kaməl''' ''n. I'' "Pan," pl. "Kəkaməl."

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kasə''' ''n. IV'' "Wind."

'''Katərə-w''' ''v.i.'' "Go hunting," redup. '''Kəgatərə-w'''.

'''Katərẽceh-u''' ''v.t.'' "Catch, hunt successfully," redup. '''Kəgatərẽceh-u'''.

'''Katəreh-u''' ''v.t.'' "Hunt for," patient trigger '''Katərec-u''', redup. '''Kəgatəreh-u'''.

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc, kəjən''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak, discuss," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədərət''' ''n. I'' "Message."

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərah-u''' ''stat.'' "Dark," redup. '''Kəgərah-u'''.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Kəstin, -tinən''' ''n. I'' "Knife," pl. '''Kəgəstin'''.
* pauc. '''Kəstĩrə''' "Plough."

'''Kũh-u''' ''stat.'' "Fat," redup '''Kugũh-u'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Ləlah-u''' ''stat.'' "Eight."

'''Lərut, -rudən''' ''n. I'' "Ocean," pl. '''Lələrut'''.

'''Ligwə, -gwan''' ''n. I'' "Parent," pauc. '''Ligwərə''' "Mother and son; father and daughter," pl. '''Liligwə'''.

'''Ligwəligwə''' ''n. IV'' "The Ancestors."

'''Ligwə-w''' ''stat.'' ''Who is (someone's) parent."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Məgəmeh-u''' ''v.t.'' "Break."
* redup. '''Məməgəmeh-u''' "Defeat."

'''Məgis-u''' ''stat.'' "Holding (something), carrying (something)."

'''Məmir-u''' ''stat.'' "Six."

'''Mənə''' ''n. I'' "Bay."

'''Məpeh-u''' ''v.t.'' "Eat," patient trigger '''Məcic-u'''.

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Mər-u''' ''stat.'' "Three."

'''Medalə, -lan''' ''n. I'' "Opponent," pauc. '''Medalərə''' "One another's opponents," pl. '''Memedalə'''.

'''Milmeh-u''' ''v.t.'' "Tell (a story)," redup. '''Mimilmeh-u'''.

'''Minə''' ''n. I'' "Woman," pl. '''Miminə'''.

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Naheh''' ''pron.'' "You,"
* '''Nahə''' "You two,"
* '''Nəra''' "You all."

'''Nətər-u''' ''stat.'' "Seven."

'''Nigət, -gətən''' ''n. III'' "Berries."

'''Nizəj-u''' ''v.t.'' "Trades or barters with."

'''Nuheh-u''' ''v.t.'' "Pour out," redup. '''Nunuheh-u'''.

'''Nuh-u''' ''stat.'' "Whistling."

'''Nugim, -imən''' ''n. I'', pauc. '''Nugĩrə''', pl. '''Nunugim'''.

'''Pahu''' ''det.'' "What sort?"

'''Pəcur-u''' ''stat.'' "Two."

'''Pəka''' ''det.'' "Which?"

'''Pəkmjeh-u''' ''v.t.'' "Hunt," patient trigger '''Pəkmic-u'''.
* caus. '''Pəkmirazeh-u''' "Banish, outlaw, cause to be hunted."

'''Pəmə''' ''n. I.'' "Father," pl. '''Pəbəmə''',
* pauc. '''Pəmərə''' "Father and son."

'''Pəmir-u''' ''v.i.'' "Call out, shout," redup. '''Pəbəmir-u'''.

'''Pətu''' ''quant.'' "How many?"

'''Piri, -rjən''' ''n. I'' "Day," pl. '''Pibiri'''.

'''Pusmeh-u''' ''v.t.'' "Swallow," redup. '''Pupusmeh-u'''.

'''Pũt, -dən''' ''n. I'' "Bridge," pl. '''Pũpũdən'''.

'''Sirəm, -əmən''' ''n. I'', pl. '''Sizirəm'''.

'''Sism-u,''' ''v.i.'' "Sink," redup. '''Sizism-u'''.

'''Sumeh''' ''quant.'' "Every, all."

'''Sunəj, -nəjən''' ''n. I'' "River," pl. '''Suzunəj'''.

'''Swəłc-u''' ''stat.'' "Sick," redup. '''Swəzwəłc-u'''.

'''Swəłcwəłcu''' ''stat.'' "Crippled."

'''Təg-u''' ''stat.'' "Blue," redup. '''Tədəg-u'''.

'''Təjək, -jəgu''' ''n. I'' "Plant; herb, esp. medicinal."

'''Təkəbeh-u''' ''v.t.'' "Capture," redup. '''Tədəkəbeh-u'''.

'''Tərun, -runən''' ''n. II'' "Leaves of a tree."
* sg. '''Tərũreh''' "Leaf; page," pl. '''Tədərun''' "Leaves; pages."

'''Təməs, -məzən''' ''n. I'' "Village, pl. '''Tədəməs'''.

'''Tərãp, -rãbu''' ''n. I'' "Traveler," pl. '''Tədərãp'''.

'''Təs-u''', ''stat.'' "Five," redup. '''Tətəs-u''' "Five by five."

'''Tilə, -lan''' ''n. I'' "Dagger," pl. '''Tedilə'''.

'''Tunəw, -nəwən''' ''n. II'' "Horns." As a relational noun, "Peak, point." pl. '''Tudunəw'''.
* sg. '''Tunəweh''' "Horn; ferocity."

'''Tuswət, -wədən''' ''n. II'' "Someone's eyes," sg. '''Tuswəceh''' "An eye," pl. '''Tuduswət''' "Many eyes."

'''Ujək-u''' ''stat.'' "Covering," redup. '''Ulujək-u'''.

'''Ujəgĩcwir-u''' ''stat.'' "Covered in," redup. '''Ulujəgĩcwir-u'''.

'''Ulwəs, -wəsən''' ''n. I'' "Worm," pl. '''Ululwəs'''.

'''Ũrəweh-u''' ''v.t.'' "Cook," redup. '''Umũrəweh-u'''.

'''Uspəneh-u''' ''v.t.'' "Weave," redup. '''Uzuspəneh-u'''.

'''Wəgudeh-u''' ''v.t.'' "Give (something)," takes recipient as an optional oblique argument.
* redup. '''Wəgũkudeh-u''' "Give (something) up."
* applic. '''Nĩkudeh-u''' "Give (to someone)," takes thing given as an optional oblique argument.

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərer-u''' ''stat.'' "Four."

'''Wərĩhəj, -həjən''' ''n. I'' "House."

'''Wərjeh-u''' ''v.t.'' "Live in, dwell in," redup. '''Wərĩrjeh-u'''

'''Wərineh-u''' ''v.t.'' "Have dwelled in, hail from."

'''Wərn-u''' ''stat.'' "Bright," redup. "Wərwərn-u."

'''Wiłk, wiłkən''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T05:57:47Z

Nebulawindphone: /* Lexicon */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməgəmeh-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməgəmeh-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Singulatives in -reh not -rə.

'''Cəmə''' ''n. I'' "Body"

'''Cəməjəmə''' ''n. IV'' "Everyone"

'''Cər-u''' ''stat.'' "One"

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cec-u''' ''v.i.'' "Defecate," redup. '''Cecec-u'''.

'''Cigic''' ''quant'' "Some"

'''Ciri, -rjən''' ''n. I'' "Cat," pl. '''Ciciri,''' pauc. '''Cirirə''' "Litter of cats."

'''Curireh-u''' ''v.t.'' "Drink," redup. '''Cucurireh-u'''.

'''Curicic-u''' ''v.i.'' "Quench one's thirst,"
* applic. '''Wədəcuriciceh-u''' "Enjoy, benefit from."

'''Cur-u''' ''v.i.'' "Fly," redup. '''Təcur-u''' "Fly around."

'''Ǝlum-u''' ''v.i.'' "Get up; go; set off on foot" redup. '''Ǝləlum-u''' "Go around, go here and there."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''ərah, -rarən''' ''n. I'' "Lover." (< PP '''*nriax''' "spouse")
* pauc. '''ərarə''' "Courting couple."

'''əramjə, -ramjan''' ''n. I'' "Bride."

'''ərahwə, -rahwan''' ''n. I'' "Groom."

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Halas, -lazən''' ''n. I'' "Chief," pl. '''Haralas'''.

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Hədər-u''' ''stat.'' "Ten," redup. '''Hihədər-u''' "Ten by ten."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh, -nərən''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Həreh''' ''pron.'' "I."
* '''Hərə''' "You and I."
* '''Hər''' "We."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed," redup. '''Hərəsur-u''' "Bleed out,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnur-u''' ''v.i.'' "Make camp," redup. '''Herẽnur-u'''.

'''Hẽnurwərəj, -rəjən''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hilucer-u''' ''stat.'' "Nine."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming," redup. '''Hĩhĩmət-u'''.

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Ira''' ''pron.'' "He, she,"
* '''Irar''' "They."

'''Ireh''' ''pron.'' "It,"
* '''Ir''' "They (inanimate)."

'''Irug-u''' ''v.i.'' "Go," redup. '''Irirug-u'''.

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Irukuj-u''' ''v.i.'' "Have gone, be gone."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kaməl''' ''n. I'' "Pan," pl. "Kəkaməl."

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kasə''' ''n. IV'' "Wind."

'''Katərə-w''' ''v.i.'' "Go hunting," redup. '''Kəgatərə-w'''.

'''Katərẽceh-u''' ''v.t.'' "Catch, hunt successfully," redup. '''Kəgatərẽceh-u'''.

'''Katəreh-u''' ''v.t.'' "Hunt for," patient trigger '''Katərec-u''', redup. '''Kəgatəreh-u'''.

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc, kəjən''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak, discuss," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərah-u''' ''stat.'' "Dark," redup. '''Kəgərah-u'''.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Kəstin, -tinən''' ''n. I'' "Knife," pl. '''Kəgəstin'''.
* pauc. '''Kəstĩrə''' "Plough."

'''Kũh-u''' ''stat.'' "Fat," redup '''Kugũh-u'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Ləlah-u''' ''stat.'' "Eight."

'''Lərut, -rudən''' ''n. I'' "Ocean," pl. '''Lələrut'''.

'''Ligwə, -gwan''' ''n. I'' "Parent," pauc. '''Ligwərə''' "Mother and son; father and daughter," pl. '''Liligwə'''.

'''Ligwəligwə''' ''n. IV'' "The Ancestors."

'''Ligwə-w''' ''stat.'' ''Who is (someone's) parent."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Məgəmeh-u''' ''v.t.'' "Break."
* redup. '''Məməgəmeh-u''' "Defeat."

'''Məgis-u''' ''stat.'' "Holding (something), carrying (something)."

'''Məmir-u''' ''stat.'' "Six."

'''Mənə''' ''n. I'' "Bay."

'''Məpeh-u''' ''v.t.'' "Eat," patient trigger '''Məcic-u'''.

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Mər-u''' ''stat.'' "Three."

'''Medalə, -lan''' ''n. I'' "Opponent," pauc. '''Medalərə''' "One another's opponents," pl. '''Memedalə'''.

'''Milmeh-u''' ''v.t.'' "Tell (a story)," redup. '''Mimilmeh-u'''.

'''Minə''' ''n. I'' "Woman," pl. '''Miminə'''.

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Naheh''' ''pron.'' "You,"
* '''Nahə''' "You two,"
* '''Nəra''' "You all."

'''Nətər-u''' ''stat.'' "Seven."

'''Nigət, -gətən''' ''n. III'' "Berries."

'''Nizəj-u''' ''v.t.'' "Trades or barters with."

'''Nuheh-u''' ''v.t.'' "Pour out," redup. '''Nunuheh-u'''.

'''Nuh-u''' ''stat.'' "Whistling."

'''Nugim, -imən''' ''n. I'', pauc. '''Nugĩrə''', pl. '''Nunugim'''.

'''Pahu''' ''det.'' "What sort?"

'''Pəcur-u''' ''stat.'' "Two."

'''Pəka''' ''det.'' "Which?"

'''Pəkmjeh-u''' ''v.t.'' "Hunt," patient trigger '''Pəkmic-u'''.
* caus. '''Pəkmirazeh-u''' "Banish, outlaw, cause to be hunted."

'''Pəmə''' ''n. I.'' "Father," pl. '''Pəbəmə''',
* pauc. '''Pəmərə''' "Father and son."

'''Pəmir-u''' ''v.i.'' "Call out, shout," redup. '''Pəbəmir-u'''.

'''Pətu''' ''quant.'' "How many?"

'''Piri, -rjən''' ''n. I'' "Day," pl. '''Pibiri'''.

'''Pusmeh-u''' ''v.t.'' "Swallow," redup. '''Pupusmeh-u'''.

'''Pũt, -dən''' ''n. I'' "Bridge," pl. '''Pũpũdən'''.

'''Sirəm, -əmən''' ''n. I'', pl. '''Sizirəm'''.

'''Sism-u,''' ''v.i.'' "Sink," redup. '''Sizism-u'''.

'''Sumeh''' ''quant.'' "Every, all."

'''Sunəj, -nəjən''' ''n. I'' "River," pl. '''Suzunəj'''.

'''Swəłc-u''' ''stat.'' "Sick," redup. '''Swəzwəłc-u'''.

'''Swəłcwəłcu''' ''stat.'' "Crippled."

'''Təg-u''' ''stat.'' "Blue," redup. '''Tədəg-u'''.

'''Təjək, -jəgu''' ''n. I'' "Plant; herb, esp. medicinal."

'''Təkəbeh-u''' ''v.t.'' "Capture," redup. '''Tədəkəbeh-u'''.

'''Tərun, -runən''' ''n. II'' "Leaves of a tree."
* sg. '''Tərũreh''' "Leaf; page," pl. '''Tədərun''' "Leaves; pages."

'''Təməs, -məzən''' ''n. I'' "Village, pl. '''Tədəməs'''.

'''Tərãp, -rãbu''' ''n. I'' "Traveler," pl. '''Tədərãp'''.

'''Təs-u''', ''stat.'' "Five," redup. '''Tətəs-u''' "Five by five."

'''Tilə, -lan''' ''n. I'' "Dagger," pl. '''Tedilə'''.

'''Tunəw, -nəwən''' ''n. II'' "Horns." As a relational noun, "Peak, point." pl. '''Tudunəw'''.
* sg. '''Tunəweh''' "Horn; ferocity."

'''Tuswət, -wədən''' ''n. II'' "Someone's eyes," sg. '''Tuswəceh''' "An eye," pl. '''Tuduswət''' "Many eyes."

'''Ujək-u''' ''stat.'' "Covering," redup. '''Ulujək-u'''.

'''Ujəgĩcwir-u''' ''stat.'' "Covered in," redup. '''Ulujəgĩcwir-u'''.

'''Ulwəs, -wəsən''' ''n. I'' "Worm," pl. '''Ululwəs'''.

'''Ũrəweh-u''' ''v.t.'' "Cook," redup. '''Umũrəweh-u'''.

'''Uspəneh-u''' ''v.t.'' "Weave," redup. '''Uzuspəneh-u'''.

'''Wəgudeh-u''' ''v.t.'' "Give (something)," takes recipient as an optional oblique argument.
* redup. '''Wəgũkudeh-u''' "Give (something) up."
* applic. '''Nĩkudeh-u''' "Give (to someone)," takes thing given as an optional oblique argument.

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərer-u''' ''stat.'' "Four."

'''Wərĩhəj, -həjən''' ''n. I'' "House."

'''Wərjeh-u''' ''v.t.'' "Live in, dwell in," redup. '''Wərĩrjeh-u'''

'''Wərineh-u''' ''v.t.'' "Have dwelled in, hail from."

'''Wərn-u''' ''stat.'' "Bright," redup. "Wərwərn-u."

'''Wiłk, wiłkən''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T05:20:45Z

Nebulawindphone: /* Bodypart terms */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməgəmeh-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməgəmeh-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Cəmə''' ''n. I'' "Body"

'''Cəməjəmə''' ''n. IV'' "Everyone"

'''Cər-u''' ''stat.'' "One"

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cec-u''' ''v.i.'' "Defecate," redup. '''Cecec-u'''.

'''Cigic''' ''quant'' "Some"

'''Ciri, -rjən''' ''n. I'' "Cat," pl. '''Ciciri,''' pauc. '''Cirirə''' "Litter of cats."

'''Curireh-u''' ''v.t.'' "Drink," redup. '''Cucurireh-u'''.

'''Curicic-u''' ''v.i.'' "Quench one's thirst,"
* applic. '''Wədəcuriciceh-u''' "Enjoy, benefit from."

'''Cur-u''' ''v.i.'' "Fly," redup. '''Təcur-u''' "Fly around."

'''Ǝlum-u''' ''v.i.'' "Get up; go; set off on foot" redup. '''Ǝləlum-u''' "Go around, go here and there."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''ərah, -rarən''' ''n. I'' "Lover." (< PP '''*nriax''' "spouse")
* pauc. '''ərarə''' "Courting couple."

'''əramjə, -ramjan''' ''n. I'' "Bride."

'''ərahwə, -rahwan''' ''n. I'' "Groom."

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Halas, -lazən''' ''n. I'' "Chief," pl. '''Haralas'''.

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Hədər-u''' ''stat.'' "Ten," redup. '''Hihədər-u''' "Ten by ten."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh, -nərən''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Həreh''' ''pron.'' "I."
* '''Hərə''' "You and I."
* '''Hər''' "We."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed," redup. '''Hərəsur-u''' "Bleed out,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnur-u''' ''v.i.'' "Make camp," redup. '''Herẽnur-u'''.

'''Hẽnurwərəj, -rəjən''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hilucer-u''' ''stat.'' "Nine."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming," redup. '''Hĩhĩmət-u'''.

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Ira''' ''pron.'' "He, she,"
* '''Irar''' "They."

'''Ireh''' ''pron.'' "It,"
* '''Ir''' "They (inanimate)."

'''Irug-u''' ''v.i.'' "Go," redup. '''Irirug-u'''.

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kaməl''' ''n. I'' "Pan," pl. "Kəkaməl."

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kasə''' ''n. IV'' "Wind."

'''Katərə-w''' ''v.i.'' "Go hunting," redup. '''Kəgatərə-w'''.

'''Katərẽceh-u''' ''v.t.'' "Catch, hunt successfully," redup. '''Kəgatərẽceh-u'''.

'''Katəreh-u''' ''v.t.'' "Hunt for," patient trigger '''Katərec-u''', redup. '''Kəgatəreh-u'''.

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc, kəjən''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak, discuss," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərah-u''' ''stat.'' "Dark," redup. '''Kəgərah-u'''.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Kəstin, -tinən''' ''n. I'' "Knife," pl. '''Kəgəstin'''.
* pauc. '''Kəstĩrə''' "Plough."

'''Kũh-u''' ''stat.'' "Fat," redup '''Kugũh-u'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Ləlah-u''' ''stat.'' "Eight."

'''Lərut, -rudən''' ''n. I'' "Ocean," pl. '''Lələrut'''.

'''Ligwə, -gwan''' ''n. I'' "Parent," pauc. '''Ligwərə''' "Mother and son; father and daughter," pl. '''Liligwə'''.

'''Ligwəligwə''' ''n. IV'' "The Ancestors."

'''Ligwə-w''' ''stat.'' ''Who is (someone's) parent."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Məgəmeh-u''' ''v.t.'' "Break."
* redup. '''Məməgəmeh-u''' "Defeat."

'''Məgis-u''' ''stat.'' "Holding (something), carrying (something)."

'''Məmir-u''' ''stat.'' "Six."

'''Mənə''' ''n. I'' "Bay."

'''Məpeh-u''' ''v.t.'' "Eat," patient trigger '''Məcic-u'''.

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Mər-u''' ''stat.'' "Three."

'''Medalə, -lan''' ''n. I'' "Opponent," pauc. '''Medalərə''' "One another's opponents," pl. '''Memedalə'''.

'''Milmeh-u''' ''v.t.'' "Tell (a story)," redup. '''Mimilmeh-u'''.

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Naheh''' ''pron.'' "You,"
* '''Nahə''' "You two,"
* '''Nəra''' "You all."

'''Nətər-u''' ''stat.'' "Seven."

'''Nigət, -gətən''' ''n. III'' "Berries."

'''Nizəj-u''' ''v.t.'' "Trades or barters with."

'''Nuheh-u''' ''v.t.'' "Pour out," redup. '''Nunuheh-u'''.

'''Nugim, -imən''' ''n. I'', pauc. '''Nugĩrə''', pl. '''Nunugim'''.

'''Pahu''' ''det.'' "What sort?"

'''Pəcur-u''' ''stat.'' "Two."

'''Pəka''' ''det.'' "Which?"

'''Pəmə''' ''n. I.'' "Father," pl. '''Pəbəmə''',
* pauc. '''Pəmərə''' "Father and son."

'''Pəmir-u''' ''v.i.'' "Call out, shout," redup. '''Pəbəmir-u'''.

'''Pətu''' ''quant.'' "How many?"

'''Piri, -rjən''' ''n. I'' "Day," pl. '''Pibiri'''.

'''Pusmeh-u''' ''v.t.'' "Swallow," redup. '''Pupusmeh-u'''.

'''Pũt, -dən''' ''n. I'' "Bridge," pl. '''Pũpũdən'''.

'''Sirəm, -əmən''' ''n. I'', pl. '''Sizirəm'''.

'''Sism-u,''' ''v.i.'' "Sink," redup. '''Sizism-u'''.

'''Sumeh''' ''quant.'' "Every, all."

'''Sunəj, -nəjən''' ''n. I'' "River," pl. '''Suzunəj'''.

'''Swəłc-u''' ''stat.'' "Sick," redup. '''Swəzwəłc-u'''.

'''Swəłcwəłcu''' ''stat.'' "Crippled."

'''Təg-u''' ''stat.'' "Blue," redup. '''Tədəg-u'''.

'''Təjək, -jəgu''' ''n. I'' "Plant; herb, esp. medicinal."

'''Təkəbeh-u''' ''v.t.'' "Capture," redup. '''Tədəkəbeh-u'''.

'''Tərun, -runən''' ''n. II'' "Leaves of a tree."
* sg. '''Tərũreh''' "Leaf; page," pl. '''Tədərun''' "Leaves; pages."

'''Təməs, -məzən''' ''n. I'' "Village, pl. '''Tədəməs'''.

'''Tərãp, -rãbu''' ''n. I'' "Traveler," pl. '''Tədərãp'''.

'''Təs-u''', ''stat.'' "Five," redup. '''Tətəs-u''' "Five by five."

'''Tilə, -lan''' ''n. I'' "Dagger," pl. '''Tedilə'''.

'''Tunəw, -nəwən''' ''n. II'' "Horns." As a relational noun, "Peak, point." pl. '''Tudunəw'''.
* sg. '''Tunəweh''' "Horn; ferocity."

'''Tuswət, -wədən''' ''n. II'' "Someone's eyes," sg. '''Tuswəceh''' "An eye," pl. '''Tuduswət''' "Many eyes."

'''Ujək-u''' ''stat.'' "Covering," redup. '''Ulujək-u'''.

'''Ujəgĩcwir-u''' ''stat.'' "Covered in," redup. '''Ulujəgĩcwir-u'''.

'''Ulwəs, -wəsən''' ''n. I'' "Worm," pl. '''Ululwəs'''.

'''Ũrəweh-u''' ''v.t.'' "Cook," redup. '''Umũrəweh-u'''.

'''Uspəneh-u''' ''v.t.'' "Weave," redup. '''Uzuspəneh-u'''.

'''Wəgudeh-u''' ''v.t.'' "Give (something)," takes recipient as an optional oblique argument.
* redup. '''Wəgũkudeh-u''' "Give (something) up."
* applic. '''Nĩkudeh-u''' "Give (to someone)," takes thing given as an optional oblique argument.

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərer-u''' ''stat.'' "Four."

'''Wərĩhəj, -həjən''' ''n. I'' "House."

'''Wərn-u''' ''stat.'' "Bright," redup. "Wərwərn-u."

'''Wiłk, wiłkən''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T05:18:35Z

Nebulawindphone: /* Bodypart terms */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməgəmeh-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməgəmeh-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Cəmə''' ''n. I'' "Body"

'''Cəməjəmə''' ''n. IV'' "Everyone"

'''Cər-u''' ''stat.'' "One"

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cec-u''' ''v.i.'' "Defecate," redup. '''Cecec-u'''.

'''Cigic''' ''quant'' "Some"

'''Ciri, -rjən''' ''n. I'' "Cat," pl. '''Ciciri,''' pauc. '''Cirirə''' "Litter of cats."

'''Curireh-u''' ''v.t.'' "Drink," redup. '''Cucurireh-u'''.

'''Curicic-u''' ''v.i.'' "Quench one's thirst,"
* applic. '''Wədəcuriciceh-u''' "Enjoy, benefit from."

'''Ǝlum-u''' ''v.i.'' "Get up; set off on foot" perf. '''Ǝləlum-u'''.

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''ərah, -rarən''' ''n. I'' "Lover." (< PP '''*nriax''' "spouse")
* pauc. '''ərarə''' "Courting couple."

'''əramjə, -ramjan''' ''n. I'' "Bride."

'''ərahwə, -rahwan''' ''n. I'' "Groom."

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Halas, -lazən''' ''n. I'' "Chief," pl. '''Haralas'''.

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Hədər-u''' ''stat.'' "Ten," redup. '''Hihədər-u''' "Ten by ten."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh, -nərən''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Həreh''' ''pron.'' "I."
* '''Hərə''' "You and I."
* '''Hər''' "We."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnur-u''' ''v.i.'' "Make camp," redup. '''Herẽnur-u'''.

'''Hẽnurwərəj, -rəjən''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hilucer-u''' ''stat.'' "Nine."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming," redup. '''Hĩhĩmət-u'''.

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Ira''' ''pron.'' "He, she,"
* '''Irar''' "They."

'''Ireh''' ''pron.'' "It,"
* '''Ir''' "They (inanimate)."

'''Irug-u''' ''v.i.'' "Go," redup. '''Irirug-u'''.

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kaməl''' ''n. I'' "Pan," pl. "Kəkaməl."

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kasə''' ''n. IV'' "Wind."

'''Katərə-w''' ''v.i.'' "Go hunting," redup. '''Kəgatərə-w'''.

'''Katərẽceh-u''' ''v.t.'' "Catch, hunt successfully," redup. '''Kəgatərẽceh-u'''.

'''Katəreh-u''' ''v.t.'' "Hunt for," patient trigger '''Katərec-u''', redup. '''Kəgatəreh-u'''.

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc, kəjən''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak, discuss," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərah-u''' ''stat.'' "Dark," redup. '''Kəgərah-u'''.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Kəstin, -tinən''' ''n. I'' "Knife," pl. '''Kəgəstin'''.
* pauc. '''Kəstĩrə''' "Plough."

'''Kũh-u''' ''stat.'' "Fat," redup '''Kugũh-u'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Ləlah-u''' ''stat.'' "Eight."

'''Lərut, -rudən''' ''n. I'' "Ocean," pl. '''Lələrut'''.

'''Ligwə, -gwan''' ''n. I'' "Parent," pauc. '''Ligwərə''' "Mother and son; father and daughter," pl. '''Liligwə'''.

'''Ligwəligwə''' ''n. IV'' "The Ancestors."

'''Ligwə-w''' ''stat.'' ''Who is (someone's) parent."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Məgəmeh-u''' ''v.t.'' "Break."
* redup. '''Məməgəmeh-u''' "Defeat."

'''Məgis-u''' ''stat.'' "Holding (something), carrying (something)."

'''Məmir-u''' ''stat.'' "Six."

'''Mənə''' ''n. I'' "Bay."

'''Məpeh-u''' ''v.t.'' "Eat," patient trigger '''Məcic-u'''.

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Mər-u''' ''stat.'' "Three."

'''Medalə, -lan''' ''n. I'' "Opponent," pauc. '''Medalərə''' "One another's opponents," pl. '''Memedalə'''.

'''Milmeh-u''' ''v.t.'' "Tell (a story)," redup. '''Mimilmeh-u'''.

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Naheh''' ''pron.'' "You,"
* '''Nahə''' "You two,"
* '''Nəra''' "You all."

'''Nətər-u''' ''stat.'' "Seven."

'''Nigət, -gətən''' ''n. III'' "Berries."

'''Nizəj-u''' ''v.t.'' "Trades or barters with."

'''Nuheh-u''' ''v.t.'' "Pour out," redup. '''Nunuheh-u'''.

'''Nugim, -imən''' ''n. I'', pauc. '''Nugĩrə''', pl. '''Nunugim'''.

'''Pahu''' ''det.'' "What sort?"

'''Pəcur-u''' ''stat.'' "Two."

'''Pəka''' ''det.'' "Which?"

'''Pəmə''' ''n. I.'' "Father," pl. '''Pəbəmə''',
* pauc. '''Pəmərə''' "Father and son."

'''Pəmir-u''' ''v.i.'' "Call out, shout," redup. '''Pəbəmir-u'''.

'''Pətu''' ''quant.'' "How many?"

'''Piri, -rjən''' ''n. I'' "Day," pl. '''Pibiri'''.

'''Pusmeh-u''' ''v.t.'' "Swallow," redup. '''Pupusmeh-u'''.

'''Pũt, -dən''' ''n. I'' "Bridge," pl. '''Pũpũdən'''.

'''Sirəm, -əmən''' ''n. I'', pl. '''Sizirəm'''.

'''Sism-u,''' ''v.i.'' "Sink," redup. '''Sizism-u'''.

'''Sumeh''' ''quant.'' "Every, all."

'''Sunəj, -nəjən''' ''n. I'' "River," pl. '''Suzunəj'''.

'''Swəłc-u''' ''stat.'' "Sick," redup. '''Swəzwəłc-u'''.

'''Swəłcwəłcu''' ''stat.'' "Crippled."

'''Təg-u''' ''stat.'' "Blue," redup. '''Tədəg-u'''.

'''Təjək, -jəgu''' ''n. I'' "Plant; herb, esp. medicinal."

'''Təkəbeh-u''' ''v.t.'' "Capture," redup. '''Tədəkəbeh-u'''.

'''Tərun, -runən''' ''n. II'' "Leaves of a tree."
* sg. '''Tərũreh''' "Leaf; page," pl. '''Tədərun''' "Leaves; pages."

'''Təməs, -məzən''' ''n. I'' "Village, pl. '''Tədəməs'''.

'''Tərãp, -rãbu''' ''n. I'' "Traveler," pl. '''Tədərãp'''.

'''Təs-u''', ''stat.'' "Five," redup. '''Tətəs-u''' "Five by five."

'''Tilə, -lan''' ''n. I'' "Dagger," pl. '''Tedilə'''.

'''Tunəw, -nəwən''' ''n. II'' "Horns." As a relational noun, "Peak, point." pl. '''Tudunəw'''.
* sg. '''Tunəweh''' "Horn; ferocity."

'''Tuswət, -wədən''' ''n. II'' "Someone's eyes," sg. '''Tuswəceh''' "An eye," pl. '''Tuduswət''' "Many eyes."

'''Ujək-u''' ''stat.'' "Covering," redup. '''Ulujək-u'''.

'''Ujəgĩcwir-u''' ''stat.'' "Covered in," redup. '''Ulujəgĩcwir-u'''.

'''Ulwəs, -wəsən''' ''n. I'' "Worm," pl. '''Ululwəs'''.

'''Ũrəweh-u''' ''v.t.'' "Cook," redup. '''Umũrəweh-u'''.

'''Uspəneh-u''' ''v.t.'' "Weave," redup. '''Uzuspəneh-u'''.

'''Wəgudeh-u''' ''v.t.'' "Give (something)," takes recipient as an optional oblique argument.
* redup. '''Wəgũkudeh-u''' "Give (something) up."
* applic. '''Nĩkudeh-u''' "Give (to someone)," takes thing given as an optional oblique argument.

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərer-u''' ''stat.'' "Four."

'''Wərĩhəj, -həjən''' ''n. I'' "House."

'''Wərn-u''' ''stat.'' "Bright," redup. "Wərwərn-u."

'''Wiłk, wiłkən''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T05:02:28Z

Nebulawindphone: /* Grounding */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməgəmeh-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməgəmeh-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Cəmə''' ''n. I'' "Body"

'''Cəməjəmə''' ''n. IV'' "Everyone"

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cec-u''' ''v.i.'' "Defecate," redup. '''Cecec-u'''.

'''Cigic''' ''quant'' "Some"

'''Ciri, -rjən''' ''n. I'' "Cat," pl. '''Ciciri,''' pauc. '''Cirirə''' "Litter of cats."

'''Curireh-u''' ''v.t.'' "Drink," redup. '''Cucurireh-u'''.

'''Curicic-u''' ''v.i.'' "Quench one's thirst,"
* applic. '''Wədəcuriciceh-u''' "Enjoy, benefit from."

'''Ǝlum-u''' ''v.i.'' "Get up; set off on foot" perf. '''Ǝləlum-u'''.

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''ərah, -rarən''' ''n. I'' "Lover." (< PP '''*nriax''' "spouse")
* pauc. '''ərarə''' "Courting couple."

'''əramjə, -ramjan''' ''n. I'' "Bride."

'''ərahwə, -rahwan''' ''n. I'' "Groom."

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Halas, -lazən''' ''n. I'' "Chief," pl. '''Haralas'''.

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh, -nərən''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnur-u''' ''v.i.'' "Make camp," redup. '''Herẽnur-u'''.

'''Hẽnurwərəj, -rəjən''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go," redup. '''Irirug-u'''.

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kaməl''' ''n. I'' "Pan," pl. "Kəkaməl."

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kasə''' ''n. IV'' "Wind."

'''Katərə-w''' ''v.i.'' "Go hunting," redup. '''Kəgatərə-w'''.

'''Katərẽceh-u''' ''v.t.'' "Catch, hunt successfully," redup. '''Kəgatərẽceh-u'''.

'''Katəreh-u''' ''v.t.'' "Hunt for," patient trigger '''Katərec-u''', redup. '''Kəgatəreh-u'''.

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc, kəjən''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak, discuss," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Kəstin, -tinən''' ''n. I'' "Knife," pl. '''Kəgəstin'''.
* pauc. '''Kəstĩrə''' "Plough."

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lərut, -rudən''' ''n. I'' "Ocean," pl. '''Lələrut'''.

'''Ligwə, -gwan''' ''n. I'' "Parent," pauc. '''Ligwərə''' "Mother and son; father and daughter," pl. '''Liligwə'''.

'''Ligwəligwə''' ''n. IV'' "The Ancestors."

'''Ligwə-w''' ''stat.'' ''Who is (someone's) parent."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Məgəmeh-u''' ''v.t.'' "Break."
* redup. '''Məməgəmeh-u''' "Defeat."

'''Məgis-u''' ''stat.'' "Holding (something), carrying (something)."

'''Mənə''' ''n. I'' "Bay."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Medalə, -lan''' ''n. I'' "Opponent," pauc. '''Medalərə''' "One another's opponents," pl. '''Memedalə'''.

'''Milmeh-u''' ''v.t.'' "Tell (a story)," redup. '''Mimilmeh-u'''.

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Nigət, -gətən''' ''n. III'' "Berries."

'''Nizəj-u''' ''v.t.'' "Trades or barters with."

'''Nuheh-u''' ''v.t.'' "Pour out," redup. '''Nunuheh-u'''.

'''Nugim, -imən''' ''n. I'', pauc. '''Nugĩrə''', pl. '''Nunugim'''.

'''Pahu''' ''det.'' "What sort?"

'''Pəka''' ''det.'' "Which?"

'''Pəmir-u''' ''v.i.'' "Call out, shout," redup. '''Pəbəmir-u'''.

'''Pətu''' ''quant.'' "How many?"

'''Pusmeh-u''' ''v.t.'' "Swallow," redup. '''Pupusmeh-u'''.

'''Sirəm, -əmən''' ''n. I'', pl. '''Sizirəm'''.

'''Sism-u,''' ''v.i.'' "Sink," redup. '''Sizism-u'''.

'''Sumeh''' ''quant.'' "Every, all."

'''Sunəj, -nəjən''' ''n. I'' "River," pl. '''Suzunəj'''.

'''Swəłc-u''' ''stat.'' "Sick," redup. '''Swəzwəłc-u'''.

'''Swəłcwəłcu''' ''stat.'' "Crippled."

'''Təg-u''' ''stat.'' "Blue," redup. '''Tədəg-u'''.

'''Təjək, -jəgu''' ''n. I'' "Plant; herb, esp. medicinal."

'''Təkəbeh-u''' ''v.t.'' "Capture," redup. '''Tədəkəbeh-u'''.

'''Tərun, -runən''' ''n. II'' "Leaves of a tree."
* sg. '''Tərũreh''' "Leaf; page," pl. '''Tədərun''' "Leaves; pages."

'''Təməs, -məzən''' ''n. I'' "Village, pl. '''Tədəməs'''.

'''Tərãp, -rãbu''' ''n. I'' "Traveler," pl. '''Tədərãp'''.

'''Tunəw, -nəwən''' ''n. II'' "Horns." As a relational noun, "Peak, point." pl. '''Tudunəw'''.
* sg. '''Tunəweh''' "Horn; ferocity."

'''Tuswət, -wədən''' ''n. II'' "Someone's eyes," sg. '''Tuswəceh''' "An eye," pl. '''Tuduswət''' "Many eyes."

'''Ujəku''' ''stat.'' "Covering."

'''Ulwəs, -wəsən''' ''n. I'' "Worm," pl. '''Ululwəs'''.

'''Ũrəweh-u''' ''v.t.'' "Cook," redup. '''Umũrəweh-u'''.

'''Uspəneh-u''' ''v.t.'' "Weave," redup. '''Uzuspəneh-u'''.

'''Wəgudeh-u''' ''v.t.'' "Give (something)," takes recipient as an optional oblique argument.
* redup. '''Wəgũkudeh-u''' "Give (something) up."
* applic. '''Nĩkudeh-u''' "Give (to someone)," takes thing given as an optional oblique argument.

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk, wiłkən''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T04:56:26Z

Nebulawindphone: /* Bodypart terms */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməgəmeh-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməgəmeh-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

* '''məp-u''' "eat (intr, foreground)"
* '''məc-i''' "eat (intr, background)"

* '''məc-u''' "deny (foreground)"
* '''məc-i''' "deny (background)"

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Cəmə''' ''n. I'' "Body"

'''Cəməjəmə''' ''n. IV'' "Everyone"

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cec-u''' ''v.i.'' "Defecate," redup. '''Cecec-u'''.

'''Cigic''' ''quant'' "Some"

'''Ciri, -rjən''' ''n. I'' "Cat," pl. '''Ciciri,''' pauc. '''Cirirə''' "Litter of cats."

'''Curireh-u''' ''v.t.'' "Drink," redup. '''Cucurireh-u'''.

'''Curicic-u''' ''v.i.'' "Quench one's thirst,"
* applic. '''Wədəcuriciceh-u''' "Enjoy, benefit from."

'''Ǝlum-u''' ''v.i.'' "Get up; set off on foot" perf. '''Ǝləlum-u'''.

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''ərah, -rarən''' ''n. I'' "Lover." (< PP '''*nriax''' "spouse")
* pauc. '''ərarə''' "Courting couple."

'''əramjə, -ramjan''' ''n. I'' "Bride."

'''ərahwə, -rahwan''' ''n. I'' "Groom."

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Halas, -lazən''' ''n. I'' "Chief," pl. '''Haralas'''.

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh, -nərən''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnur-u''' ''v.i.'' "Make camp," redup. '''Herẽnur-u'''.

'''Hẽnurwərəj, -rəjən''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go," redup. '''Irirug-u'''.

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kaməl''' ''n. I'' "Pan," pl. "Kəkaməl."

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kasə''' ''n. IV'' "Wind."

'''Katərə-w''' ''v.i.'' "Go hunting," redup. '''Kəgatərə-w'''.

'''Katərẽceh-u''' ''v.t.'' "Catch, hunt successfully," redup. '''Kəgatərẽceh-u'''.

'''Katəreh-u''' ''v.t.'' "Hunt for," patient trigger '''Katərec-u''', redup. '''Kəgatəreh-u'''.

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc, kəjən''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak, discuss," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Kəstin, -tinən''' ''n. I'' "Knife," pl. '''Kəgəstin'''.
* pauc. '''Kəstĩrə''' "Plough."

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lərut, -rudən''' ''n. I'' "Ocean," pl. '''Lələrut'''.

'''Ligwə, -gwan''' ''n. I'' "Parent," pauc. '''Ligwərə''' "Mother and son; father and daughter," pl. '''Liligwə'''.

'''Ligwəligwə''' ''n. IV'' "The Ancestors."

'''Ligwə-w''' ''stat.'' ''Who is (someone's) parent."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Məgəmeh-u''' ''v.t.'' "Break."
* redup. '''Məməgəmeh-u''' "Defeat."

'''Məgis-u''' ''stat.'' "Holding (something), carrying (something)."

'''Mənə''' ''n. I'' "Bay."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Medalə, -lan''' ''n. I'' "Opponent," pauc. '''Medalərə''' "One another's opponents," pl. '''Memedalə'''.

'''Milmeh-u''' ''v.t.'' "Tell (a story)," redup. '''Mimilmeh-u'''.

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Nigət, -gətən''' ''n. III'' "Berries."

'''Nizəj-u''' ''v.t.'' "Trades or barters with."

'''Nuheh-u''' ''v.t.'' "Pour out," redup. '''Nunuheh-u'''.

'''Nugim, -imən''' ''n. I'', pauc. '''Nugĩrə''', pl. '''Nunugim'''.

'''Pahu''' ''det.'' "What sort?"

'''Pəka''' ''det.'' "Which?"

'''Pəmir-u''' ''v.i.'' "Call out, shout," redup. '''Pəbəmir-u'''.

'''Pətu''' ''quant.'' "How many?"

'''Pusmeh-u''' ''v.t.'' "Swallow," redup. '''Pupusmeh-u'''.

'''Sirəm, -əmən''' ''n. I'', pl. '''Sizirəm'''.

'''Sism-u,''' ''v.i.'' "Sink," redup. '''Sizism-u'''.

'''Sumeh''' ''quant.'' "Every, all."

'''Sunəj, -nəjən''' ''n. I'' "River," pl. '''Suzunəj'''.

'''Swəłc-u''' ''stat.'' "Sick," redup. '''Swəzwəłc-u'''.

'''Swəłcwəłcu''' ''stat.'' "Crippled."

'''Təg-u''' ''stat.'' "Blue," redup. '''Tədəg-u'''.

'''Təjək, -jəgu''' ''n. I'' "Plant; herb, esp. medicinal."

'''Təkəbeh-u''' ''v.t.'' "Capture," redup. '''Tədəkəbeh-u'''.

'''Tərun, -runən''' ''n. II'' "Leaves of a tree."
* sg. '''Tərũreh''' "Leaf; page," pl. '''Tədərun''' "Leaves; pages."

'''Təməs, -məzən''' ''n. I'' "Village, pl. '''Tədəməs'''.

'''Tərãp, -rãbu''' ''n. I'' "Traveler," pl. '''Tədərãp'''.

'''Tunəw, -nəwən''' ''n. II'' "Horns." As a relational noun, "Peak, point." pl. '''Tudunəw'''.
* sg. '''Tunəweh''' "Horn; ferocity."

'''Tuswət, -wədən''' ''n. II'' "Someone's eyes," sg. '''Tuswəceh''' "An eye," pl. '''Tuduswət''' "Many eyes."

'''Ujəku''' ''stat.'' "Covering."

'''Ulwəs, -wəsən''' ''n. I'' "Worm," pl. '''Ululwəs'''.

'''Ũrəweh-u''' ''v.t.'' "Cook," redup. '''Umũrəweh-u'''.

'''Uspəneh-u''' ''v.t.'' "Weave," redup. '''Uzuspəneh-u'''.

'''Wəgudeh-u''' ''v.t.'' "Give (something)," takes recipient as an optional oblique argument.
* redup. '''Wəgũkudeh-u''' "Give (something) up."
* applic. '''Nĩkudeh-u''' "Give (to someone)," takes thing given as an optional oblique argument.

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk, wiłkən''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T04:56:03Z

Nebulawindphone: /* Headless NPs */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməgəmeh-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməgəmeh-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

* '''məp-u''' "eat (intr, foreground)"
* '''məc-i''' "eat (intr, background)"

* '''məc-u''' "deny (foreground)"
* '''məc-i''' "deny (background)"

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Cəmə''' ''n. I'' "Body"

'''Cəməjəmə''' ''n. IV'' "Everyone"

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cec-u''' ''v.i.'' "Defecate," redup. '''Cecec-u'''.

'''Cigic''' ''quant'' "Some"

'''Curireh-u''' ''v.t.'' "Drink," redup. '''Cucurireh-u'''.

'''Ǝlum-u''' ''v.i.'' "Get up; set off on foot" perf. '''Ǝləlum-u'''.

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''ərah, -rarən''' ''n. I'' "Lover." (< PP '''*nriax''' "spouse")
* pauc. '''ərarə''' "Courting couple."

'''əramjə, -ramjan''' ''n. I'' "Bride."

'''ərahwə, -rahwan''' ''n. I'' "Groom."

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh, -nərən''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnur-u''' ''v.i.'' "Make camp," redup. '''Herẽnur-u'''.

'''Hẽnurwərəj, -rəjən''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go," redup. '''Irirug-u'''.

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kaməl''' ''n. I'' "Pan," pl. "Kəkaməl."

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kasə''' ''n. IV'' "Wind."

'''Katərə-w''' ''v.i.'' "Go hunting," redup. '''Kəgatərə-w'''.

'''Katərẽceh-u''' ''v.t.'' "Catch, hunt successfully," redup. '''Kəgatərẽceh-u'''.

'''Katəreh-u''' ''v.t.'' "Hunt for," patient trigger '''Katərec-u''', redup. '''Kəgatəreh-u'''.

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc, kəjən''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak, discuss," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lərut, -rudən''' ''n. I'' "Ocean," pl. '''Lələrut'''.

'''Ligwə, -gwan''' ''n. I'' "Parent," pauc. '''Ligwərə''' "Mother and son; father and daughter," pl. '''Liligwə'''.

'''Ligwəligwə''' ''n. IV'' "The Ancestors."

'''Ligwə-w''' ''stat.'' ''Who is (someone's) parent."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Milmeh-u''' ''v.t.'' "Tell (a story)," redup. '''Mimilmeh-u'''.

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Nigət, -gətən''' ''n. III'' "Berries."

'''Nizəj-u''' ''v.t.'' "Trades or barters with."

'''Nuheh-u''' ''v.t.'' "Pour out," redup. '''Nunuheh-u'''.

'''Nugim, -imən''' ''n. I'', pauc. '''Nugĩrə''', pl. '''Nunugim'''.

'''Pəmir-u''' ''v.i.'' "Call out, shout," redup. '''Pəbəmir-u'''.

'''Pusmeh-u''' ''v.t.'' "Swallow," redup. '''Pupusmeh-u'''.

'''Sirəm, -əmən''' ''n. I'', pl. '''Sizirəm'''.

'''Sism-u,''' ''v.i.'' "Sink," redup. '''Sizism-u'''.

'''Sunəj, -nəjən''' ''n. I'' "River," pl. '''Suzunəj'''.

'''Swəłc-u''' ''stat.'' "Sick," redup. '''Swəzwəłc-u'''.

'''Swəłcwəłcu''' ''stat.'' "Crippled."

'''Təjək, -jəgu''' ''n. I'' "Plant; herb, esp. medicinal."

'''Təkəbeh-u''' ''v.t.'' "Capture," redup. '''Tədəkəbeh-u'''.

'''Tərun, -runən''' ''n. II'' "Leaves of a tree."
* sg. '''Tərũreh''' "Leaf; page," pl. '''Tədərun''' "Leaves; pages."

'''Təməs, -məzən''' ''n. I'' "Village, pl. '''Tədəməs'''.

'''Tərãp, -rãbu''' ''n. I'' "Traveler," pl. '''Tədərãp'''.

'''Tunəw, -nəwən''' ''n. II'' "Horns." As a relational noun, "Peak, point." pl. '''Tudunəw'''.
* sg. '''Tunəweh''' "Horn; ferocity."

'''Ujəku''' ''stat.'' "Covering."

'''Ulwəs, -wəsən''' ''n. I'' "Worm," pl. '''Ululwəs'''.

'''Ũrəweh-u''' ''v.t.'' "Cook," redup. '''Umũrəweh-u'''.

'''Uspəneh-u''' ''v.t.'' "Weave," redup. '''Uzuspəneh-u'''.

'''Wəgudeh-u''' ''v.t.'' "Give (something)," takes recipient as an optional oblique argument.
* redup. '''Wəgũkudeh-u''' "Give (something) up."
* applic. '''Nĩkudeh-u''' "Give (to someone)," takes thing given as an optional oblique argument.

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk, wiłkən''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T04:41:34Z

Nebulawindphone: /* Bodypart terms */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

* '''məp-u''' "eat (intr, foreground)"
* '''məc-i''' "eat (intr, background)"

* '''məc-u''' "deny (foreground)"
* '''məc-i''' "deny (background)"

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Cəmə''' ''n. I'' "Body"

'''Cəməjəmə''' ''n. IV'' "Everyone"

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cec-u''' ''v.i.'' "Defecate," redup. '''Cecec-u'''.

'''Cigic''' ''quant'' "Some"

'''Curireh-u''' ''v.t.'' "Drink," redup. '''Cucurireh-u'''.

'''Ǝlum-u''' ''v.i.'' "Get up; set off on foot" perf. '''Ǝləlum-u'''.

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''ərah, -rarən''' ''n. I'' "Lover." (< PP '''*nriax''' "spouse")
* pauc. '''ərarə''' "Courting couple."

'''əramjə, -ramjan''' ''n. I'' "Bride."

'''ərahwə, -rahwan''' ''n. I'' "Groom."

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh, -nərən''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnur-u''' ''v.i.'' "Make camp," redup. '''Herẽnur-u'''.

'''Hẽnurwərəj, -rəjən''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go," redup. '''Irirug-u'''.

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kaməl''' ''n. I'' "Pan," pl. "Kəkaməl."

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kasə''' ''n. IV'' "Wind."

'''Katərə-w''' ''v.i.'' "Go hunting," redup. '''Kəgatərə-w'''.

'''Katərẽceh-u''' ''v.t.'' "Catch, hunt successfully," redup. '''Kəgatərẽceh-u'''.

'''Katəreh-u''' ''v.t.'' "Hunt for," patient trigger '''Katərec-u''', redup. '''Kəgatəreh-u'''.

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc, kəjən''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak, discuss," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lərut, -rudən''' ''n. I'' "Ocean," pl. '''Lələrut'''.

'''Ligwə, -gwan''' ''n. I'' "Parent," pauc. '''Ligwərə''' "Mother and son; father and daughter," pl. '''Liligwə'''.

'''Ligwəligwə''' ''n. IV'' "The Ancestors."

'''Ligwə-w''' ''stat.'' ''Who is (someone's) parent."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Milmeh-u''' ''v.t.'' "Tell (a story)," redup. '''Mimilmeh-u'''.

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Nigət, -gətən''' ''n. III'' "Berries."

'''Nizəj-u''' ''v.t.'' "Trades or barters with."

'''Nuheh-u''' ''v.t.'' "Pour out," redup. '''Nunuheh-u'''.

'''Nugim, -imən''' ''n. I'', pauc. '''Nugĩrə''', pl. '''Nunugim'''.

'''Pəmir-u''' ''v.i.'' "Call out, shout," redup. '''Pəbəmir-u'''.

'''Pusmeh-u''' ''v.t.'' "Swallow," redup. '''Pupusmeh-u'''.

'''Sirəm, -əmən''' ''n. I'', pl. '''Sizirəm'''.

'''Sism-u,''' ''v.i.'' "Sink," redup. '''Sizism-u'''.

'''Sunəj, -nəjən''' ''n. I'' "River," pl. '''Suzunəj'''.

'''Swəłc-u''' ''stat.'' "Sick," redup. '''Swəzwəłc-u'''.

'''Swəłcwəłcu''' ''stat.'' "Crippled."

'''Təjək, -jəgu''' ''n. I'' "Plant; herb, esp. medicinal."

'''Təkəbeh-u''' ''v.t.'' "Capture," redup. '''Tədəkəbeh-u'''.

'''Tərun, -runən''' ''n. II'' "Leaves of a tree."
* sg. '''Tərũreh''' "Leaf; page," pl. '''Tədərun''' "Leaves; pages."

'''Təməs, -məzən''' ''n. I'' "Village, pl. '''Tədəməs'''.

'''Tərãp, -rãbu''' ''n. I'' "Traveler," pl. '''Tədərãp'''.

'''Tunəw, -nəwən''' ''n. II'' "Horns." As a relational noun, "Peak, point." pl. '''Tudunəw'''.
* sg. '''Tunəweh''' "Horn; ferocity."

'''Ujəku''' ''stat.'' "Covering."

'''Ulwəs, -wəsən''' ''n. I'' "Worm," pl. '''Ululwəs'''.

'''Ũrəweh-u''' ''v.t.'' "Cook," redup. '''Umũrəweh-u'''.

'''Uspəneh-u''' ''v.t.'' "Weave," redup. '''Uzuspəneh-u'''.

'''Wəgudeh-u''' ''v.t.'' "Give (something)," takes recipient as an optional oblique argument.
* redup. '''Wəgũkudeh-u''' "Give (something) up."
* applic. '''Nĩkudeh-u''' "Give (to someone)," takes thing given as an optional oblique argument.

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk, wiłkən''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T04:38:16Z

Nebulawindphone: /* Bodypart terms */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

* '''məp-u''' "eat (intr, foreground)"
* '''məc-i''' "eat (intr, background)"

* '''məc-u''' "deny (foreground)"
* '''məc-i''' "deny (background)"

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Cəmə''' ''n. I'' "Body"

'''Cəməjəmə''' ''n. IV'' "Everyone"

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cec-u''' ''v.i.'' "Defecate," redup. '''Cecec-u'''.

'''Cigic''' ''quant'' "Some"

'''Curireh-u''' ''v.t.'' "Drink," redup. '''Cucurireh-u'''.

'''Ǝlum-u''' ''v.i.'' "Get up; set off on foot" perf. '''Ǝləlum-u'''.

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''ərah, -rarən''' ''n. I'' "Lover." (< PP '''*nriax''' "spouse")
* pauc. '''ərarə''' "Courting couple."

'''əramjə, -ramjan''' ''n. I'' "Bride."

'''ərahwə, -rahwan''' ''n. I'' "Groom."

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh, -nərən''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnur-u''' ''v.i.'' "Make camp," redup. '''Herẽnur-u'''.

'''Hẽnurwərəj, -rəjən''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go," redup. '''Irirug-u'''.

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kaməl''' ''n. I'' "Pan," pl. "Kəkaməl."

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kasə''' ''n. IV'' "Wind."

'''Katərə-w''' ''v.i.'' "Go hunting," redup. '''Kəgatərə-w'''.

'''Katərẽceh-u''' ''v.t.'' "Catch, hunt successfully," redup. '''Kəgatərẽceh-u'''.

'''Katəreh-u''' ''v.t.'' "Hunt for," patient trigger '''Katərec-u''', redup. '''Kəgatəreh-u'''.

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc, kəjən''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lərut, -rudən''' ''n. I'' "Ocean," pl. '''Lələrut'''.

'''Ligwə, -gwan''' ''n. I'' "Parent," pauc. '''Ligwərə''' "Mother and son; father and daughter," pl. '''Liligwə'''.

'''Ligwəligwə''' ''n. IV'' "The Ancestors."

'''Ligwə-w''' ''stat.'' ''Who is (someone's) parent."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Milmeh-u''' ''v.t.'' "Tell (a story)," redup. '''Mimilmeh-u'''.

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Nigət, -gətən''' ''n. III'' "Berries."

'''Nizəj-u''' ''v.t.'' "Trades or barters with."

'''Nuheh-u''' ''v.t.'' "Pour out," redup. '''Nunuheh-u'''.

'''Nugim, -imən''' ''n. I'', pauc. '''Nugĩrə''', pl. '''Nunugim'''.

'''Pusmeh-u''' ''v.t.'' "Swallow," redup. '''Pupusmeh-u'''.

'''Sirəm, -əmən''' ''n. I'', pl. '''Sizirəm'''.

'''Sism-u,''' ''v.i.'' "Sink," redup. '''Sizism-u'''.

'''Sunəj, -nəjən''' ''n. I'' "River," pl. '''Suzunəj'''.

'''Təjək, -jəgu''' ''n. I'' "Plant; herb, esp. medicinal."

'''Tərun, -runən''' ''n. II'' "Leaves of a tree."
* sg. '''Tərũreh''' "Leaf; page," pl. '''Tədərun''' "Leaves; pages."

'''Təməs, -məzən''' ''n. I'' "Village, pl. '''Tədəməs'''.

'''Tərãp, -rãbu''' ''n. I'' "Traveler," pl. '''Tədərãp'''.

'''Tunəw, -nəwən''' ''n. II'' "Horns." As a relational noun, "Peak, point." pl. '''Tudunəw'''.
* sg. '''Tunəweh''' "Horn; ferocity."

'''Ujəku''' ''stat.'' "Covering."

'''Ulwəs, -wəsən''' ''n. I'' "Worm," pl. '''Ululwəs'''.

'''Ũrəweh-u''' ''v.t.'' "Cook," redup. '''Umũrəweh-u'''.

'''Uspəneh-u''' ''v.t.'' "Weave," redup. '''Uzuspəneh-u'''.

'''Wəgudeh-u''' ''v.t.'' "Give (something)," takes recipient as an optional oblique argument.
* redup. '''Wəgũkudeh-u''' "Give (something) up."
* applic. '''Nĩkudeh-u''' "Give (to someone)," takes thing given as an optional oblique argument.

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk, wiłkən''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T04:35:29Z

Nebulawindphone: /* Bodypart terms */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

* '''məp-u''' "eat (intr, foreground)"
* '''məc-i''' "eat (intr, background)"

* '''məc-u''' "deny (foreground)"
* '''məc-i''' "deny (background)"

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Cəmə''' ''n. I'' "Body"

'''Cəməjəmə''' ''n. IV'' "Everyone"

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cec-u''' ''v.i.'' "Defecate," redup. '''Cecec-u'''.

'''Cigic''' ''quant'' "Some"

'''Curireh-u''' ''v.t.'' "Drink," redup. '''Cucurireh-u'''.

'''Ǝlum-u''' ''v.i.'' "Get up; set off on foot" perf. '''Ǝləlum-u'''.

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''ərah, -rarən''' ''n. I'' "Lover." (< PP '''*nriax''' "spouse")
* pauc. '''ərarə''' "Courting couple."

'''əramjə, -ramjan''' ''n. I'' "Bride."

'''ərahwə, -rahwan''' ''n. I'' "Groom."

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh, -nərən''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnurwərəj, -rəjən''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go."

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kaməl''' ''n. I'' "Pan," pl. "Kəkaməl."

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kasə''' ''n. IV'' "Wind."

'''Katərə-w''' ''v.i.'' "Go hunting," redup. '''Kəgatərə-w'''.

'''Katərẽceh-u''' ''v.t.'' "Catch, hunt successfully," redup. '''Kəgatərẽceh-u'''.

'''Katəreh-u''' ''v.t.'' "Hunt for," patient trigger '''Katərec-u''', redup. '''Kəgatəreh-u'''.

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc, kəjən''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lərut, -rudən''' ''n. I'' "Ocean," pl. '''Lələrut'''.

'''Ligwə, -gwan''' ''n. I'' "Parent," pauc. '''Ligwərə''' "Mother and son; father and daughter," pl. '''Liligwə'''.

'''Ligwəligwə''' ''n. IV'' "The Ancestors."

'''Ligwə-w''' ''stat.'' ''Who is (someone's) parent."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Milmeh-u''' ''v.t.'' "Tell (a story)," redup. '''Mimilmeh-u'''.

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Nigət, -gətən''' ''n. III'' "Berries."

'''Nizəj-u''' ''v.t.'' "Trades or barters with."

'''Nuheh-u''' ''v.t.'' "Pour out," redup. '''Nunuheh-u'''.

'''Nugim, -imən''' ''n. I'', pauc. '''Nugĩrə''', pl. '''Nunugim'''.

'''Pusmeh-u''' ''v.t.'' "Swallow," redup. '''Pupusmeh-u'''.

'''Sirəm, -əmən''' ''n. I'', pl. '''Sizirəm'''.

'''Sunəj, -nəjən''' ''n. I'' "River," pl. '''Suzunəj'''.

'''Təjək, -jəgu''' ''n. I'' "Plant; herb, esp. medicinal."

'''Tərun, -runən''' ''n. II'' "Leaves of a tree."
* sg. '''Tərũreh''' "Leaf; page," pl. '''Tədərun''' "Leaves; pages."

'''Təməs, -məzən''' ''n. I'' "Village, pl. '''Tədəməs'''.

'''Tərãp, -rãbu''' ''n. I'' "Traveler," pl. '''Tədərãp'''.

'''Tunəw, -nəwən''' ''n. II'' "Horns." As a relational noun, "Peak, point." pl. '''Tudunəw'''.
* sg. '''Tunəweh''' "Horn; ferocity."

'''Ujəku''' ''stat.'' "Covering."

'''Ulwəs, -wəsən''' ''n. I'' "Worm," pl. '''Ululwəs'''.

'''Ũrəweh-u''' ''v.t.'' "Cook," redup. '''Umũrəweh-u'''.

'''Wəgudeh-u''' ''v.t.'' "Give (something)," takes recipient as an optional oblique argument.
* redup. '''Wəgũkudeh-u''' "Give (something) up."
* applic. '''Nĩkudeh-u''' "Give (to someone)," takes thing given as an optional oblique argument.

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk, wiłkən''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T03:41:03Z

Nebulawindphone: /* Bodypart terms */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

* '''məp-u''' "eat (intr, foreground)"
* '''məc-i''' "eat (intr, background)"

* '''məc-u''' "deny (foreground)"
* '''məc-i''' "deny (background)"

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cigic''' ''quant'' "Some"

'''Ǝlum-u''' ''v.i.'' "Walk," perf. '''Ǝləlum-u''',
* appl. '''Wədalumeh-u''' "Walk with (someone)," '''Wədalumuz-u''' "Walk together, court with."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnurwərəj''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go."

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kasə''' ''n. IV'' "Wind."

'''Katərə-w''' ''v.i.'' "Go hunting," redup. '''Kəgatərə-w'''.

'''Katərẽceh-u''' ''v.t.'' "Catch, hunt successfully," redup. '''Kəgatərẽceh-u'''.

'''Katəreh-u''' ''v.t.'' "Hunt for," patient trigger '''Katərec-u''', redup. '''Kəgatəreh-u'''.

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T03:36:33Z

Nebulawindphone: /* Bodypart terms */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

* '''məp-u''' "eat (intr, foreground)"
* '''məc-i''' "eat (intr, background)"

* '''məc-u''' "deny (foreground)"
* '''məc-i''' "deny (background)"

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cigic''' ''quant'' "Some"

'''Ǝlum-u''' ''v.i.'' "Walk," perf. '''Ǝləlum-u''',
* appl. '''Wədalumeh-u''' "Walk with (someone)," '''Wədalumuz-u''' "Walk together, court with."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* redup. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnurwərəj''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," redup. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," redup. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go."

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Irab-u''' ''v.i.'' "Hiccup; flinch; tremble."

'''Jən''' ''n. I'' "Chief," du. '''Jərə,''' pl. '''Jərjən'''.

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Karəjsk-u''' ''stat.'' "Big."

'''Karəri-w''' ''v.i.'' "Grow bigger."

'''Kəbah''' ''n. II'' "Someone's skin; something's hide," pl. '''Kəgəbah'''.
* pauc. '''Kəbahə''' "Animal hide; coat or blanket."

'''Kəc''' ''n. II'' "(One another's) siblings," sg. '''Kəceh''' "Someone's sibling," pl. '''Kəjgəc''' "Various people's siblings."

'''Kəc-u''' ''stat.'' "Who is (someone's) sibling," redup. '''Kəjgəcər-u''' "Who are one another's siblings."

'''Kəcwə''' ''n. I'' "Gate," pl. '''Kəgəcwə'''.

'''Kədər-u''' ''v.i.'' "Talk, speak," redup. '''Kəgədər-u'''
* applic. '''Nigədəreh-u''' "Address," redup. '''Kənigədəreh-u'''

'''Kədəreh-u''' ''v.t.'' "Say," redup. '''Kəgədəreh-u'''

'''Kədərwərjər-u''' ''v.i.'' "Have spoken; hold one's peace."

'''Kəhek''' ''n. I'' "Wife," pl. '''Kəgəhek'''.
* pauc. '''Kəhekə''' "Wife and mother-in-law."

'''Kəhi''' ''n. IV'' "Hair."

'''Kəkər-u''' ''v.i.'' "Vomit," redup. '''Kəgəkər-u'''.

'''Kəłcu''' ''n. I'' "Parent's sibling," pl. '''Kəgəłcu'''.
* pauc. '''Kəłcurə''' "Uncle and niece; aunt and nephew."

'''Kəłcurwə''' ''n. I'' "Uncle (parent's brother)," pl. '''Kəgəłcurwə'''.
* pauc. '''Kəłcurwərə''' "Uncle and nephew."

'''Kəłcumjə''' ''n. I'' "Aunt (parent's sister)," pl. '''Kəgəłcumjə'''.
* pauc. '''Kəłcumjərə''' "Aunt and niece."

'''Kəmeh''' ''n. I'' "Wall," pl. '''Kəgəmeh'''.

'''Kəmehə''' ''n. II'' "Walled town," pl. '''Kəgəmehə''', no sg. form attested.

'''Kərjul''' ''n. IV'' "Tree bark."

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T03:15:23Z

Nebulawindphone: /* Grounding */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize, which caused a few mergers.

* '''məp-u''' "eat (intr, foreground)"
* '''məc-i''' "eat (intr, background)"

* '''məc-u''' "deny (foreground)"
* '''məc-i''' "deny (background)"

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cigic''' ''quant'' "Some"

'''Ǝlum-u''' ''v.i.'' "Walk," perf. '''Ǝləlum-u''',
* appl. '''Wədalumeh-u''' "Walk with (someone)," '''Wədalumuz-u''' "Walk together, court with."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* perf. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnurwərəj''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," perf. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," perf. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go."

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Jən''' ''n. I'' "Chief," du. '''Jəreh,''' pl. '''Jərjən'''.

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T03:13:38Z

Nebulawindphone: /* Grounding */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

The citation forms are the foreground forms in '''-u'''. The background clitic caused preceding '''d''' or '''t''' to palatalize.

* '''məp-u''' "eat (intr, foreground)"
* '''məc-i''' "eat (intr, background)"

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cigic''' ''quant'' "Some"

'''Ǝlum-u''' ''v.i.'' "Walk," perf. '''Ǝləlum-u''',
* appl. '''Wədalumeh-u''' "Walk with (someone)," '''Wədalumuz-u''' "Walk together, court with."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* perf. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnurwərəj''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," perf. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," perf. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go."

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Jən''' ''n. I'' "Chief," du. '''Jəreh,''' pl. '''Jərjən'''.

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T03:11:51Z

Nebulawindphone: /* Grounding */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

Like in its sister language Kibülʌiṅ, Merneha marked statives and verbs in the indicative mood for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel); these were clear reflexes of the Proto-Peninsular ''u''-phase and ''i''-phase morphemes.

Verbs outside the indicative mood had no explicit grounding clitic. However, the markers for non-indicative mood all incorporated a reflex of the Proto-Peninsular ''a''-phase morpheme, cognate with the Kibülʌiṅ non-indicative background marker '''ʌ'''. Unlike in Kibülʌiṅ, these same morphemes were used when the non-indicative verb was foregrounded.

As a rule, relative clause verbs or statives that appeared as modifiers in the sentence topic had to be marked as background, since they were presumably old information. Elsewhere in the sentence, they were marked as background if they were restrictive modifiers, and as foreground if they were nonrestrictive.

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cigic''' ''quant'' "Some"

'''Ǝlum-u''' ''v.i.'' "Walk," perf. '''Ǝləlum-u''',
* appl. '''Wədalumeh-u''' "Walk with (someone)," '''Wədalumuz-u''' "Walk together, court with."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* perf. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnurwərəj''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," perf. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," perf. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go."

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Jən''' ''n. I'' "Chief," du. '''Jəreh,''' pl. '''Jərjən'''.

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T02:59:39Z

Nebulawindphone:

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

All predicates, whether they were statives or true verbs, were marked for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel).

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cigic''' ''quant'' "Some"

'''Ǝlum-u''' ''v.i.'' "Walk," perf. '''Ǝləlum-u''',
* appl. '''Wədalumeh-u''' "Walk with (someone)," '''Wədalumuz-u''' "Walk together, court with."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* perf. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnurwərəj''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," perf. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," perf. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go."

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Jən''' ''n. I'' "Chief," du. '''Jəreh,''' pl. '''Jərjən'''.

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T02:58:55Z

Nebulawindphone: /* Clause particles and */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Adverbs and particles =

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

All predicates, whether they were statives or true verbs, were marked for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel).

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Other ==

Four postverbal particles are attested: the interrogative particle '''ka''', the mirative particle '''la''', and the coordinating conjunctions '''cə''' and '''je'''. There were no subordinating conjunctions; the subordinate clause was relativized instead and made an oblique argument of the main clause.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

=== Coordination ===

There were only two true coordinating conjunctions: '''cə''' (from Proto-Peninsular '''*kta''') corresponding to English "and" or "but," and '''je''' (of unknown origin) corresponding to "or." Both followed the first and optionally the second of the coordinated clauses: CLAUSE-'''cə''' CLAUSE or CLAUSE-'''cə''' CLAUSE-'''cə'''.

Linking a negative clause to a positive one, '''cə''' could mean something like "except."

=== Subordination ===

The main mechanism of subordination was to form a relative clause and attach one of the incorporating statives to the head.

===

== Questions ==

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cigic''' ''quant'' "Some"

'''Ǝlum-u''' ''v.i.'' "Walk," perf. '''Ǝləlum-u''',
* appl. '''Wədalumeh-u''' "Walk with (someone)," '''Wədalumuz-u''' "Walk together, court with."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* perf. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnurwərəj''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," perf. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," perf. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go."

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Jən''' ''n. I'' "Chief," du. '''Jəreh,''' pl. '''Jərjən'''.

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T02:55:18Z

Nebulawindphone: /* Incorporating statives */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Adverbs and particles =

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

All predicates, whether they were statives or true verbs, were marked for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel).

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During/under the circumstances || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Clause particles and ==

A few particles could follow the main verb of a clause. These had several functions — forming questions, coordinating clauses, and expressing uncertainty, certainty or surprise.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

===

== Questions ==

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cigic''' ''quant'' "Some"

'''Ǝlum-u''' ''v.i.'' "Walk," perf. '''Ǝləlum-u''',
* appl. '''Wədalumeh-u''' "Walk with (someone)," '''Wədalumuz-u''' "Walk together, court with."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* perf. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnurwərəj''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," perf. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," perf. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go."

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Jən''' ''n. I'' "Chief," du. '''Jəreh,''' pl. '''Jərjən'''.

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T02:32:01Z

Nebulawindphone: /* Incorporating statives */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Adverbs and particles =

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

All predicates, whether they were statives or true verbs, were marked for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel).

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Ipəka''' and '''ipahu''' were commonly combined with incorporating statives to compound WH-words like the following:

* '''ipahu-tər-u''' "how?" ("just like what sort of thing?")
* '''ipəka-nir-u''' "when?" ("during which thing?")
* '''ipahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''ipəka-t-u''' "where?" ("at which thing?")
* '''ipəka-tə-w''' "by what means?" ("with what thing?")
* '''ipahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Clause particles and ==

A few particles could follow the main verb of a clause. These had several functions — forming questions, coordinating clauses, and expressing uncertainty, certainty or surprise.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

===

== Questions ==

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cigic''' ''quant'' "Some"

'''Ǝlum-u''' ''v.i.'' "Walk," perf. '''Ǝləlum-u''',
* appl. '''Wədalumeh-u''' "Walk with (someone)," '''Wədalumuz-u''' "Walk together, court with."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* perf. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnurwərəj''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," perf. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," perf. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go."

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Jən''' ''n. I'' "Chief," du. '''Jəreh,''' pl. '''Jərjən'''.

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T02:31:14Z

Nebulawindphone: /* Incorporating statives */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Adverbs and particles =

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

All predicates, whether they were statives or true verbs, were marked for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel).

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

'''Pəka''' and '''pahu''' were commonly combined with incorporating statives to form the following compound WH-words:

* '''pahu-tər-u''' "how?" ("just like what sort of thing?")
* '''pəka-nir-u''' "when?" ("during which thing?")
* '''pahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''pəka-t-u''' "where?" ("at which thing?")
* '''pəka-tə-w''' "by what means?" ("with what thing?")
* '''pahu-tə-w''' "in what manner/to what extent?" ("(compared) with what sort of thing?")

And the demonstratives '''igu''', '''iju''' and '''ire''' could be used in similar combinations.

* '''igu-tər-u''' "like this"
* '''iju-tər-u''' "like that"
* '''ire-tər-u''' "like that over there"
* '''iju-nir-u''' "then, during that"
* '''iju-t-u''' "there, at that"
* '''iju-tə-w''' "to that extent"

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Clause particles and ==

A few particles could follow the main verb of a clause. These had several functions — forming questions, coordinating clauses, and expressing uncertainty, certainty or surprise.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

===

== Questions ==

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cigic''' ''quant'' "Some"

'''Ǝlum-u''' ''v.i.'' "Walk," perf. '''Ǝləlum-u''',
* appl. '''Wədalumeh-u''' "Walk with (someone)," '''Wədalumuz-u''' "Walk together, court with."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* perf. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnurwərəj''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," perf. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," perf. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go."

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Jən''' ''n. I'' "Chief," du. '''Jəreh,''' pl. '''Jərjən'''.

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T02:26:41Z

Nebulawindphone: /* Clause particles */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Adverbs and particles =

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

All predicates, whether they were statives or true verbs, were marked for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel).

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Clause particles and ==

A few particles could follow the main verb of a clause. These had several functions — forming questions, coordinating clauses, and expressing uncertainty, certainty or surprise.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

===

== Questions ==

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cigic''' ''quant'' "Some"

'''Ǝlum-u''' ''v.i.'' "Walk," perf. '''Ǝləlum-u''',
* appl. '''Wədalumeh-u''' "Walk with (someone)," '''Wədalumuz-u''' "Walk together, court with."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* perf. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnurwərəj''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," perf. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," perf. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go."

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Jən''' ''n. I'' "Chief," du. '''Jəreh,''' pl. '''Jərjən'''.

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T02:26:07Z

Nebulawindphone: /* Questions */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Adverbs and particles =

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

All predicates, whether they were statives or true verbs, were marked for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel).

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Clause particles ==

A few particles could follow the main verb of a clause. These had several functions — forming questions, coordinating clauses, and expressing uncertainty, certainty or surprise.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

'''Pəka''' and '''pahu''' were commonly combined with incorporating statives to form the following compound WH-words:

* '''pahu-tər-u''' "how?" ("just like what sort of thing?")
* '''pəka-nir-u''' "when?" ("during which thing?")
* '''pahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''pəka-t-u''' "where?" ("at which thing?")
* '''pəka-tə-w''' "by what means?" ("with what thing?")
* '''pahu-tə-w''' "in what manner/to what extent?" ("with what sort of thing?")

== Questions ==

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cigic''' ''quant'' "Some"

'''Ǝlum-u''' ''v.i.'' "Walk," perf. '''Ǝləlum-u''',
* appl. '''Wədalumeh-u''' "Walk with (someone)," '''Wədalumuz-u''' "Walk together, court with."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* perf. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnurwərəj''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," perf. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," perf. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go."

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Jən''' ''n. I'' "Chief," du. '''Jəreh,''' pl. '''Jərjən'''.

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T02:24:05Z

Nebulawindphone: /* Coordination and subordination */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Adverbs and particles =

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

All predicates, whether they were statives or true verbs, were marked for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel).

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Clause particles ==

A few particles could follow the main verb of a clause. These had several functions — forming questions, coordinating clauses, and expressing uncertainty, certainty or surprise.

=== Questions ===

Questions were formed using the interrogative particle '''ka'''. If no other interrogative words were present in the clause, it was interpreted as a polar question.

If the question expected a negative answer, a verb in the negative mood could be used.

If it expected a positive answer, the potential mood was sometimes used. Probably this was perceived as softening the question in some way.

Content questions could be formed using the interrogative determiners '''pəka''' "which?" and '''pahu''' "what sort of?" and the interrogative quantifier '''pətu''' "how many?"

'''Pəka''' and '''pahu''' were commonly combined with the incorporating statives.

* '''Pahu-tər-u''' "how?" ("just like what sort of thing?")
* '''Pəka-nir-u''' "when?" ("during which thing?")
* '''Pahu-nir-u''' "in what circumstances?" ("during what sort of thing?")
* '''Pəka-t-u''' "where?" ("at which thing?")
* '''Pəka-tə-w''' "by what means?" ("with what thing?")
* '''Pahu-tə-w''' "in what manner?" ("with what sort of thing?")

== Questions ==

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cigic''' ''quant'' "Some"

'''Ǝlum-u''' ''v.i.'' "Walk," perf. '''Ǝləlum-u''',
* appl. '''Wədalumeh-u''' "Walk with (someone)," '''Wədalumuz-u''' "Walk together, court with."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* perf. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnurwərəj''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," perf. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," perf. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go."

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Jən''' ''n. I'' "Chief," du. '''Jəreh,''' pl. '''Jərjən'''.

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."

Merneha

2009-12-01T02:10:17Z

Nebulawindphone: /* Relative clauses */

= Phonology =

The three-vowel system of proto-Peninsular opened out into a slightly unusual five-vowel system in Mernesha: ''a ə e i u''. This can be thought of as a rectangular system (likely pronounced /ɑ e i u/) with a central vowel added. If vowel length was present in proto-Peninsular — some reconstructions include it, some don't — it has been lost here. The peripheral vowels could be oral or nasalized. There was no nasalization contrast on the central vowel.

The general shape of the consonant system is well known, although some of the pronunciations are disputed. We know ot gained a voicing distinction and the semivowels ''y w'', and lost the contrast between proto-Peninsular ''*h'' and ''*x'', which have merged in most contexts as ''h''. The precise points of articulation for some consonants are impossible to reconstruct, so for instance we cannot say for sure whether ''h r'' were velar or uvular. Similarly, it is unclear whether ''l'' was a lateral approximant or a voiced lateral fricative.

== Phonemic inventory ==

=== Consonants ===

{| {{bluetable|lightbluebg c}}
|-
! !! labial || apical || laminal || dorsal
|-
! stop
| p b || t d || c || k g
|-
! fricative
| || s z || || h r
|-
! lateral
| || ł l ||
|-
! sonorant
| m || n || j || w
|-
|}

=== Vowels ===

{| {{bluetable|lightbluebg c}}
|-
! !! front || central || back
|-
! high
| i ĩ || || u ũ
|-
! mid
| || ə ||
|-
! low
| e ẽ || || a ã
|-
|}

== Word-level phonology ==

Clusters were quite rare, and most involved the glides ''j'' and ''w.'' The maximal syllable was CłgVgC (where "g" stands for "glide"). The attested consonant clusters ''not'' involving glides were:

* Stop plus nasal — ''sũnipma'' "apple (dir)," ''picmirugu'' "become a man"
* ''S'' plus non-fricative — ''təməsma'' "village (dir)," ''təskis'' "root"
* ''L'' plus ''m, n, r'' — ''milmu'' "singing," ''ləlnər'' "branches"
* ''Ł'' before or after ''p, t, c, k, s'' — ''płəzə'' "many," ''swəłku'' "sick"
* ''Skł'' — ''wəriskła'' "could have gone," ''məgiziskła'' "could have carried"

Only the stop-plus-''ł'' clusters could occur word initially.

Any of these clusters could be preceded or followed by a glide, although this was ''quite'' rare. A few attested examples are:

* ''picmwə'' "man"
* ''swəłcwəłku'' "quite sick"
* ''pələjsku'' "this forest"

Four- or five-consonant clusters, while theoretically possible given the syllable structure, are not attested.

Stress fell on the penult, or on the antepenult if the last two vowels were ''ə''.

Allophony is difficult to reconstruct. There is evidence from daughter languages of final consonant devoicing, and of the syncopation of some unstressed vowels.

== Words, phrases and clitics ==

A phonological word in Merneha often consisted of more than one one syntactic word. A number of important words in Merneha were postclitics. Some of these were grammatical particles — determiners, quantifiers, verb tense and aspect markers, and so on. But some were lexical words in their own right. The copula was a clitic, as were a few other verbs ("act like," "get," "use," "come/go") and a number of statives. (I've referred to these later in this sketch as "incorporating verbs" and "incorporating statives." That's not quite right, since what they do isn't really incorporation at all, but calling them "clitic verbs" invites confusion with the "verb clitics" — postverbal clitic particles like the past tense marker ''-isk'' — and I decided not to open that can of worms.)

The result was that phrase structure and division into phonological words often did not match. The two words ''werĩhəj sumehəlu'' "From every village" correspond to the noun phrase ''werĩhəj sume'' "every village" and the stative ''-həl-u'' "from" (itself built from ''-həl'' and the phase clitic ''-u''). To make the phrase structure visible, I will write a dash between clitics in most of the examples here, and will use square brackets to reveal phrase structure when it might be unclear.

* [''werĩhəj sume'']-''rəl-u''

This could be taken further; it was not uncommon for three or four levels of syntactic structure to contribute morphemes to a single phonological word.

* [[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'' "Come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma'' "(Ones) that come from every village"
* [[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'' "Some that come from every village"
* [[[[''Werĩhəj sume'']-''rəl-ũ'']-''ri-w'']''-ma cigic'']-''ar-isk-u'' "There were some from every village."

The last example above would have been pronounced as three words: ''Werĩhəj sumehəlũriwma cigicarisku.''

From a phonological point of view, what is important is that all of the morphophonological processes listed below applied between clitics as well as within the syntactic word. So for instance, ''-həl-u'' lenites to ''-rəl-u'' after a vowel, despite the intervening syntactic word boundary.

== Morphophonology ==

=== Lenition ===

Voiceless consonants often became voiced intervocalically. This did not affect clusters in PPI, some of which simplified; the result is that certain morphemes specifically resit lenition. We will speak of "strong" suffixes or postclitics (with initial consonants that resist lenition) and "weak" ones (with initial consonants that lenite).

There are also some words whose final consonants resist lenition before a vowel. In particular, some nouns lenite and some don't, and this is indicated in the lexicon by giving both the bare noun and the noun plus genitive clitic.

* ''kəc, kəjən'' "sibling"
* ''hənəh, hənərən'' "blossom"
* ''lərut, lərudən'' "ocean"
* ''ulwəs, ulwəsən'' "worm"
* ''wiłk, wiłkən'' "crop"
* ''nigət, nigətən'' "berry"

Consonants subject to lenition were also affected between an vowel and a semivowel, or between two semivowels.

* ''hẽnurwərəj'' "bed" + ''ku'' "this" > ''hẽnurwərəjgu'' "this bed"

=== "Palatalization" ===

The palatal affricate ''c'' and the palatal glide ''j'' could arise from other consonants in a few different ways. Only some would count as processes of palatalization in the strict sense, but it is useful to treat them all together.

The simplest was the palatalization of ''t d'' to ''c j'' before or after ''i'' or ''j''.

* ''wəgudeh-u'' "give (agt focus)," ''wəgujic-u'' "give (pt focus)"
* ''sunəj'' "river" + ''-tər-u'' > ''sunə-jər-u'' "as a river, like a river"

A number of PPI clusters ending in an alveolar or post-alveolar also reduced to ''c''. The result is that certain "palatalizing suffixes" combined with a stem-final voiceless stops to form ''c''.

* ''tunəw'' "horns" > ''tunəw-t-u'' "horns (loc)" (''*tunaufthu'')
* ''təjək'' "plant" > ''təjə-c-u'' "plant (loc)" (''*taliakthu'')
* ''ərah'' "spouse" > ''əra-tər-u'' "as a spouse" (''*nriaxtari'')
* ''tərãp'' "traveler" > ''tərã-cər-u'' "as a traveler" (''*taxnptari'')

Other clusters only palatalized before or after ''i''. The result is that, for instance, the imperfect stative suffix ''-(r)i'' was palatalizing while the aorist stative suffix ''-(r)u'' was not.

* ''ujəku'' "covering (aor)", ''ujəci'' "covering (impf)" (''*uliakru, *uliakri'')

Finally, the combinations ''lj tlj'' palatalized, while ''li tli'' did not.

* ''nizəli-'' "trade with" > ''nizəju'' (<''nizəlj-u'') "trades with," ''nizəlisku'' "traded with"

=== Nasalization ===

Vowel-nasal-consonant sequences tended to lose the nasal consonant and nasalize the vowel. The nasalization contrast did not occur on ''ə''; between a schwa and another consonant, nasal consonants simply vanished. (This suggests ''*ənt'' > ''*ə˜t'' > ''ət''.)

* ''nugim'' "mother-in-law", ''nugĩrə'' "mothers-in-law (dual)"
* ''sirəm'' "lice", ''sirərə'' "louse"

=== Vowel reduction ===

When two vowels are brought together across a morpheme boundary:

* If the first vowel is ''i'' or ''u'', it reduces to a glide: ''j'' or ''w'' respectively.
* If the first vowel is ''a'' or ''ə'' and the second is ''i'' or ''u,'' the second reduces to a glide.
* The combinations ''aa, aə'' and ''əa'' contract to ''a''.

''E'' in some morphemes behaved like ''i'', and in others like ''ə''.

=== Cluster reduction ===

A stop-initial suffix that was not palatalizing simply deleted a stem-final stop.

* ''tərãp'' "traveler" > ''tərã-ku'' "that traveler"
* ''nigət'' "berries" > ''nigə-ku'' "those berries"

=== H and R ===

At a fairly late stage in the evolution of of Merneha, ''h'' vanished in all clusters, and ''r'' vanished in clusters with voiceless consonants. The combinations ''hl'' and ''lh'' became ''ł''; ''hr'', ''rh'' and ''hh'' became ''h''.

This occurred after lenition (so that voiceless consonants in VCrV and VChV combinations remained voiceless) and after nasalization (so that the outcome of VNhV was ṼhV, and the outcome of VNrV was ṼrV).

* ''təməs'' "village" + ''-rar-i'' copula --> ''təməs-ar-i'' "it's a village"
* ''tədərun'' "leaves" --> ''tədərũ-rar-i'' "they're all leaves"
* ''kaməl'' "pan" --> ''kaməl-rar-i'' "it's a pan"
* ''təməs'' + ''-he'' "that there" --> ''təməs-e'' "that village there"
* ''tədərun'' --> ''tədərũ-he'' "those leaves there"
* ''kaməl'' --> ''kaməł-e'' "that pan there"

=== Reduplication ===

Full reduplication of a word occurred in some derivational processes.

* ''ligwə'' "parent" > ''ligwəligwə'' "ancestors"
* ''cəmə'' "body" > ''cəməjəmə'' "everyone"

Partial reduplication is much more pervasive, occurring as an inflectional process for nouns, verbs and statives alike. (Here it is demonstrated with verbs.) The usual form was reduplication of the initial syllable's onset and vowel

{| {{bluetable|lightbluebg c}}
|-
! !! Swallow || Drink || Tell (a story) || Pour || Defecate
|-
! plain
| ''pusmu'' || ''curiru'' || ''milmu'' || ''nuhu'' || ''cecu''
|-
! augment
| ''pu-pusmu'' || ''cu-curiru'' || ''mi-milmu'' || ''nu-nuhu'' || ''ce-cecu''
|-
|}

or of the vowel and first subsequent consonant of a vowel-initial word. Words that begin with a nasal vowel have a ''Vm-'' or ''Vn-'' reduplicant.

{| {{bluetable|lightbluebg c}}
|-
! !! Fall || Get up || Cook || Go || Weave
|-
! plain
| ''əjubu'' || ''əlumu'' || ''ũrəw'' || ''irugu'' || ''uspənu''
|-
! augment
| ''ər-əjubu'' || ''əl-əlumu'' || ''um-ũrəw'' || ''ir-irugu'' || ''uz-uspənu''
|-
|}

The various other morphophonological processes listed above also applied during reduplication. So for instance, while some "strong" stems beginning with a consonant resist lenition, other "weak" stems lenite in their reduplicated form.

{| {{bluetable|lightbluebg c}}
|-
! !! Make camp || Sink || Discuss || Capture || Shout, cry out
|-
! plain
| ''hẽnuru'' || ''sismu'' || ''kədəru'' || ''təkəbu'' || ''pəmiru''
|-
! augment
| ''he-rẽnuru'' || ''si-zismu'' || ''kə-gədəru'' || ''tə-dəkəbu'' || ''pə-bəmiru''
|-
|}

With full reduplication, it is possible to find instances of palatalization as well.

* ''swəłku'' "sick" > ''swəłcwəlku'' "crippled"

= Morphosyntactic Overview =

== Parts of speech ==

There were three open word classes in Mernesha: nouns, verbs and ''statives''.

The distinction between verbs and statives was largely aspectual, although the name "stative" is something of a misnomer — the class included activities and even some iterative and semelfactive words as well as states. Verbs were characteristically ''telic'', statives characteristicaly ''atelic''.

Only verbs could serve as the head of a finite clause. Only verbs could undergo voice or valence changing operations, be marked for tense, or be used in the negative, subjunctive, potential, imperative, or prohibitive voices. For statives, various periphrastic constructions involving the copula had to be used instead. Statives also lacked infinitive or gerund forms. Statives could take a noun phrase complement just like verbs could, although these complements were not marked for case. Indeed, many statives had a relational meaning, playing the role prepositions or postpositions might in another language.

Closed classes include pronouns, nominal particles (determiners, case markers and quantifiers, some of which were independent words and some of which were clitics), conjunctions, a few true adverbs, and interjections.

== Phase ==

Every predicate — that is, every verb and every stative — was marked for a category known to Peninsularists as ''phase''. In Gaadràmarneš — and as far as can be reconstructed, in Proto-Peninsular — a predicate's phase was determined by its aspect and mood. In Merneha phase was largely aspectual, but showed some signs of influence from the foreground/background-marking system found in Kibülʌiṅ. As a rule, all the verbs and statives in a clause had to agree in phase, although there were some signs that this pattern was breaking down.

The ''u''-phase, marked with the postclitic ''-u'', ''-w'' or ''-a'', was used in clauses describing

The ''i''-phase, marked with the postclitic ''-i'', ''-j'' or ''-e'', is used in

The ''a''-phase, marked with the postclitic ''-a'', ''-e'' or ''-ə'', was used in irrealis clauses — imperatives, prohibitives, subjunctive clauses, and so on — and on a few idiosyncratic statives regardless of mood, such as ''t-ə'' "compared to."

== Constituent order ==

Mernesha was relentlessly head-final. Modifiers of all types preceded the nouns the modified; nominal arguments, complement clauses and adverbial modifiers preceded verbs; the finite verb was the last element of every clause. Order in transitive clauses was SOV — but given the strong tendency to make the most topical argument in every clause the subject, this can also be seen as topic-comment.

== Alignment ==

Mernesha combined PPI's rich system of voice operations with a tendency to put topics in subject position at all costs to arrive at something like a terrestrial Austronesian or direct/inverse alignment system. Transitive verbs could be found in "agent trigger" or "patient trigger" forms, with the trigger in either voice in direct case, the patient of agent trigger verbs in genitive case, and the agent of patient trigger verbs in ergative case.

* ''In-ma sũnip-ən məpeh-isk-u.'' Name DIR apple-GEN eat.AGT-PAST-PHASE "In ate an apple."

* ''Sũnip-ma In-lu məcic-isk-u.'' apple-DIR name-ERG eat.PAT-PAST-PHASE "It was In who ate the apple."

Instruments and benefactive arguments could also be triggers, although for these two changes had to be made to the verb — first an applicative prefix was added to promote the instrument or benefactive argument to the "patient" role, then the patient trigger form of the verb was used.

* ''Kəstĩ-ku-ma In-lu sũnip-ən məgis-u wədəməcic-isk-u.'' knife-this-DIR name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE "With this knife, In ate an apple."

The trigger was also the only element of the verb phrase that could be relativized.

* ''In-lu sũnip-ən məgis-u wədəməcic-isk-u kəstin'' Name-ERG apple-GEN hold-PHASE INSTR.eat.PAT-PAST-PHASE knife "The knife that In ate an apple with."

= Nominal Morphosyntax =

A maximal noun phrase consisted of an inflected noun preceded by modifiers and followed by a series of noun particles — determiners, case particles, quantifiers and conjunctions.

In fact, though, every element was optional. In particular, the noun itself was optional: a ''headless'' noun phrase could consist of a modifier followed by a determiner and possibly other particles. On the other hand, if a noun was present, all the noun particles were optional.

== The noun ==

The only inflectional category marked on the noun itself was number.

=== Number ===

Merneha distinguished three numbers: singular, paucal and plural. The paucal, derived from the Proto-Peninsular dual number, was used primarily for naturally occurring sets — often pairs (as in the two hands or feet of a person) but not always. It was semantically extended to some "natural sets of one" — for instance, naturally singular bodyparts such as the head, nose or backside, and had some other idiosyncratic semantic properties: see below under class II.

There were four noun classes, corresponding to four different patterns of number marking. Class I nouns were inherently singular, marking paucal number with the suffix '''-rə''' and plural number with CV-reduplication. Nouns referring to humans, some other animals, and a mixed bag of culturally important inanimates fell into class I.

* '''ciri''' "cat" (sg.) '''cirirə''' "litter of cats" '''ciciri''' "cats"

Class II nouns were inherently paucal, marking singulative number with the palatalizing suffix '''-reh''' and plural number with CV-reduplication.

* '''tuswət''' "pair of eyes" (pauc.) '''tuswəceh''' "eye" '''tuduswət''' "many eyes"

Class II nouns fell into three overlapping semantic categories: those referring to objects which occurred in natural sets (eyes, siblings, seasons, hands, petals, teeth), those referring to parts of a larger whole (eye, head, petal, branch, tail, base, corner, inside, outside), and those referring to locations (base, corner, inside, outside, beach, valley, . Apparently, the fact that most bodyparts fall into natural pairs encouraged the paucal to be extended first to unpaired bodyparts like heads, and then to other meronyms like "branch." Often the unpaired bodyparts in class II had an idiosyncratic meaning in the singular:

* '''isip''' "someone's head," sg. '''isipeh''' "a head; stubbornness."
* '''lĩrah''' "someone's tongue," sg. '''lĩraheh''' "a tongue; eloquence."
* '''ceci''' "someone's back or bottom," sg. '''cecireh''' "a back; genitals."
* '''mirəw''' "someone's nose," sg. '''mirəwreh''' "a nose; luck."

The paucal forms of class II bodypart nouns could also be extended to refer to parts of an inanimate object: '''ceci''' for the base of a bench or the bottom of a jar, '''isip''' for the lid of a jar or the top of a bench, and so on. The singular forms of these nouns could not be extended in this way.

Finally, the paucal forms of class II nouns tended to be used for reference to objects that were possessed in the natural way, or occurred in the natural context, while the singular or plural forms were used when the objects were unpossessed or occurred out of context. This can be seen with the bodypart terms above — '''isipeh''' could be used for a disembodied head, and the plural '''izisip''' for a number of heads, attached or disembodied; '''isip''' could only be used for a head attached to its rightful owner. But the same alternation is found on other class II nouns.

* '''larudən suna''' "The seashore" sg. '''sunareh''' "A beach or shore (in general)" pl. '''suzuna''' "Beaches (in general); beaches of many seas"
* '''Ẽtəl''' "Branches (of a tree)" pl. '''Ǝjnẽtəl''' "Sticks; fallen branches; branches of many trees"

Class III nouns, mostly inanimates, were inherently plural. A paucal form in '''-rə''' was used when a natural subset of the things referred to needed to be picked out; the singular was formed with the suffix '''-rə'''.

* '''cirer''' "stars" (pl.) '''cirereh''' "star" '''cirerə''' "constellation"

Class IV was made up of a small number of mass nouns. These were inherently grammatically singular, and had no paucal or plural forms.

=== Noun derivation ===

''-ra'' (strong) derived agent nouns from verbs. ''-ica'' (weak, palatalizing) — combining ''-ha'' with the patient trigger suffix ''-ic'' — derived patient nouns. These suffixes were already becoming less productive in Merneha, and most of the words formed with them have cognates in the other Peninsular languages.

''-wə'' (weak) and ''-mjə'' derived specifically masculine and feminine nouns, either from other nouns that were unspecified for gender or from verbs. The weak, palatalizing patient forms ''-icwə'' and ''-imjə'' ( < ''-icmjə'') are occasionally found.

''-wərəj'' after a consonant (weak) or ''-Ṽhəj'' after a vowel, derived location or instrument nouns from verbs.

== Noun particles ==

=== Determiners ===

Some determiners were freestanding words; others were postclitics. There were no definite or indefinite articles. A noun phrase with no determiner was reliably interpreted as definite if it occurred as a topic, and could be interpreted as definite or indefinite as the context required elsewhere in a sentence.

*''-ku'' — "this" (weak)
*''-su'' — "that" (weak, palatalizing)
*''-he'' — "that over there" (weak)
*''pəka'' — "which?"
*''pahu'' — "what sort of?"

A headless noun phrase was required to have a determiner. If none of the others were semantically appropriate, the absolutive clitic determiner ''-ha'' could be used. This was most likely related to the nominalizing derivational suffix ''-ha'', but unlike that suffix it could be attached to phrasal modifiers as well as single words.

=== Case marking ===

The three true case markers were all postclitics, attaching to the determiner if there was one and the noun stem if there was not. Like many other ancient Peninsular languages, Merneha placed case markers after determiners but before quantifiers.

*''-ma'' — nominative
*''-lu'' — ergative
*''-ən'' — genitive

The nominative case was used for the subject of a clause — which could be an intransitive subject, transitive agent in an agent-trigger clause, a transitive patient in a patient-trigger clause, or an oblique argument in a clause with an applicative-marked verb. See the section on voice for more details.

The ergative case was used for agents that were not subjects. (Again, see under voice for more details on this).

The genitive case had two uses: one for possessors, and one for patients that were not subjects (see under voice). It resulted from the merger of the Proto-Peninsular accusative and genitive cases.

A number of the Proto-Peninsular case particles were reinterpreted as clitic statives in Merneha. They participated in phase concord, and they no longer appeared between the determiner and the quantifier. This includes for instance most of the locative particles. See the next section on statives for more information on these.

=== Quantification ===

Like many other ancient Peninsular languages, Mernesha placed quantifiers ''after'' the case marker.

*''sumeh'' — "every, all"
*''cigic'' — "some"
*''-mə'' — "no, none"
*''pətu'' — "how many?"

Quantifiers could coexist with a determiner, in which case they tended to take a partitive reading (''cirerkuma sumeh'' "all of these stars," ''sũnipemə'' "none of those apples"), or they could be attached to a determinerless noun (''cirerma sumeh'' "each star," ''sũnipmamə'' "no apple[s]"). The only restriction was that a given noun could only have one interrogative attached: ''*suzũnip pəkama pədə'' (intended meaning "how many of which apples?") and similar constructions did not occur.

The cardinal and ordinal numbers, on the other hand, were statives in Mernesha rather than quantifiers.

== Modification ==

A noun could be modified by

* a noun phrase in the genitive case,
* a stative or stative phrase, or
* a finite verb phrase serving as a relative clause.

Modifiers preceded the noun they modified (but see "apposition" in the section on sentence syntax).

=== Relative clauses ===

Relative clause syntax was very simple. First of all, there was no special relative clause marker. Diachronically speaking, this was caused by the reduction of Proto-Peninsular ''*r'' — which formed the participle suffix — to ''*h'' and then ''Ø'' in most contexts within the verb where it could occur. It was sustainable because the ''main'' verb of a sentence always occurred at the end, and was usually followed by a discourse particle. Any verb occurring earlier and not marked by a discourse particle could safely be interepreted as a relative.

Second, there was also no need to indicate the head noun's role in the relative clause. That was because only subjects could be relativized. To relativize an object or an oblique argument, one first took advantage of the voice and valence changing operations available to shift it into the subject slot. See the relevant sections on verbs for more details.

== Headless NPs ==

Any modifier

* ''halaz-ən'' "the chief's"
* ''təg-u'' "blue"
* ''curicic-isk-u'' "(who) quenched their thirst"

can be converted into a headless NP by adding any noun particle.

* ''halaz-ə-ku'' "this one that belongs to the chief"
* ''təg-u-ra'' "a/the blue one"
* ''curicic-isk-u sumeh'' "everyone who quenched their thirst"

Note that even though we are dealing with at least two syntactic words here (a modifier and a determiner/quantifier/case marker) there may only be one phonological word. Word-internal phonological processes — such as the loss of ''n'' before a consonant in ''halazəku'' — proceed as normal.

Multiple noun particles are of course permitted — but remember that ''-ha'' is only used to form headless NPs when there is no other particle present to do the job, so it will vanish if another is added.

* ''halaz-ə-ku-lu'' "this one that belongs to the chief (erg)"
* ''təg-u-lu'' "a/the blue one (erg)" (not ''*təg-u-ra-lu'')
* ''curicic-isk-u-lu sumeh'' "everyone who drank (erg)"

And multiword modifiers can form headless NPs just as easily as one-word modifiers.

* ''hərə pəm-an'' "my father's"
* ''hərə pəm-a-ma'' "nothing of my father's"
* ''hũt-u wərĩhəj-nəmnir-u'' "into the old town"
* ''hũt-u wərĩhəj-nəmnir-u-ma'' "the one (that goes) into the old town (dir)"
* ''medal-an sumeh məməkəmic-isk-i'' "(who) have beaten every opponent"
* ''medal-an sumeh məməkəmic-isk-j-ən'' "of the ones who have beaten every opponent"

= Pronouns =

Unlike the nouns, the Merneha pronouns have distinct forms for the direct, genitive and ergative cases, and do not take any postclitics.

The local pronouns were spared the merger of the Proto-Peninsular singulative and dual suffixes ''*rix'' and ''*ra'', and so distinguish all three numbers. The first person dual form was semantically inclusive, meaning "you and I," and was likely used as a plural inclusive ("you and they and I") as well. The first person plural form was semantically exclusive.

The dual and plural number were not distinguished in the genitive or ergative case.

{| {{bluetable|lightbluebg c}}
|-
! First person
! Direct || Genitive || Ergative
|-
! singular
| həreh || hən || hərełu
|-
! dual
| hərə || hərən || hərəlu
|-
! plural
| hər || hərən || hərəlu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
!Second person
! Direct || Genitive || Ergative
|-
! singular
| naheh || nərã || nahełu
|-
! dual
| nahə || nahən || nəralu
|-
! plural
| nəra || nahən || nəralu
|-
|}

The third person pronouns participated in the singulative/dual merger and so behave like class III nouns — unmarked in the plural, marked in the singular and dual with a suffix. Number was not distinguished at all in the genitive or ergative.

{| {{bluetable|lightbluebg c}}
|-
! Animate
! Direct || Genitive || Ergative
|-
! singular/dual
| ira || iran || iralu
|-
! plural
| irar || iran || iralu
|-
|}

{| {{bluetable|lightbluebg c}}
|-
! Inanimate
! Direct || Genitive || Ergative
|-
! singular/dual
| ireh || irirã || irełu
|-
! plural
| ir || irirã || irełu
|-
|}

= Adverbs and particles =

= Predicates =

Merneha had two types of predicates: true verbs and ''statives''. The distinction between the two was both semantic and morphosyntactic.

From a semantic point of view, it was a distinction based on lexical aspect. Goal-oriented (or ''telic'') events were described using verbs, as were events that were instantaneous or nearly instantaneous. Prolonged atelic events were described using statives. (In Zeno Vendler and Bernard Comrie's system of aktionsarten, semelfactives, achievements and accomplishments were described with verbs, states and activities with statives.)

When it came to describing motion, this meant that verbs encoded things like destination and path shape, while statives encoded the manner of motion. Instead of "run out" one said "exit running," where "exit" was a verb and "running" a stative. Static position was also described using statives: there were verbs for "exit," "enter" and "get up on," but statives for "be inside," "be outside" and "be on top of." Verbs tended to encode the result of a process, while statives were used for the process itself, so instead of "hammer flat" one said "flatten hammering" — "flatten" was a verb, "hammering" a stative. The closest counterparts to most English adjectives were statives. ("Flat," for instance, was a stative; the verb "flatten" was derived from it.)

Surprisingly for speakers of terrestrial European languages, the numbers were also statives. Instead of "three birds" one said "birds being-three."

Morphologically, the statives were much simpler than the verbs. They could be reduplicated, with intensive meaning, and they took a phase suffix (see next section) but were not otherwise marked for aspect or mood, and were not marked at all for tense or voice.

Syntactically, the statives were always modifiers, either in a noun phrase or a clause, while the verbs were usually the heads of their own clauses.

== Grounding ==

All predicates, whether they were statives or true verbs, were marked for ''grounding'' — that is, marked to indicate whether they described foreground or background information. The foreground clitic was '''-u''' (which became '''-w''' after a vowel) and the background clitic was '''-i''' (which became '''-j''' after a vowel).

== Statives ==

The statives were quite morphologically simple. Each could occur in two inflected forms: the plain stative with basic meaning, and a reduplicated form that generally had either intensive, frequentive, distributive or iterative meaning.

* '''kũh-u''' "fat" '''kugũh-u''' "very fat"
* '''kərah-u''' "dark" '''kəgərah-u''' "very dark, dark all over"
* '''hĩmet-u''' "swimming" '''hĩhĩmet-u''' "swimming here and there"

The only clitics they took were the aforementioned grounding clitics '''-i''' and '''-u'''.

A fair number of statives had irregular reduplicated forms; often, this went hand in hand with irregular semantics. For instance, the reciprocal kinship terms add a suffix in their reduplicated forms that's related to the paucal suffix '''-rə''', and they indicate ''mutual'' kinship: "one another's RELATION" rather than "someone's RELATION"

* '''ira kəc-u cigic''' "some of his siblings" '''kəc-u cigic''' "some of (someone's) sibling's" '''kəjgəc.ər-u cigic''' "some people who are one another's siblings"

=== NP complements and adjuncts ===

As the above example suggests, many statives could take an NP complement. This was unmarked for case, but could be marked with a determiner or quantifier as normal.

* '''In kəc-u''' "brother to John"
* '''Hərə pəmə-w''' "my father"
* '''kəhi ujəgĩcwir-u''' "covered in hair" (*uliakinsthuiru)
* '''tilə məgis-u''' "carrying a knife" (*mnkisru)

Comparatives could be formed by giving a stative an adjunct with the strong palatalizing clitic '''-tə-w'''.

* '''piri-tə-w wərn-u''' "bright as day" '''piri-tə-w wərwərn-u''' "brighter than daylight"
* '''Ĩ-tə-w kũh-u''' "as fat as John" '''Ĩ-tə-w kugũh-u''' "fatter than John"
* '''kəhi Ĩ-tə-w ujəgĩcwir-u''' "as covered in hair as John" '''kəhi Ĩ-tə-w ulujəgĩcwir-u''' "more thoroughly covered in hair than John."

=== Incorporating statives ===

A few important statives cliticized onto the last phonological word of their complement. These statives did not have reduplicated forms of their own; rather, they had intensive meaning when attached to the (reduplicated) plural form of a class I or class II noun.

{| {{bluetable|lightbluebg c}}
|-
! !! Just like || During || At/in || From || For/to (benefactive) || Into || Without || With (instrumental)/Compared to
|-
! plain
| ''-tər-u'' || ''-nir-u'' || ''-t-u'' || ''-həl-u'' || ''-n-u'' || ''-nəmnir-u'' || ''-mə-w'' || ''-tə-w''
|-
|}

What set these statives apart from case particles was, first, the fact that they participated in phase concord. (This was an innovation in Merneha; their cognates in other ancient Peninsular languages ''were'' in fact case markers, and were not marked for phase.) The second thing that set them apart from case particles was their syntactic position — they could not appear before the quantifier in a noun phrase, while case particles could.

The incorporating statives beginning in '''t''' were phonologically strong and palatalizing. '''-Həl-u''' was phonologically weak.

=== The relational noun construction ===

The spatial incorporating statives '''-t-u''', '''-həl-u''' and '''-nəmnir-u''' are used in combination with class II bodypart nouns to form locative expressions.

* '''ceci''' "Bottom, backside"
* '''pũd-ən ceci''' "The underside of the bridge" '''[pũd-ən ceci]-c-u''' "Under the bridge (static location); in the space below the bridge" '''[pũd-ən ceci]-nəmnir-u''' "Under the bridge (movement); into the space below the bridge" '''[pũd-ən ceci]-rəl-u''' "Out from under the bridge; from the space below the bridge"
* '''isip''' "Head"
* '''wərĩhəj-ən isip''' "The top of the house" '''[wərĩhəj-ən isi]-c-u''' "On top of the house" '''[wərĩhəj-ən isip]-əl-u''' "Off of the house" '''[wərĩhəj-ən isip]-nəmnir-u''' "Onto the house"

=== Numbers ===

The numbers formed their own paradigm of statives. The reduplicated forms were used to mean "''n'' by ''n''" or "in groups of ''n''".

{| {{bluetable|lightbluebg c}}
|-
! !! 1 || 2 || 3 || 4 || 5 || 6 || 7 || 8 || 9 || 10
|-
! cardinal
| ''cəru'' || ''pəcuru'' || ''məru'' || ''wəreru'' || ''təsu'' || ''məmiru'' || ''nətəru'' || ''ləlahu'' || ''hilucəru'' || ''hədəru''
|-
! ''n'' by ''n''
| ''cəcəru'' || ''pəpəcuru'' || ''məməru'' || ''luwəreru'' || ''tətəsu'' || || || || || ''hihədəru''
|-
|}

== Verb morphology ==

A finite verb consisted of a root — which could be plain or reduplicated — plus an optional applicative prefix or infix and a voice suffix. A verb was followed by a series of clitic particles indicating tense, mood and grounding.

=== Reduplication ===

The reduplicated form of the verb root, like the reduplicated stative, could have frequentive, distributive or iterative meaning. More often, though, especially for transitive verbs, it had an intensive or completive meaning.

* '''məpeh-u''' "eat" '''məməpeh-u''' "eat up"
* '''kətərə-w''' "die" '''kəgətərə-w''' "die off, be exterminated"
* '''təkəbeh-u''' "capture" '''tədəkəbeh-u''' "keep capturing, capture entirely"
* '''cur-u''' "fly" '''təcur-u''' "fly around"
* '''həsur-u''' "bleed" '''hərəsur-u''' "bleed out"
* '''əlum-u''' "go" '''ələlum-u''' "go around"

Certain stems underwent internal, rather than initial reduplication. (These were mostly regular verbs in PPI; this behavior is a result of sound changes: ''*halxasu > həłazu, *halhalxazu > həłəłazu'', etc.)

{| {{bluetable|lightbluebg c}}
|-
! !! Plant || Bleed || Give || Go pale || Urinate
|-
! plain
| ''hiyibu'' || ''həsuru'' || ''wəgudu'' || ''wədəri'' || ''həłazu''
|-
! augment
| ''hiciyibu'' || ''həsəsuru'' || ''wəgũkudu'' || ''wədətəri'' || ''həłəłazu''
|-
|}

A milder form of irregularity caused vowel changes in some reduplicated forms.
{| {{bluetable|lightbluebg c}}
|-
! !! Hunt || Hiccup, flinch || Blow || Testify || Go around
|-
! plain
| ''katərəw'' || ''irapu'' || ''kasəru'' || ''mepəw'' || ''cenu''
|-
! augment
| ''kə-gatərəw'' || ''ir-erapu'' || ''kə-gasəru'' || ''mi-mepəw'' || ''cə-cenu''
|-
|}

A few common roots simply do not fall into any pattern.
{| {{bluetable|lightbluebg c}}
|-
! !! Inhabit || Come, enter
|-
! plain
| ''wəryu'' || ''wəru''
|-
! augment
| ''wərĩyu'' || ''wərĩru''
|-
|}

=== Applicatives ===

There were two applicatives in Merneha, both weak prefixes. ''Ni-'' (from a merger of PPI ''*niç'' and ''*ŋi'') promoted a recipient, beneficiary or destination to direct object. ''Wədə-'' from (''*mta'') promoted an instrument or means.

Most verbs placed the applicative in between the reduplicant and the verb stem.

* ''kə-ni-gatərə-j'' "have hunted on someone's behalf"

Often this lead to irregularities.

* ''tə-ni-jəkəb-i'' "have captured for someone" (''*tanitankapi'')
* ''wəgu-nĩ-kud-i'' "have given (gifts) to someone" (''*mkunimkuti'')

In many cases the applicative form was reanalyzed as a verb root (often with a specialized meaning) and then reduplicated. This could lead to lexical doublets.

* ''nj-ũrə-w'' "cook for"
* ''ũ-nj-ũrə-j'' "have cooked for"
* ''njũrə-w'' "host"
* ''nju-njũrə-j'' "have hosted, be friends with"

=== Voice ===

There were three voice suffixes found on transitive verbs.

* ''-eh'' — agent trigger
* ''-ic'' — patient trigger
* ''-us'' — reflexive/reciprocal

The final consonants of the agent and patient trigger suffixes resist lenition. The final consonant of the reflexive suffix does lenite. Indeed, it only shows up unlenited in two contexts: before the potential particle (''kədərus-ła'' "might talk to each other") and before the optative (''kədərus-nə'' "let them talk to each other").

These could combine freely with the applicative prefixes.

* ''nĩ.kud.eh-u'' "give (gifts) to (agt trigger)"
* ''nĩ.kuj.ic-u'' "give (gifts) to (pt trigger)"
* ''nĩ.kud.uz-u'' "give (gifts) to one another"
* ''wədə.bəkmj.eh-u'' "hunt with (agt trigger)"
* ''wədə.bəkmj.ic-u'' "hunt with (pt trigger)"
* ''wədə.bəkmj.uz-u'' "hunt together"

All three transitive voice suffixes were weak (that is, they caused lenition of a preceding consonant). The agent and patient trigger suffixes formed strong stems (that is, their final consonant did not itself lenite), while the reflexive/reciprocal suffix formed weak stems.

A transitive verb could also be used with the passivizing suffix ''-rl'' after a vowel or ''-əl'' after a consonant.

* ''wəgud.əl-u'' "be given (as a gift)"
* ''nĩ.kud.əl-u'' "receive (a gift)"
* ''pəkmi.rl-u'' "be hunted"
* ''wədə.bəkmi.rl-u'' "be used for hunting (of a weapon), be taken along hunting (of a person)"

=== Derivation ===

Aspect in Merneha was lexical rather than grammatical. Many of the verb derivation operations caused changes in lexical aspect.

A number of suffixes were used to form perfects of result: '''-ir''', '''-in''', '''-ic''', '''-wi''', '''-huj''', and '''-Ṽt''' (after a vowel) or '''-ət''' (after a consonant).

*'''hidibeh-u''' "plant" '''hidibiceh-u''' "have planted"
*'''kədər-u''' "speak" '''kədərəd-u''' "have spoken; finish speaking; hold one's peace"
*'''irug-u''' "go" '''irukuj-u''' "have gone, be gone"
*'''inər-u''' "pray" '''inərəd-u''' "have prayed; receive"
*'''wərjeh-u''' "live, dwell in" '''wərineh-u''' "have dwelled in; hail from"

'''-haz''' forms causatives, either replacing the agent of a transitive verb with a cause or adding a cause to an intransitive verb.

* '''əjub-u''' "fall" '''əjupazeh-u''' "make fall, topple (agt trigger)"
* '''pəkmjeh-u''' "hunt" '''pəkmirazeh-u''' "cause to be hunted; banish, shun, outlaw"

== Verb particles ==

Any verb may be followed by the past tense clitic '''-isk'''.

Following the tense clitic (if any) comes a mood and grounding clitic.

* '''-u''' — foreground indicative
* '''-i''' — background indicative
* '''-ənə''' — negative
* '''-əha''' — subjunctive
* '''-la''' — potential
* '''-jə''' — imperative
* '''-wənə''' — prohibitive

== Derivational morphology ==

= Clause-level syntax =

== The copula ==

The copula was the strong postclitic ''-rar-u'', past tense ''-rar-isk-u''.

* '''Turəha-rar-u''' "it's a fisherman"
* '''[Hərə pəmə]-rar-u''' "it's my father"

Like a verb, it could take the full complement of non-indicative final suffixes.

* '''[Hərə pəmə]-rar-ənə''' "it's not my father"

Unlike a verb, it did not participate in any of the voice or transitivity alternations. It also did not have an reduplicated form of its own, but like the incorporating statives, it could be attached to a plural noun phrase with intensive, completive, frequentive or generic meaning.

* '''Tudurəha-rar-u''' "it's always a fisherman; they're all fishermen; they keep being fishermen"

Often copular sentences with plural noun objects can be translated using an adverbial quantifier like "always."

* '''[Kurir pəbəmə]-rar-u''' "it's always his father."
* '''Pwərwə-n cəras-i medalə-ma [kurir pəbəmə]-rar-u.''' "A boy's first opponent is always his father."
* ''Pwərwə-n cəras-i medalə-ma [kurir pəmə]-rar-u.''' "This particular boy's first opponent is (in this case) his father."

Similarly, copular sentences with dual noun objects often have reciprocal meanings.

* '''Kura-ma kəjwə-rar-u.''' "They're (someone's) brothers."
* '''Kura-ma kəjwərə-rar-u.''' "They're (one another's) brothers."

The object of a copula phrase could also be a stative.

* '''Nuh-u-rar-u''' "(s)he's whistling"
* '''[Lwəgwə kəhir-u]-rar-u''' "it's over the river"
* '''[Hərə pəmə-t-u]-rar-u''' "it's for my father"
* '''[Hərə pəmə-həl-u]-rar-u''' "it's from my father"

Recall, though, that what looks like a bare stative can be analyzed as a headless noun phrase with a stative modifier. If you look at it that way, the rule is simpler: a copula phrase ''always'' consisted of a noun phrase — with or without a head — plus ''-rar-u'' or ''-rar-isk-u''.

* '''[[Nuh-u] minə-zu]-rar-u''' "she's that woman who's whistling"
* '''[[Nuh-u]-zu]-rar-u''' "(s)he's that one who's whistling"
* '''[[Nuh-u] ]-rar-u''' "(s)he's someone who's whistling" --> "(s)he's whistling"

== Copular sentences ==

Existential sentences could consist solely of a copular phrase, with no subject. (Note that in the first example, there is no determiner and the copula attaches to the noun; in the second, there is a freestanding determiner, and it's the determiner that the copula attaches to.)

* '''Turəha-rar-isk-u.''' "There was a fisherman."
* '''[Kinək cigic]-ar-u.''' "There are some beans."

=== Case ===

If there was a subject in a copular sentence, it was in the nominative case.

* '''[Hərə pəmə]-ma turəha-rar-isk-u.''' "My father was a fisherman."
* '''Kədərət-ma [hərə pəmə-həl-u]-rar-u.''' "The message is from my father."

Note that copular sentences are symmetrical in meaning: if X IS Y then Y IS X. So rearranging the sentence to keep the topic in subject position did not require any voice transformations — just a change in constituent order.

* '''Turəha-ma [hərə pəmə]-rar-isk-u.''' "The fisherman was my father."
* '''[Hərə pəmə-həl-i]-ma kədərət-ar-u.''' "The thing that came from my father is a message."

(Note that '''hərə pəməhəlu''' "from my father" was foregrounded when it was part of the comment. When the same phrase is topical — and hence old information — it is backgrounded: '''hərə pəməhəli'''.)

=== Locative sentences ===

Locative sentences were constructed in the same way, using a headless NP with a locative stative modifier as either the subject or the object of the clitic.

* '''[Lwəgwə-he ceci-c-i]-ma təməs-ar-u.''' "Across [''lit.'' "behind"] yonder river is a village."
* '''Təməs-ma [lwəgwə-he ceci-c-i]-rar-u.''' "The village is across yonder river."

(Note that '''lwəgwəhe cecici''' is backgrounded in the second example above. This would be used if the hearer already knew about the river — and, presumably, knew that the river ''had'' a far side — and he just needed to be informed that the village was ''on'' the far side. If the river was new information too, one would say '''Təməsma lwəgwəhe cecituraru'''.)

=== Possessive sentences ===

Possessive sentences were constructed using a NP with the instrumental stative clitic ''-tə-w.''

* '''Hə [kurə-tə-w]-rar-i'''. "I have [''lit.'' "am with"] a brother."
* '''[Kurə-tə-j]-ma hə-rar-i'''. "The one with a brother is me."

=== Clefts ===

Finally, true verbs could occur in the subject of a copular sentence, as relative clause modifiers of a headless NP. This formed a cleft, in which the verb itself was old information and one of its arguments was asserted as new.

* '''Katərarwə-ma hərə-n nĩkujic-i.''' "The hunter is giving [gifts] to me."
* '''[Katərarwə nĩkujic-u]-ma hə-rar-i.''' "The one the hunter is giving [gifts] to is me."
* '''[Hərə nĩkudeh-u]-ma katərarwə-rar-i.''' "The one who's giving [gifts] to me is the hunter."

== Verbal clauses ==

In a clause with a full lexical verb, voice operations and the applicative prefixes were used to ensure that the most topical constituent was the subject.

=== Case ===

In a transitive clause, the second constituent after the subject was always the non-subject argument — the object of an agent-trigger verb, or the agent of a patient-trigger verb.

Objects were marked for genitive case

and the agents of patient-trigger verbs were marked for ergative case.

=== Clausal modifiers and obliques ===

Between the non-subject argument and the verb itself came statives modifying the whole sentence. These could express location,

source, destination or beneficiary,

instrument or co-agent,

temporal circumstances

or the manner in which an action was done.

If the patient of a transitive verb was demoted — for instance, because an applicative prefix replaced it as one of the verb's arguments — it could be expressed using the instrumental stative '''-tə-w'''.

== Coordination and subordination ==

== Questions ==

= Fish story =

'''Piriniru, turəhama lərut təwməri sunaci piri sumeniri turisku ca hicərihanmə ca cəru cijiskwən ira mihisku turicehisku.'''

Piri-nir-u, turəha-ma lərut təwmər-i suna-c-i piri sume-nir-i tur-isk-u ca hicə.rih-an-mə ca cər-u cijisk-w-ən ira mih-isk-u turic.eh-isk-u

Day-cop-P fisherman-NOM sea beside-P beach-at-P day all-cop-P fish-PAST-P and fish.SG-GEN-no and one-P small-P-GEN he.NOM see-PAST-P catch.agt-PAST-P

"One day, a fisherman fished all day on the beach by the sea, but he didn't catch anything except for one small fish."

"See + VERB" is idiomatic for "Manage to verb." Also, note '''ca hicərihanmə ca cəru cijiskwən''' "Both no fish and one small one" for "No fish except one small one."

= Lexicon =

Add to SCA — əjNC --> ẽC, əwNC --> ãC, VjNC, VwNC --> ṼjC, ṼwC
Duals in -reh not -rə.

=== Bodypart terms ===

Bodypart terms and some other meronyms are inherently ''dual,'' reflecting the bilateral symmetry of the body. The plural form denotes more than one body's worth: '''hukus''' "pair of arms," '''hurukus''' "many arms"; '''tuswət''' "pair of eyes" '''tuduswət''' "many eyes."

For parts that appear more than two to a person, the dual refers to a complete set: '''nit''' "a person's teeth," '''nitə''' "a single tooth," '''ninit''' "many teeth"; '''łər''' "ten fingers," '''łərər''' "a finger," '''həłər''' "many fingers." For these, the plural does not necessarily describe ''more'' than one body's worth of parts, but it definitely suggests more than one person is involved. You can say that a group of people '''həłərən kudehu''' "raised (their) fingers" if they all raised just one finger.

For singular bodyparts — "head," "back," "tail," "nose," and so on — the dual has a singular meaning, and the singulative form often has a figurative meaning of some sort. '''Isip''' "head," '''isipe''' "face"; '''lĩrah''' "tongue," '''lĩrarər''' "eloquence"; '''ceci''' "back," '''cecirə''' "genitals"; '''mirəw''' "nose," '''mirəwrə''' "luck."

'''Ceci''' ''n. II'' "Back, bottom,"
* sg. '''Cecirə''' "Genitals (lit. 'back part')"

'''Cigic''' ''quant'' "Some"

'''Ǝlum-u''' ''v.i.'' "Walk," perf. '''Ǝləlum-u''',
* appl. '''Wədalumeh-u''' "Walk with (someone)," '''Wədalumuz-u''' "Walk together, court with."

'''Ǝtłəjn''' ''n. I'' "Rope," du. '''Ǝtłẽre''', pl. '''Ǝrətłəjn'''.

'''Ǝrjub-u''' ''v.i.'' "Fall," perf. '''Ǝrərjub-u'''.

'''Ẽtəl''' ''n. II'' "Branches of a tree," sg. '''Ẽtəlrə''' "Stick, branch," pl. '''Ǝjnẽtəl''' "Sticks."

'''Həbə''' ''n. II'' "Clan," sg. '''Həbərə''' "Clansman, relative," pl. '''Hərəbə''' "Clansmen."

'''Həłaz-u''' ''v.i.'' "Urinate."

'''Hənəh''' ''n. II'' "Flower," sg. '''Hənəhə''' "Petal," pl. '''Hərənəh''' "Flowers."

'''Hərnə-w''' ''v.i.'' "Return, go back,"
* perf. '''Hərərnə-w''' "Have returned, be home,"
* appl. '''Wədərərnəj-u''' "Return to," '''Nirərnəj-u''' "Go back for,"
* caus. '''Hərnərazeh-u''' "Send back, complete, finish."

'''Həsuretər-u''' ''stat.'' "Weak."

'''Həsur-u''' ''v.i.'' "Bleed,"
* caus. '''Həsuhazeh-u''' "Empty (something) out."

'''Həzut''' ''n. III'' "Fungus, mushroom," sing. '''Həzutə'''.

'''Hẽnuh-u''' ''stat.'' "Asleep."

'''Hẽnurwərəj''' ''n. I'' "Bed," pl. '''Herẽnuwərəj'''.

'''Hihəceh-u''' ''v.t.'' "Make wider."

'''Hihər-u''' ''stat.'' "Wide, broad."

'''Hijib-u''' ''v.i.'' "Become pregnant," perf. '''Hicijib-u'''

'''Hijibeh-u''' ''v.t.'' "Plant."

'''Hijibic-u''' ''stat.'' "Wealthy."

'''Hĩmətazeh-u''' ''v.t.'' "Bathe (someone)," perf. '''Hĩhĩmətazeh-u'''
* refl. '''Hĩmətazuz-u''' "Bathe oneself."

'''Hĩmət-u''' ''stat.'' "Swimming."

'''Hiniruk-u''' ''stat.'' "Mortal, frail."

'''Hin-u''' ''v.i.'' "Die."

'''Hirẽraj-u''' ''stat.'' "Dead, passed on (of a person, esp. an ancestor)."

'''Hizər-u''' ''stat.'' "Narrow."

'''Hugək''' ''n. III'' "Trees," sg. '''Hugəkə'''.

'''Hũhir-u''' ''v.i.'' "Dream,"
*appl. '''Nihũhireh-u''' "Dream about, dream of."

'''Hukus''' ''n. II'' "A person's arms," sg. '''Hukusə''' "Arm," pl. '''Hurukus''' "Various arms."

'''Humə''' ''n. I'' "Mother," pl. '''Hurumə,'''
*du. '''Huməreh''' "Mother and child."

'''Hunəjn''' ''n. II'' "A set of bones, a person's bones, a skeleton." Sg. '''Hunəjrə''' "A bone," pl. '''Hurunəjn''' "Various bones."

'''Hũt-u''' ''stat.'' "Old."

'''Hũtũri-w''' ''v.i.'' "Age, grow old,"
* appl. '''Nirũtũreh-u''' "Wait for."

'''Inər''' ''n. II'' "Prayer," sg. '''Inərə''' "Verse," pl. '''Ininər''' "Prayers."

'''Inərwə''' ''n. I'' "Priest."

'''Inərmjə''' ''n. I'' "Priestess."

'''Inər-u''' ''v.i.'' "Pray."

'''Inərwərəj''' ''n. I'' "Temple"

'''Irug-u''' ''v.i.'' "Go."

'''Irukaj-u''' ''v.i.'' "Be on its way, be about to happen."

'''Isip''' ''n. II'' "Head," pl. '''Izisip''' "Heads,"
* sg. '''Isipə''' "Face."

'''Jən''' ''n. I'' "Chief," du. '''Jəreh,''' pl. '''Jərjən'''.

'''Kasəra''' ''n. II'' "The Winds as an object of worship," sg. '''Kasərarə'''.

'''Łər''' ''n. II'' "A person's fingers," sg. '''Łərər''' "A finger," pl. '''Həłər''' "Various fingers."

'''Lĩrah''' ''n. II'' "Tongue,"
* sg. '''Lĩrarər''' "Eloquence."

'''Mənə''' ''n. I'' "Bay."

'''Mirəw''' ''n. II'' "Nose,"
* du. '''Mirəwrə''' "Luck."

'''Mər-u''' ''stat.'' "In-law,"
* intens. '''Məmərər-u''' "One another's in-laws."

'''Wəjlu''' ''n. II'' "Tail."

'''Wərərugək''' ''n. III'' "Deciduous trees," sg. '''Wərərugəkə.'''

'''Wərnu''' ''stat.'' "Red."

'''Wiłk''' ''n. III'' "Crop, harvest"
* sing. '''Wiłkə''' "One harvested fruit."