Jamna Kopiai

Jamna Kopiai [ˈʤamna koˈpi.aj]
Period	c. 1500 YP
Spoken in	northern Zeluzhia
Total speakers	c. 500,000
Writing system	unknown
Classification	unknown
Typology
Basic word order	SVO
Morphology	agglutinative
Alignment	accusative
Credits
Created by	Radius

Jamna Kopiai (adjective: Jamna) is a language spoken natively by perhaps half a million people, on the savanna of northern Zeluzhia around 1500 YP. There are additionally several million second-language speakers, as Jamna is a regional lingua franca.

Jamna Kopiai is a synthetic, agglutinating language of principally right-branching syntax and nominative-accusative morphosyntactic alignment. The verb's morphological complexity is comparable to that of many polysynthetic languages, but we will avoid that label here, as it is prone to evoking expectations of noun incorporation or strong polysynthetic personal marking - both of which are largely absent in this language.

The language also has a very small phonology comprising only fourteen contrasting segments, and is highly vocalic, permitting few consonant clusters but quite long vowel sequences.

phonology

consonants

	labial	coronal	velar	glottal
stop	p	t	k - g
nasal	m	n
fricative	f	s		h
affricate		ʤ
flap		ɾ

spelling

We will spell ʤ and ɾ with j r.

clusters

Only seven consonant clusters appear frequently: pr kr gr sr mn nj ng. Two additional clusters, fr hr, are far less common. Clusters with r are permitted in word-initial position; all may appear medially.

gemination

Consonant gemination is generally contrastive, but there remain many predictable alternations of single consonants with geminates, especially in nominal collective forms. In this alternation, m j g become mn nj ng. There are no phonetically geminated versions of these consonants. Clustered consonants do not geminate, and geminated consonants (including mn nj ng) do not appear in initial or final position.

final consonants

Final consonants are permitted in all word types except verb roots, but are nevertheless uncommon, except for those which are suffixes. In no case whatsoever may a word end in a consonant cluster.

the break

Many morphemes behave as though they have a consonant that isn't pronounced. This is the "break".

There is a hidden "consonant" present in the underlying representation of many morphemes, that is normally realized on the surface by zero, except for its effect on stem reduction. Deriving from a historical glottal stop, the "break" is now a null entity that remains necessary in order to explain syllabification patterns and stem reduction behavior (see below). It is unwritten, and can only be found in two environments: at the ends of some words, and in the midst of some word-final vowel sequences. In all other environments, no traces remain of former glottal stops.

All monoconsonantal content roots contain or end in a break.

vowels

	front	central	back
high	ɪ
mid			o
low		a

distribution

Vowels can appear in arbitrarily long sequences, but no more than two like vowels can appear in a row. That is, /aaa/ reduces to aa. Doubled vowels reduce to single vowels before a geminated consonant - but not a cluster - unless the underlying sequence is of three or more like vowels, which are never reduced to a quantity fewer than two.

syllabification

Syllabification is not completely predictable. The surface representations of the vowels /i o/ often consist of the glides [j w]; the rules for when they become glides are complex, and have not been fully investigated. The best treatment we can give the topic at this time is to list four rules that appear to hold true in most words, though they often conflict, and in such conflicts the precedence of some over others may vary by speaker or speech variety.

• Vowel sequences that can become falling diphthongs, ai ao oi, are preferentially treated as syllable nuclei - unless a break intervenes between the two vowels, in which case they are assigned to separate syllables.
• A sonority hierarchy is observed: a > o > i. A less sonorous vowel generally becomes a glide before a more sonorous one - again, unless a break intervenes, and except for the oa-rule below.
• A three-vowel sequence is preferentially broken into a bisyllable rather than a triphthong, unless the central vowel is a. That is, oio is always [o.jo], while the more problematic ioi is often [ɪ.wɪ], though this varies with [joj] for some speakers, which is more consistent with the previous two constraints. (Variations like this are what leads us not to posit phonemic glides or diphthongs.)
• Most speakers break oa into a bisyllable rather than [wa] if it directly follows a consonant. For most of these speakers, it is poorly or not at all distinct from ooa. Many, but fewer, treat ia io in a parallel manner.

further notes

1. Intervocalic glides do not clearly syllabify to the left or right, but show a gliding transition in both directions. For example, though naoi is always [ˈnawɪ], neither syllable nucleus is realized as a monophthong: they are [naw] and [wi]. Thus the syllable break in such words may be considered to consist of the medial glide, rather than being placed to the left or right of it.

2. A double vowel unadjacent to any others does not show gliding, but behaves like a long vowel. The double vowels ii oo, when pronounced as long vowels, are both fully high: [iː uː], but the short varieties (which appear far more often) are typically lower in articulation: [ɪ o].

stress

Stress is consistently placed on the penultimate syllable of a word, unless monosyllabic. Stress placement is unpredictable only as a result of situations where the relevant syllabification is unpredictable; universally, words ending in a sequence of CVCV(C) will stress the first vowel in this sequence (i.e. the penultimate), but stress assignment in words ending with other sequences occurs only after they are syllabified.

The stress accent is realized by both dynamic and pitch-contour effects. The accented syllable is pronounced more volubly and carefully than surrounding ones; and in native words, it is also the point of lowest pitch in the word, with the following syllable displaying a notable rise in pitch. However, a substantial number of loanwords gained from the nearby (and substrate) language Tipatirápai were borrowed with their native pitch contours approximately intact. In general, these words have a peak in pitch on the stressed syllable, which we mark here with an acute accent. Though Tipatirápai does not always place its stress on the penultimate syllable, the Jamna have generalized this as their pattern for loanwords, sometimes even for those borrowed from languages other than Tipatirápai.

Notably, however, when accents are moved to the right as morphology lengthens the word, pitch peaks remain where they were and the moved accent takes the native low pitch instead. Thus the accusative of the language name Tipatirápaioa still bears the acute mark in the same location despite it having gained an extra syllable: the true accent is on the pai syllable, but the preceding ra still bears a pitch peak.

example words

• naoi "to find" [ˈnawɪ]
• raaipai "village" [ɾaˈajpaj]
• haooi "field" [ˈhawoj]
• koai "ball, sphere" [ˈko.aj]
• kiori "year" [ˈkjoɾɪ] or [kɪˈoɾɪ]
• ojoo "to die" [ˈoʤuː]
• miitoisai "hallway" [miːtojˈsa.i] (this word has a break)
• jittiga "to awaken" [ʤɪˈtːɪga]
• srami "priest" [ˈsɾamɪ]
• Jamna "the Jamna people" [ˈʤamna]
• Hiiaona (man's name) [hɪjawna]

morphology

issues in Jamna morphology

categorization

Some of the grammatical morphology of Jamna Kopiai is simultaneously inflectional and derivational; often, morphemes display traits characteristic of both categories. For example, the verbal stem formants operate in both the lexical and inflectional realms, fundamentally altering meanings of roots and deriving new verbs from other parts of speech, while at the same time occurring in a mandatory and organized paradigm, but from which no member is fully productive. With this overlap between the traditional morphological types, we have chosen instead to divide everything between syntactically important morphology that is required in order to speak grammatically on the one hand, and purely semantic morphology, which changes the meaning of roots optionally or as a lexical process, on the other. The two major subdivisions of this morphology section address these types.

stem forms

Most nouns and verbs have two stems apiece - the full stem, and the reduced (or "short") stem. As a morphophonological behavior, some suffixes on verbs and nouns trigger stem reduction and others do not. Suffixes beginning with vowels typically trigger reduction and those with consonants typically don't, but there are various exceptions to both tendencies, which will be noted when appropriate in this sketch. It is also generally true that suffixes which cause reduction are of older vintage while those that don't tend to be more recent grammaticalizations. Reduction also applies to the first element in most compound words, and sometimes to other affixes.

Stem reduction normally consists of deleting one or more final vowels. Although a few words display irregular reduction, the general rule is that all vowels are removed after the root's last consonant, or a "break" (see above). If the root ends in a consonant or break, no reduction can occur.

There is only one systematic exception to stem reduction, and that is proper nouns. Certain proper nouns do undergo reduction, but most do not.

Numerous examples of stem reduction can be found in this document. For instance, the accusative of Jamna is Jamnoa, not *Jamnaoa.

grammatical morphology

nouns

case and class

Nouns in Jamna Kopiai inflect to show accusative case with the suffix -oa, which attaches to the reduced stem, and to show the dative case with the suffix -an, which does not.

The nominative case of a noun varies by its class: A-class nouns are unmarked in the nominative, while I-class nouns take the suffix -i.

A-class
nioo child
niooa child.ACC
niooan child.DAT

I-class
gaihi wife
gaihoa wife.ACC
gaihan wife.DAT

It is important to distinguish the I-class nominative suffix from A-class nouns whose roots happen to end with i. For example, jigi "eye" is A-class. Such final i is only deleted before the accusative suffix as part of stem reduction, and always retained before the non-reducing dative suffix, whereas the I-class suffix -i is always replaced by the accusative or dative suffixes.

A-class
jigi eye
jigoa eye.ACC
jigian eye.DAT

In certain situations the dative case can be "stacked" onto the accusative case, resulting in -oan.

number

plurals

Most nouns can be made plural by full reduplication. This is optional, and limited mainly to situations where confusion might otherwise arise.

sori "bone" > sori-sori "bones"
krai "horse" > krai-krai "horses"

In the accusative and dative cases, both instances of the reduplicated noun take the case suffix.

soroa "bone" > soroa-soroa "bones"

duals

Certain nouns can instead take the suffix -nan. This is a dual form, and is used for things that naturally come in pairs. However, -nan is mutually exclusive with any case suffix. Therefore, only A-class nouns mark dual number, and only in the nominative.

jigi "eye" > jiginan "eyes"
jigoa "eye, eyes"

collectives

Collective nouns can be formed by reduplicating only the first CV. The newly medial C is then geminated, unless part of a cluster.

sori "bone" > sossori "skeleton"
krai "horse" > krakrai "herd of horses"
jipoi "chip" > jinjipoi "mosaic"

If the root begins with a single or double vowel, the null onset is represented by /h/ in the reduplicant.

ikaa "mountain" > hiikaa "mountain range"

There are a number of irregular collectives. Noun roots with three or more syllables are more likely to be irregular, as are those with a medial /h/, and those beginning in two or more vowels. Some historical reduplicated collectives that have resulted in irregular forms are not always recognized by speakers as based on the same root, and may have undergone substantial semantic drift. A few of these have even developed new regular collective forms of their own.

nihii "person" > niinii "tribe, nation"
kahapai "action" > kakkapai "temperament"
iaofa "twig" > haoffa "thicket, complicated situation" > hahhaoffa "politics"

Normally, plural and collective reduplications are mutually exclusive. When needed, quantifiers are employed for indicating plurality.

*sossori-sossori "skeletons"

pronouns

The main pronouns of Jamna Kopiai come in two flavors, personal pronouns and haaioari. All members of both sets behave in syntax as A-class nominals.

personal pronouns

The set of free personal pronouns is limited to speech act participants, and they distinguish only case, person, and clusivity, not grammatical number as such.

	Nominative	Accusative	Dative	gloss
1 (excl)	taa	taai	taan	I and maybe others but not you
1+2 (incl)	aago	agoa	agoan	both of us and maybe others
2	sii	sioa	siian	you and maybe others but not me

Additionally, the verbal transitive formant -k sometimes acts as a bound pronoun for third person accusative.

haaioari

The haaioari (collective form of ai-o-ari, meaning "this-that-which") are the Jamna demonstratives and closely related words, which are all formed from a tightly organized set of only twelve agglutinative morphemes (or thirteen, if you include the zero root as a morpheme). However, the haaioari are definite words only; indefinite pronouns (and their kin), like "somewhere" or "nobody" or "anything", are poorly systematic and often lacking entirely.

The first set may be considered the root morphemes, if any can, although one of them is represented by zero and none of them can appear independently.

• Ø- thing
• sa- person
• ho- person (familiar)
• ji- place
• ki- time
• ana- manner
• po- reason

The zero root is also used whenever an aioari is used as a determiner for a noun.

These roots may take the suffix -jin to form basic, A-class nouns meaning the same thing: sajin, "person"; pojin, "reason"; and so forth. Such nouns are considered the most basic terms for each of these things, but all of them have various synonyms that are not part of the system, and hojin has come to specifically refer to children although ho- in its other uses is not so specific. There is no such noun form for the zero root.

For all other (non-noun) uses, there are optional suffixes for "to" and "from", followed by mandatory suffixes for "this", "that", and "which?". The to- and from- forms, which keep their vowels only before -ri, are:

• -m(o)- to
• -g(i)- from

The mandatory suffixes, from which the system takes its name, are:

• -ai this
• -o that
• -ri (query)

These final morphemes are required, whereas the middle set are optional and the first set includes zero, so these three can appear on their own.

The following tables list the full range of nominative forms for two roots: 0 "thing" and ji- "place".

	this	that	which?
basic	ai "this"	o "that"	ari "which?"
to	mai "to this"	mo "to that"	mori "to which?"
from	gai "from this"	go "from that"	giri "from which?"

	this	that	which?	noun
basic	jiai "here"	jio "there"	jiri "where?"	jijin "place"
to	jimai "hither"	jimo "thither"	jimori "whither?"
from	jigai "hence"	jigo "thence"	jigiri "whence?"

notes

• The root ho- is not used in query forms; only sa- is used for asking "who?".
• Otherwise, all other roots have comparable tables to these.
• After sa- and ana-, the -ai suffix takes the form -oi.
• When serving as pronouns, all haaioari take accusative and dative cases wherever appropriate, and are grammatically A-class. They do not reduce before case suffixes.

ai haooi
this field-NOM
this field

Taa poai kogriingai jimo.
1.excl-NOM reason-this INT-travel-FMT place-to-that
For this reason I will travel to there.

verbs

morphological structure

Jamna verbal morphology is complex. A verb can appear with as many as ten grammatical affixes at one time, not including the rich semantic morphology. However, only some of these occur in well-defined slots; the remainder are ordered on a syntactic rather than a templatic basis. That is to say, many morphemes take positions that are dependent on the presence or absence of others.

The suffixal morphology is found in three well-defined slots, which we will name S1, S2, and S3. There are two well-defined prefixal slots, P1 and P2, but the remainder of the morphology - especially the TAM Complex ("TC") - is syntactically ordered. In particular, the plural marker has the appearance of changing position dramatically, appearing sometimes in S2 and sometimes clear on the other side of the verb among the prefixes, although this too is rule-governed.

The general template can be described as follows, with morpheme types listed to the extent possible. Only the nucleus and a formant are required.

P1	P2	TC	NUCLEUS	S1	S2	S3
negative or interrogative	modal	tense, aspect, mode	NUCLEUS	formant	plural or participle	formant

Note: Strictly speaking, the verb "stem" consists of the root, any semantic morphology, and the formant. Thus neither "root" nor "stem" are appropropriate terms for the core element of the verb, so we will term this the "nucleus" instead.

number of verb forms

Between these slots there are a total of 22 non-zero morphemes that may attach to the verb. Of these, formants are not evenly distributed (no verb can take them all), so we must count them out of this calculation. That leaves 16 non-formant morphemes, which in their possible combinations can produce about 2,700 possible verb forms. This is further extended in poorly calculable ways by the formants, and by semantic morphology within the nucleus. Certain verbs may exceed a hundred thousand possible morphological instantiations, perhaps even a million.

P1: negative and interrogative

The first slot is home to the negative and interrogative prefixes. These are sri- and to- respectively.

Taa srikrinoso.
taa sri-krino-so
1.excl-NOM NEG-die-FMT
"I'm not dying."

Taa tokrinoso.
taa to-krino-so
1.excl-NOM ?-die-FMT
"Am I dying?"

P2: modals

The second slot houses optional morphemes that supplement the modal and aspectual possibilities offered in the TC.

• jo - deontic
• him - hypothetical
• goa - inceptive
• jasooi - opinionated

TAM complex

The expression of tense, aspect, and mode on the verbs of Jamna Kopiai is manifest by an intricate interplay of prefixes. The structure of their combination is covered in the syntax section, under "TAM usage: the TC".

First, we shall list the morphemes that can be found in the TAM Complex.

• i - general past
• inni - relative past
• na - imperfective
• so - perfect
• kok - intentive
• ago - assertive
• fi - abilitative
• haat - energetic
• a - plural (sometimes)

Only nine affixes, but up to four of them can appear in any given verb, and the past-perfective i can even appear twice, giving a maximum total of five morphemes in the TC. All of the above (save for the plural) can appear as sole prefixes on a verb, in addition to being combinable.

Taa agokrinoso.
taa ago-krino-so
1.excl-NOM ASRT-die-FMT
"I'm really dying."

Taa ifiagojooik.
taa i-fi-ago-jooi-k
1.excl-NOM PAST-ABL-ASRT-see-FMT
"I was indeed able to see it."

S1 and S3: stem formants

The argument structure of verbs in Jamna Kopiai is strongly reflected in morphology. It is marked via a set of suffixed verbal formants, which are - except on naked verbs - required in order to situate a root into discourse as a finite verb. Except for transitive -k, the nature of the system defies any quick listing of meanings for particular suffixes; they operate in both the lexical and grammatical realms and require a close examination to be at all understood. This is given in the syntax section under formant usage.

There are five S1 formants:

• -gai
• -s, -so
• -r, -ar
• -nja
• -ti

And one S3 formant:

• -k (transitive)

The appearance of the allomorphs of -s and -r is poorly predictable.

S2: participle and plural

The S2 slot may be considered the "native" location of the plural suffix -a - but if there is any material in the TC, the plural marker is moved to that part of the verb, under rules covered previously. To recap: the plural suffix is attached to the furthest verbal element to the left, whether that is the main verb stem or a prefixed auxiliary.

The plural marker agrees only with plural A-class subjects; I-class subjects do not trigger agreement.

Participles are formed with the suffix -iag in slot S2. This is one of two non-finite verb formatives, along with the infinitive (which consists of the bare verb nucleus). Unlike infinitives, participles allow the full range of verbal morphology to be expressed, except for the plural -a which is mutually exclusive with -iag (except where it has been moved to the TC).

classes of irregular verb

Regular verbs are those which behave exactly as described in the previous sections. But we must also briefly discuss the two major classes of irregular verb: naked verbs and rhotic verbs.

naked verbs

Certain verbs are unable to take stem formants. Their argument structures are lexically determined, and subject to little if any modification. This class includes the copular verbs, and certain others that appear frequently in discourse, as well as many loanwords. A minority of naked verbs can or always take the S3 formant -k.

Naked verbs always take plural agreement and participial -iag as usual, but they do also display a preference for avoiding most prefixes, especially in the TAM Complex. To be more explicit about these preferences:

• P1 - regular
• P2 - dispreferred
• TC - dispreferred
• S1 - absent
• S2 - regular
• S3 - only certain verbs

Diachronically, it appears likely that the naked verbs derive from what had been a small class of intransitive-only verbs in a prior stage of the language, whose other verbs were all default-transitive.

rhotic verbs

Rhotic verbs are verbs with vowel-initial stems that prefix a meaningless, thematic r- to the stem if it is directly preceded by any of the following prefixes: past i-, anterior inni-, deontic jo-, or inceptive goa-.

Example: apiik, "to hold/carry"

Taa apiik.
taa apii-k
1.excl-NOM hold-FMT
"I'm holding it."

Taa irapiik. (not *iapiik)
taa i-r-apii-k
1.excl-NOM PAST-THM-hold-FMT
"I held it."

There are at least a hundred rhotic verbs in Jamna Kopiai. A much smaller number - perhaps a dozen, in total - instead prefix a thematic h- that works in exactly the same way. These have been called "h-rhotic" verbs.

semantic morphology

We use the term semantic morphology to describe affixes in Jamna Kopiai which are optionally added when the speaker finds them semantically appropriate or which primarily serve to derive new words. These two functions are not clearly differentiable: it could well be said instead that speakers easily create new words "on the fly".

The great majority of Jamna semantic morphology is found in the verb nucleus.

nuclear structure

The nucleus of a Jamna verb is often only a single root, but just as often, it consists of multiple morphemes. A nucleus may contain zero, one, or more true lexical roots, and any number of affixes. These affixes tend to carry classic adverbial meanings.

Nucleic affixes are broadly dividable into prefixes and suffixes. Beyond this, their order is not fixed, but rather reflects syntactic scope: each affix is added to a valid nucleus, creating another in turn. In principle, this process can continue recursively with few limitations, though combinations rarely reach any remarkable length outside of poetry. Nevertheless, it is a powerfully expressive system, as we will see with the following examples:

a) Taa isimitik.
taa i-simi-ti-k (no semantic affixes)
1.excl PAST-roll.down-FMT-FMT
I made it roll down.

b) Taa isimimotik.
taa i-simi-mo-ti-k (one semantic affix)
1.excl PAST-roll.down-by.foot-FMT-FMT
I kicked it such that it rolled down.

c) Taa isimimoiitik.
taa i-simi-mo-ii-ti-k (two semantic affixes)
1.excl PAST-roll.down-by.foot-backwards-FMT-FMT
I kicked it such that it rolled down backwards.

d) Taa isimimoiittik.
taa i-simi-mo-ii-k-ti-k (three semantic affixes)
1.excl PAST-roll.down-by.foot-backwards-abrupt.stop-FMT-FMT
I kicked it such that it rolled down backwards and came to an abrupt stop.

e) Taa isissimimoiittik.
taa i-sig-simi-mo-ii-k-ti-k (four semantic affixes)
1.excl PAST-violently-roll.down-by.foot-backwards-abrupt.stop-FMT-FMT
I kicked violently it such that it rolled down backwards and crashed.

Semantic scope is especially apparent in example e): it is the entire scene expressed by the nucleus in d) that is now presented as happening violently, not just the kicking and not just the abrupt stop.

Nucleic suffixes often govern a particular verb formant, while the prefixes never do so, and the one infix does not either. When there are multiple suffixes, the last one's formant preference wins.

bipartite verbs

Some Jamna verbs appear to have no roots: their nuclei consist entirely of semantic affixes. We cannot treat these affixes as being roots unto themselves, because they are obligatorily dependent - that is, a nuclear suffix must have material to its left, and a prefix must have material to its right. The only catch is that a prefix can satisfy this dependence of a suffix, and vice versa, such that the pair constitutes a valid nucleus. No rootless verbs are possible without at least one semantic prefix and at least one semantic suffix, though many have more.

For example, there is no verb root meaning "to dance". Instead this verb nucleus is rootless, consisting wholly of a prefix and two suffixes.

Taa homoaragai.
taa ho-mo-ara-gai
1.excl playfully-by.foot-repetitive-FMT
I am dancing.

syntax

Looking at the big picture of Jamna syntax, the language is principally right-branching (or "head-initial"), while heavily employing the movement of elements to the immediate left of phrasal heads for various purposes. In the following discussions, we will use the term extraction exclusively for this operation. Extraction can be seen repeatedly in the sections below, and even in basic constituent order.

notation conventions

The most important transformations in Jamna Kopiai will be provided among the sections below, in the following Zompistian manner.

A → B means "whenever A is a legal structure, so is B".

Symbols used:
S = subject
V = verb
O = direct object
I = indirect object
N = noun
VP = verb phrase
NP = noun phrase
PR = pronoun or aioari
H = aioari
PP = prepositional phrase
Adv = adverbial information
Adj = adjectival information
Det = determiner
SEN = sentence
TOP = topic
AGR = agreement (of whatever sort)
- = morpheme boundary (for instance, V-k means "verb with the -k suffix")

Brackets attached to a symbol specify further information: for example, S[N-i] means a subject, specifying that this must be a noun with the nominative -i suffix, while TOP[S] specifies a topic that is the subject of a sentence, and PR[coref] a co-referential pronoun or aioari.

Parentheses, which are not attached to symbols, indicate optional elements - e.g. S V (O) means a clause with or without a direct object.

major phrase types

noun phrases

A noun phrase follows the order (Det)-N-(Comp)-(Adj)-(PP) - that is, first an optional determiner, then the noun, then its complement if there is one, and finally any adjuncts: adjectives or prepositional phrases, in that order.

haooi
field-NOM
a/the field

Niakai haooi jami pao ha siisi
NAME field-NOM wheat-NOM big LOC river-NOM
Niakai's big wheat field by the river

noun class

There are two classes, A and I, named for their broadly corresponding to "animate" and "inanimate" semantic categories. Despite the tendency of A-class and I-class nouns to be associated with animate and inanimate referents, Jamna noun class is fundamentally a grammatical property of each word, not a semantic property of its referent. For instance, an A-class noun can be co-referential with an I-class noun - and indeed, all pronouns are A-class, even when their antecedents are I-class, a fact much exploited in locution. Compounded and derived nouns typically inherit the class of the root (or head root) on which they are based, regardless of animacy.

There are numerous morphosyntactic differences between A-class and I-class nouns, which are detailed in the appropriate sections. To summarize the most important differences:

• A-class nouns (pronouns, etc) take no nominative suffix, while I-class words take the nominative suffix -i.
• A-class nouns may optionally take the dual suffix (in the nominative case). This is not possible for I-class nouns.
• A-class direct objects and indirect objects can be extracted from VPs, and A-class possessors from NPs, none of which is possible with I-class nouns (though no such limitation holds for topic extraction). The preverbal position permits only A-class direct objects.

I-class nouns are unable to participate in syntax as fully as A-class nouns do. There are various work-arounds, but they all depend on the fact that haaioari are A-class; for example, though an I-class noun cannot be extracted as a determiner (see below), it can be topicalized, and a co-referential aioari is left in its place - which can then be extracted.

case usage

The three morphological noun cases correspond directly with the grammatical relations of subject, direct object, and indirect object. These usually take the nominative, accusative, and dative cases respectively.

Other usage:

• Objects of prepositions are always nominative.
• "Second objects", for example in causatives, also take the accusative case.
• When what would normally be a direct object is demoted to indirect object, the dative case is stacked onto the "original" accusative case, rather than simply replacing it.
• Predicate nominals take the nominative case.
• Topicalized nominals take the nominative case regardless of their grammatical relationship in the comment.
• Subjects of "experiential dative" constructions are in the dative case.
• NP elements extracted as determiners agree in case with their head nouns; otherwise they are always nominative.

possession

The least-marked indication of possession is a simple prepositional phrase with ia, "of".

jingioa ia nagahoi
handle-ACC of basket-NOM
handle of the basket

I-class possessors can be indicated in no other way, but A-class ones are typically extracted as determiners (next section).

determination

Another instance of extraction: N → Det N

Jamna Kopiai does not have articles as such, though the haaioari ai and o sometimes serve simply to mark definiteness. In this language, "determiner" is not a word class but rather a syntactic position, to which several types of elements can be placed or fronted: chiefly haaioari, adjectives, and A-class possessors. Filling this slot makes the noun definite - that is, it has a specific referent identifiable by the listener.

If the determiner position is empty, the noun is not necessarily indefinite, but rather simply unmarked for either status. Except for indefinite quantifiers, there are no direct markers for indefinite material. (Free English translations in such instances tend to use "the" or "a" as needed in English, but this does not reflect a behavior of Jamna.)

In general, it is the element most salient to identifying the noun that is fronted. This results in a common order of priority: an aioari usually takes the prenominal position, if one is present; if not, A-class possessors are typically the next most pragmatically salient, followed by adjectives (including numbers). This hierarchy is not a strict rule, merely an observed tendency.

Determiners agree with the head noun for case.

with haaioari and pronouns: N → Det[PR]-AGR N

ai haooi
this-NOM field-NOM
this field

taai mainoa
1.excl-ACC mother-ACC
my mother

with nominal possessors: N ia NP[A-class] → Det[NP[A-class]]-AGR N

Kopiai ia Jamna
language-NOM of Jamna-NOM
language of the Jamna
→
Jamna Kopiai
Jamna-NOM language-NOM
the Jamna language

with adjectives: N Adj → Det[Adj]-AGR N

haooa kinjo
field-ACC green-NOM
green field
→
kinjoa haooa
green-ACC field-ACC
the green field

Determining possessors can be whole noun phrases, which can result in recursive possession:

taai mainoa nagahoa
1.excl-ACC mother-ACC basket-ACC
my mother's basket

pronouns

Personal pronouns distinguish "clusivity" (whether "we" includes "you"), but not grammatical number as such. Thus the 1.excl pronoun taa can mean either "I" or "we (but not you)", while the 1.incl pronoun aago means "I and you (and maybe others)". However, note that the number (singular or plural) of an A-class subject is invariably marked on the verb.

pronominal usage

The use of pronouns is socially restricted:

• In the third person, there are no personal pronouns except for haaioari. Any third-person referent can be indicated with haaioari, but human referents are typically indicated by their title.
• The second-person pronoun is restricted to those one is on a comfortable, familiar basis with: family (in the same generation or lower), friends, lovers, and peers. In most other situations, people are addressed by their title.
• The first-person pronouns are unrestricted, except that royals and certain other august personages may refer to themselves by title as well.

There are several general-purpose titles in use, which may be on the path to becoming directly pronominal. The most important of these are:

• iaar, "sir/ma'am"
• gaih, "wife"
• ikka, "merchant"
• kafari, "teacher"
• praan, "(your/his) majesty, highness"

numbers and quantifiers

Both numeric and non-numeric quantifiers fall into the category of adjectives, and behave like any other in regard to noun determination, case, and all other behaviors within a noun phrase. Quantifiers cannot independently serve as pronouns in Jamna Kopiai and usually appear only as modifiers to nominal elements; independent uses are non-referential without at least a token noun. The most basic non-numeric quantifiers are as follows.

• rai - no, none
• niia - few, a minority of
• ta - some but not all
• inna - half-ish
• haaii - many but countable
• okro - most
• gii - uncountably many
• janar - every, all

prepositional phrases

Jamna Kopiai is very poor in basic prepositions: there are only two. These are:

• ni - comitative, instrumental, "with respect to"
• sam - "from", "by", "without", "sans"

All other prepositions are verbs that have this as an additional function. There are several dozen verbs of relationship and motion that are used as prepositions, but this is not a productive process - except for playful language use, not just any verb can be made into a preposition, only a particular set.

But even though "preposition" as a word class is poorly instantiated in Jamna, prepositional phrase (PP) as a syntactic constituent is not. These are similar to infinitive phrases, but differ from them in their inability to take even the transitive -k formant, and the inability of the prepositional object to inflect for non-nominative case.

PPs with basic prepositions

Taa kiaagai ni ai tafi.
1.excl-NOM cut-FMT with this knife-NOM
I'm cutting with this knife.

Ja sam gaihi najoaiag ni kokki.
man-NOM without wife-NOM IMPF-be-PTCPL with hunger-NOM
A man without a wife is always hungry.

PPs with verbal prepositions

Taa ijooik ooa ha raaipai.
1.excl-NOM PAST-see-FMT that-ACC LOC village.NOM
I saw that at the village.

Taa ha tihi piini.
1.excl-NOM be.at over house
I am atop the house.

Taa joa ikka tapoak tafoa-tafoa.
1.excl-NOM be merchant-NOM sell-FMT (REDUP)-knife-ACC
I, as a merchant, am selling knives.

This last example is especially interesting, for it displays an important property of prepositions that are copular verbs - that sentences constructed this way are structurally ambiguous. This one could just as well be a complex sentence with two verb phrases, "am a merchant" and "am selling knives". However, note how close in meaning the English translations are for the two structures: "I, as a merchant, am selling knives" and "I am a merchant and I'm selling knives". In Jamna Kopiai there is no semantic difference at all. This shows that PPs and VPs are not always morphsyntactically distinct, at least for the copular verbs, and sometimes with other naked verbs used as prepositions.

adverbs

Adverbial adjuncts are most neutrally located at the end of the clause, although many one-word adverbs instead prefer to be placed at the beginning of the verb phrase, after the subject - particularly adverbs of time, direction, and location, as well as adverbs pertaining to the speech act rather than the verb. This is not extraction, just an alternative location for many adverbs.

Taa kiai tittik koaioa ha nahai.
1.excl-NOM time-THIS rise-CAUS ball-ACC LOC sky-NOM
Now I'm lifting the ball in the air.

argument structure

basic constituent order

Simple Jamna sentences are SV (subject-verb), while the position of the O (direct object) exhibits a split based mainly on noun class. SVO is the default word order with I-class direct objects, and most basic overall - while A-class direct objects are normally extracted from the verb phrase, resulting in SOV. Though this SOV is a syntactic transformation, it is nevertheless the default word order when the direct object is A-class. For more on noun class, see below, under "noun phrases".

I-class:

Taa tittik koaioa.
1.excl-NOM rise-FMT ball-ACC
I'm lifting the ball.

A-class:

Taa jigoa tittik.
1.excl-NOM eye-ACC rise-FMT
I'm lifting the eyes. (i.e. looking up)

intransitive:

Taa tihiso.
1.excl-NOM rise-FMT
I'm rising.

grammatical relations

Jamna Kopiai is sensitive to three grammatical relations - subject, direct object, and indirect object - which correspond with the three morphological noun cases.

subjects

Subject is a mandatory category in Jamna Kopiai. Subject pronouns cannot be omitted from any clause, except for syntactically important gaps in coordinated and subordinate clauses. There also exist no "impersonal" constructions - even weather verbs require a meaningful subject.

Nahai mioinja.
sky.NOM rain.FMT
It's raining.

The subject is always the first element in any main clause; no other clause elements can be placed before it, except by dislocation (see topicalization below), or in some subclauses, and except for conjunctions and complementizers.

direct objects

Direct objects normally indicate entities that are acted upon, or receive the action indicated by the verb. The presence of a direct object (in the accusative case) is directly tied to the verbal formant: if the verb has the transitive formant, -k, there is normally a direct object. If the verb has an intransitive formant, there is never a direct object. Some naked verbs do take direct objects as well.

Taa maina jooik nagahoa.
1.excl-NOM mother-NOM see-FMT basket-ACC
My mother sees the basket.

The use of the accusative case -oa is mandatory for any direct object, but the presence of the object itself is not (see below).

indirect objects

Indirect objects most often indicate recipients or benefactees: participants to whom or some other thing is sent, given, etc., or for whom something is done. Indirect objects are not tied to verb morphology, and one can be included in most sentences whether they also have a direct object or not.

Ikka tapoak nagahoa-nagahoa taan.
merchant-NOM sell-FMT (REDUP)-basket-ACC 1.excl-DAT
The merchant sells baskets to me.

Very frequently, an adversative meaning is intended instead.

Nahai mioinja taan.
sky.NOM rain.FMT 1.excl-DAT
It's raining on me. (ugh, I'm getting wet!)

voice and valence adjustment

verbal formants

The formant suffixes that appear on most Jamna verbs form a somewhat unusual system. The formants are mostly not a direct manifestation of specific grammatical operations, per se. Instead, each verb nucleus prefers one or two particular formants, which can then be changed to manipulate the verb's argument structure. Each formant correlates with one or more grammatical argument structures, and to bring such a structure into play on a verb that doesn't normally have it, the verb is re-assigned to an appropriate formant. The "structures" indicated by the various intransitive formants often differ principally by the semantic role of the subject, and sometimes for the indirect object as well.

This introductory explanation should serve as an adequate starting point, but it must be noted that the system is messy: there is substantial overlap between the structures covered by each formant, and many individual verb-formant combinations (hereafter called stems) have undergone independent semantic drift - resulting in their having semantic derivational properties. Additionally, many verb nuclei are rarely or never found with certain formants. Due to these facts, each stem must be cited independently in the dictionary.

Following is a list of the usual behaviors of verbs with their preferred and and other formants.

-gai

as a preferred formant

Verbs in -gai are intransitive, and show a mixed bag of subject types: agents, patients, experiencers, themes, sources, forces, instruments, and agent+patient (i.e. "reflexive"). This does not mean such verbs can normally change their argument structures without changing formants! Each -gai verb lexically specifies one or more roles for the subject. But this is still the most catch-all of the intransitive formants, the most frequently occurring after -k, and perhaps the most neutral. Indirect objects on these verbs can be of any kind, but are most often benefactees or anti-benefactees (adversatives).

as an alternative

When verbs are changed to the -gai formant from some other, the effect depends on their preferred formant. -k verbs usually take this formant when demoting or dropping a direct object, and their subjects typically remain agents, but there is high lexical variation in what the detransitivized verb means. For -ti-k verbs that indicate actions or events, -gai is a dispreferred intransitive form, but it is the normal intransitive of stative or derived -ti-k verbs. When other intransitive verbs take this formant, it is frequently coupled with a benefactive or adversative indirect object.

-so

as a preferred formant

Verbs in -so can have almost any sort of subject, as with -gai, but there is a clear tendency toward reflexive or middle-type argument structures - for example, the verbs of personal hygiene all prefer -so. These verbs are otherwise much like -gai verbs, with less diversity of the subject's semantic role.

as an alternative

The major difference between -gai and -so becomes clearer in their use as alternative formants, which are most readily describable with traditional voice terminology: as alternatives, -gai is consistently "active" and -so is consistently "reflexive". Verbs preferring either transitive formant can be changed to -so to make them reflexive. This achieves classic reflexivity, not any sort of middle-like voice - the subject is portrayed as simultaneously agent and patient, performing an action upon itself (whereas middles portray it as undergoing a process).

Notably, any verb with -so (preferred or otherwise) can optionally take an accusative A-class argument in the preverbal position normally taken by A-class direct objects; but we hesitate to call these direct objects, as their purpose is to indicate reciprocal action. The accusative argument has the same semantic status as the subject: each acts upon the other.

-r

as a preferred formant

The third intransitive formant is -r, and verbs preferring it are much more homogenous than with the first two. -r verbs have two major trait in common: their subjects are always patients or experiencers, and their indirect objects - if any are present - are always agents (or agent-like participants such as causes or instruments).

as an alternative

Verbs changed to this formant from a transitive one can straightforwardly be described as passivized: the patient of the transitive action is now in the subject role, while the agent is absent from the clause, but can optionally be re-added in the form of an indirect object.

Verbs changed to this formant from another intransitive one are usually verbs that take only agent-like subjects (including instruments, causes, and so forth) in their preferred formant, and replace this with a patient-like subject when in -r.

-nja

as a preferred formant

Much like -r verbs, verbs preferring the final intransitive formant -nja take subjects that are experiencers or patients. These two verb groups are not highly distinct from each other, except in one regard: any indirect object present is a normal indirect object (recipient, benefactee, or similar) and not a re-added agent.

as an alternative

As between the first two formants, the principal difference between -r and -nja becomes clearest in their alternative uses, which again can be described with voice terminology: -r is "passive" and -nja is "middle". Verbs changed to -nja take patient or experiencer subjects that are portrayed as undergoing a process as opposed to those changed to -r which are portrayed as receiving an action.

An exception to this is when an indirect object needs to be specified that is not of the type each formant takes: for example, passivizing a transitive verb when you don't need to re-add the action's agent but you do need to specify a recipient or benefactee, will be done with -nja instead of -r. The reverse situation occurs too.

-nja is the preferred formant for detransitivizing -ti-k verbs. A change between these two formants is best understood as removing the cause from the sentence. This is a frequent operation with bivalent -ti-k verbs, but less so with trivalent ones, as that leaves an argument with nowhere to go: verbs with the -nja formant (preferred or otherwise) never have direct objects. When such an operation is necessary, the erstwhile direct object can either be demoted to an indirect one or omitted entirely.

-k

Though the position of the transitive formant is a more complex matter than for the others (occuring in S3 rather than S1), it is the most straightforward in terms of what it does.

as a preferred formant

Verbs in -k are those which normally take direct objects. Their subjects can be of any agentive-type role and their indirect objects can be of any sort.

as an alternative

Verbs changed to -k from some other formant retain their other properties, but add a direct object.

Note that direct objects after -k are omissible (see below, under other adjustments). In such cases the suffix becomes loosely referential; this may be considered a direct equivalent to English clauses where the direct object is "it" or "something".

-ti-k

The messiest of all formants is the causative, -ti-k. Except in special cases, -ti does not stand alone, but requires the transitive formant -k in S3. Verbs preferring this formant and those changed to it from another all work the same way, as follows:

1. The subject is the cause (or causer) - the participant who instigates the expressed situation.
2. The preverbal direct object, if present, is the causee - he or she whom the subject causes to perform the action of the verb. This argument is required to be A-class.
3. The postverbal direct object, if present, is generally a patient - that on which the causee acts - or of any related role (anything that could be a direct object otherwise).
4. The indirect object, if there is any, is a normal IDO: recipient, benefactee, etc.

That makes four possible core arguments of a causative clause. This is the only formant that permits a greater valence than three. For this reason, in certain circumstances -ti-k may be employed as an applicative. Both "direct objects" require the accusative case, as normal.

Frequently, -ti-k is employed as a verbalizing derivation from nouns or adjectives, resulting in a "make X" meaning (for example kinjo "green" -> kinjotik "make green"). In this usage there is no preverbal direct object. This must be distinguished from indirect causation with an action verb, in which there is also no causee present in the clause.

other adjustments

postposing (de-extraction)

O V → V O

As previously noted, the constituent order of transitive clauses varies by the class and willing participation of the direct object. There are three basic possibilities:

1. If the direct object is I-class, it always follows the verb.
2. If it is A-class, it normally immediately precedes the verb, taking the preverbal position.
3. If it is A-class, but does not refer to a living thing, OR if it is an unknowing or involuntary participant, this may optionally be emphasized by postposing the direct object to the postverbal position (where I-class direct objects go).

Though common, this postposing operation is highly marked: it treats the direct object as being acted on like one would act on a sack of potatoes or a rock, rather than an animate and participatory entity. For this reason, postposing second-person references can be very rude, and may be employed as a calculated insult or a marker of relative social status.

Postposing gives rise to several implicit subdistinctions. For instance "I him love" normally implies that the love is returned, while "I love him" implies that it is not. "I him see" is a simple default way to indicate this situation, but postposing to "I see him" implies that he may not want to have been seen - perhaps he was caught doing something bad, or simply hiding. Acts of harm nearly always take postposed direct objects.

object omission

V-k O → V-k

A transitive clause with no explicitly given direct object implies that it is irrelevent to the speaker's communicative purpose, or that it is already established well enough by context that it need not be mentioned. Essentially, the transitive -k on the verb constitutes a third-person reference to a direct object, whenever no other is present. (The suffix itself derives historically from a third-person pronoun otherwise lost in the language.)

Taa maina jooik.
1.excl-NOM mother-NOM see-FMT
My mother sees it.

Only in this regard may Jamna Kopiai be considered a "pro-drop" language; the same pattern does not hold for subjects.

object demotion

V-k O[N-oa] → V I[N-oan]

To downplay the centrality or relevence of a direct object, without dropping it entirely, it is also possible to downgrade its status to that of indirect object. Notably, it retains the accusative -oa while so doing; the dative -an is appended to the accusative form rather than the root. Demoted objects are rarely if ever IDO-extracted (as the two operations serve poorly-compatible purposes), but this is not strictly ungrammatical.

Of course -k cannot simply be dropped; it must be replaced by an intransitive formant appropriate for the verb in question. However, if the verb is causative, this is one of the rare circumstances in which -ti can stand alone: V-ti-k O[N-oa] → V-ti I[N-oan]

IDO extraction

S V (O[I-class]) I[A-class] → S I[A-class] V (O[I-class])

Frequently, the indirect object is a voluntary A-class participant with high pragmatic salience while the direct object, if present, is neither. This circumstance is indicating by extracting the indirect object, which may be considered a weak form of morphosyntactic promotion.

IDO extraction moves an indirect object to the preverbal position. This is possible only if there is not already a direct object in that position; however, it is possible to use both extract and postpose at the same time: O V I -> I V O (that is, O and I switch places, not roles).

experiential datives

S VP → S[experiencer]-an VP

It is never wrong to use the nominative case for the subject of a clause. But subjects of clauses expressing a physical or emotional experience (like cold, happiness, or illness) are often given the dative case, if its semantic role is experiencer. This can have the function of expressing nonvolitionality - in particular, it can convert the semantic role of the subject to experiencer in a clause that otherwise would appear to indicate some other semantic role for it. However, some uses of the construction do not fit this description, and may simply be idiomatic.

Taan joa naataro.
1.excl-DAT be too-hot
I'm too hot.

The experiential dative is typically limited to imperfective verb forms, but can be extended to others for rhetorical effect. The great majority of experiential datives involve verbs with the -nja or -r formants, or else copular verbs.

topic-comment structures

SEN → TOP, SEN

Topicalization in Jamna Kopiai is not restricted to nominal elements; almost any sort of element may be made a sentence topic, including entire clauses - even subclauses. Topics may extracted from the sentence (comment) that follows, or may not be represented in it at all: a topic is not required to have any grammatical connection to its comment, only a contextual connection.

Topics employ no special particle, only an intonation break that we will represent with the comma. Only one topic is permitted at a time (but multiple comments may follow: see complex sentences below).

extracted subjects and indirect objects

A subject or indirect object, or a prepositional object, extracted from a sentence and topicalized must be represented by an appropriately cased pronoun or aioari left in the original position.

• S VP → TOP[S], PR[coref] VP
• S V (O) I → TOP[I], S V (O) PR[coref]

Of course, all the usual operations may be performed in the comment, including other extractions of the pronoun. For instance, IDO extraction can look like this: S V I → TOP[I], S V PR[coref] → TOP[I], S PR[coref] V

extracted direct objects

An extracted direct object topic requires no such in situ pronoun.

• S V O or S O V → TOP[O], S V

Extracted nominals are restricted to arguments of the verb and prepositional objects.

other topic extractions

• S VP PP → TOP[PP], S VP H
• S VP Adv → TOP[Adv], S VP H

disjoined topics

A disjoined topic is one that was not extracted from the following comment, and bears no grammatical relationship to it other than the topic-comment structure itself. This pertains to the topical constituent as a whole; elements within it may still be represented in the comment.

Panani, sijai maaso.
dinner-NOM | rice-NOM cook-FMT
"As for dinner, rice is cooking."

copulas

There are three copular verbs in Jamna Kopiai: joa, ha, and ia, which can be paraphrased "to be", "to be at", and "to be of". The first is used for predicate nominals (and adjectives):

Taa joa kafari.
1.excl-NOM be teacher-NOM
I am a teacher.

The second is used for predicate locatives (where the predicate is a prepositional phrase or equivalent infinitive phrase):

Taa ha piini.
1.excl-NOM be.at house-NOM
I am at home.

Taa ha tihi piini.
1.excl-NOM be.at over house
I am atop the house.

The third is used for predicate possessives:

Piini ia taa.
house-NOM be.of 1.excl-NOM
I have a house.

All three are naked verbs: they do not take verbal formants, but otherwise may take the entire range of verbal morphology. However, on copular verbs, most morphology (except for P1 and number agreement) becomes optional - it is typically abandoned whenever a bare copula is good enough.

The copulas are among the verbs that serve double duty as prepositions.

TAM usage

the TC

Below we list the most usual uses and functions of each of the TAM Complex morphemes, individually and in their combinations.

structure

Combinations are structured and restricted. The system can be understood best by treating certain of the morphemes as prefixed auxiliary verbs, and the remainder as morphology that they take. Let us define, then, a category VE: verbal element, which includes both main-verb stems and all six of the auxiliary-like elements: na, so, kok, ago, fi, and haat. From this definition, describing the other three affixes becomes approachable:

1. the general past marker, i-, is prefixed to the first VE in the word.
2. the relative past marker, inni-, is also prefixed to the first VE in the word (mutually exclusive with i-; these form a paradigm)
3. the plural agreement marker, -a, is suffixed to the first VE in the word. If that is the main verb root, the suffix appears in S2.

Two VEs also govern morphology of subsequent ones:

1. the perfect auxiliary requires a supporting i- prefix on the next VE, independently from whether the verb as a whole is past-perfective.
2. the imperfective auxiliary requires a supporting -iag in S2, which in this usage does not result in a participle - the verb remains finite.

as single prefixes

Ø - unmarked - continuous aspect
The continuous is unspecified for tense, beyond that the action cannot (yet) have achieved consequences. Dozens of examples throughout this document.

i - general past (gloss PAST)
By itself, this prefix simply indicates actions that occurred prior to the time of speech, and generally treats them as completed or perfective events. Taa ikrinoso. - "I died."

inni - relative past or anterior (gloss ANT)
The relative past has a similar meaning to the general past, except that it is non-deictic: it indicates actions that happened prior to a time specified in discourse, rather than prior to the time of speech. Taa innikrinok. - "I had killed it."

ago - assertive (gloss ASRT)
This is used when countering mistaken expectations, or for making any sort of lightly assertive statement. It's very frequently found in answers to questions. But in the interrogative, the effect is rather to request such an assertion from the listener; in effect it says "answer me!". Taa agokamar. - "I do smell bad."

fi - abilitative (gloss ABL)
This prefix means, simply, "able to X"; although the notion of epistemic possibility is not conflated with that of ability in Jamna Kopiai, the Jamna abilitative is nevertheless irrealis. It can also be employed for instances that are more properly a matter of permission than capability. Taa figoagai. - "I am able to begin."

haat - energetic (gloss ENG)
This makes a forceful assertion. Indeed, so forceful that its use is typically rude, bordering on profane. This affix is almost completely absent from calm and neutral speech, and is a good indicator that the speaker is upset or otherwise highly emotionally involved in his speech act. Notably, it sometimes appears on adjectives as well. It is clearly related to the verb haata, "to burn". Taa hakkamar. - "I smell fucking terrible!"

kok - intentive or future (gloss INT)
The "future" mainly indicates intentions, and sometimes expectations, on the part of the subject. There is a cultural constraint against treating the non-immediate future as something it is possible to make declarative statements about; any and all description of the future is seen as intrinsically counterfactual, and as such, the affix is strongly irrealis. Taa koggoagai. - "I intend/expect to begin."

na - imperfective (gloss IMP)
On its own, usually a habitual or generic meaning is intended with this prefix: it is generally found in statements of what is customarily done or what is normally or always true. Taa nakamariag. - "I always stink."

so - perfect (gloss PRF)
The fundamental notion of the perfect is the time of consequences: where past actions simply and baldly assert that something happened, regardless of when its consequences were felt, perfect actions are those in which the consequences (of the described situation) are involved in - or chiefly relevant to - the situational context of the time of speech. As such, it is not totally unakin to the perfect aspects found in many other languages - however, note that the Jamna perfect does not actually specify the past occurrance of the situation; this may be typically assumed, but in combination with time adverbials, it could just as well be a present action, or even a future one. What matters is that its consequences are manifest at the time of speech. Taa soigoagai. - "I have begun."

The given list of VEs is not strictly exhaustive for colloquial Jamna. The standard language is essentially limited to these (perhaps even omitting haat), but a moderately productive slang process is to bring other verb roots into the TC, often in clipped form. This usually achieves some sociolinguistic, rhetorical, or "attitude-marking" purpose, like that seen in the energetic. As slang, these constructions are in constant flux.

in combinations

The past and relative past are both combinable with all of the other elements on the list (VEs), resulting in transparent meanings all but one of them: the perfect.

i-so - "past perfect"
The "past perfect" does not give a pluperfect type of meaning; that function is covered by the relative past already. Essentially, the past does not serve to shift the deictic center of the perfect, because this is always anchored to the time of speech. Instead, it merely reinforces that the action/situation happened before the present time. As such, this combination is not frequently employed. Taa isoigoagai. - "I have begun."

inni-so - "anterior perfect"
The anterior perfect on the other hand, though uncommon, has an interesting effect: it explicitly separates the time of an action or situation from the time of its consequences. An anterior perfect action is one that happened before some other, but its consequences were delayed or irrelevent until now. The use of the anterior perfect is mainly limited to experiential statements (those in which the perfect marks possession of relevant experience), and serves to relate the time of gaining that experience to something else. Taa innisoijooik. - the best we can do for a free English translation might be "I have seen it and this was prior to (whatever)." or "I had seen it, thus I have seen it."

The remaining combinations we'll examine are those in which there are two VE prefixes; eight such occur, but only two require usage explanation.

so-na - perfect imperfective
Rather than a habitual or generic meaning, na instead gives a continuous aspect when combined with so. This is much like the unmarked continuous, except that the continuing action/situation is portrayed as having already achieved consequences relevent to the time of speech. Taa soinajooik. - "I've been seeing it."

kok-na - intentive imperfective (reduces to konna)
Instead of "intend to be Xing", this combination merely specifies that the future action/situation is an expectation rather than an intention. Taa konnajooiiak. - "I expect to see it."

The following have meanings we find transparent and self-explanatory:

• so-ago - perfect and assertive
• so-haat - perfect and energetic
• kok-so - intentive and perfect (reduces to kosso)
• kok-ago - intentive and assertive (reduces to kokko)
• kok-haat - intentive and energetic (reduces to kokaat)
• fi-ago - abilitative and assertive

modal prefixes (P2)

The P2 modal prefixes work as follows. They are wholly distinct from the morphological structure of the TC, and behave as simple prefixes.

jo - deontic (gloss DEO)
Jamna Kopiai has only one deontic mood; no subtypes are distinguished, not even an imperative. The deontic mood presents the content of the statement variously as desirable, obligatory, commanded, exhorted, etc. Sii johaatak. - "You should burn it." or "Burn it!"

him - hypothetical (gloss HYPO)
The hypothetical mood expresses situations that might happen or might be true. In this use there is no speaker judgement of likelihood or non-likelihood, rather a bare assertion of possibility. The hypothetical also marks the protasis of most conditional sentences. Sii hihhaatak. - "You might be burning it." or "If you are burning it ..."

goa - inceptive (gloss INC)
Strictly speaking this is an aspect rather than a modality. The inceptive marks situations that are, or were, beginning. It is transparently related to the verb goa, "to begin". Taa goahaatak. - "I'm starting to burn it."

jasooi - opinionated (gloss OPN)
The opinionated mood indicates that the statement is an expression of opinion about what is or what should be, and can be paraphrased "I think" or "in my opinion". Sii jasooikamar. - "I think you smell bad."

semantics

phonosemantics

Jamna Kopiai has a relatively robust phonosemantic system, with a number of pervasive phonesthemes - associations of sounds with meanings found among many otherwise-unrelated words. Phonesthemes are sub-morphemic: similar to morphemes, except indivisible from the roots in which they are found. If a phonestheme is removed from a root, what remains is not a separate morpheme, but rather meaningless.

pair phonesthemes

In this document we will focus on one interesting type of sound symbolism to be found in Jamna, the pair phonestheme.

Many Jamna phonesthemes are discontinuous - that is, many of them consist of a consonant pair, generally the first two consonants in the root, which are separated by one or more vowels in any sort of sequence. Or no vowel at all, for the few pairs where this is phonologically possible. A few "pair"-type phonesthemes are single consonants instead; these form weaker associations of sound with meaning, as there are many other words beginning with the same consonants too, but their historical origin as pair phonesthemes is clear and remains significant. There are also other variations on the consonant-pair theme.

To illustrate, following are ten roots with the phonestheme s-r, which can be summarized as "things like or having to do with the fingers". Srako may be derived from sra, but the rest are not known to be related words.

• sora - finger
• sori - bone
• soroi - gesture, signal
• sirroi - knot
• sair - thread
• srako - to poke
• srafana - to tickle
• sra - to touch
• saroati - to twist
• siriso - to point

A case can be made for most consonant combinations possible in the language being pair phonesthemes to at least some limited extent - depending on the standards chosen for counting them, there may be over a hundred. Only some of the most frequently appearing are listed here.

• s-r - the fingers
• f- - the mouth, and speech
• t-r - heat and fire
• kr-k - cold and ice
• j-p - sharp or angular things
• g-k - loud noises
• h-s - quietness
• m-s - flood or excess
• j- - shiny things and light
• k-k - fast or abrupt motion
• h-i - slender things
• k-f - knowledge and the divine
• h-o - space, openness, extent
• s-f - pale or white things
• ni-n - small or cute things (-ni is also a diminutive suffix)
• p- - big, dumb, or hollow things (sometimes pronounced [b], a phone not otherwise found in the language, to increase the phonosemantic force)

Certain "pair" phonaesthemes are actually triplets, or have triplet sub-phonaesthemes. For example, many more words with the k-f pair have r as their third consonants than sound frequencies alone would suggest. It has been suggested that these, and pair phonaesthemes in general, may be relics of a triliteral root system present in a very ancient stage of the language.

other sound symbolism

Particularly in nouns and adjectives, there are additional symbolic significances of most sounds, especially in final consonants. Root-final consonants are found in a relatively low proportion of Jamna words overall, but those which do occur very frequently have some symbolic force, and affective words are much more likely to have them. We will not go into these in detail at this time, as these phonesthemes follow quite typical cross-linguistic sound symbolism trends (such as the bouba/kiki effect) - for example, /i/ is associated with small or endearing things, plosives with stopped motion, /p/ with popping and bursting, and so forth.

some semantic domains

motion verbs

Motion verbs in Jamna Kopiai lexically encode both manner of motion and direction of motion. As a result there are a great number of them.

Up to four direction categories are lexically distinguished: forwards or basic, up, down, and sideways or curved. For some verb sets one or more of these is absent, for example there are three verbs for "to swim", indicating forward, downward, and upward swimming, but none for sideways. To illustrate one verb set, we will list the types of meanings encompassed by one of the most basic, which simply means "to go".

• hoo - to go forward; proceed; move as expected
• tihi - to go up; ascend; rise; lift
• goraa - to go down; descend; fall; sink
• tigo - to go sideways; go around; detour

A table of selected verb roots follows. Each set encompasses a range of meanings broadly parallel to those of the hoo set.

	forward	up	down	sideways
go	hoo	tihi	goraa	tigo
walk, run (2 legs)	roi	oaji	hoogi	hiho
walk, run, crawl (4+ legs)	moga	pioo	aiga	otii
jump	noa	ioo	roo	noi
scoot, sneak	hiiona	tana	higa	asoo
swim	kita	pakro	kitara	--
fly, float	kata	riko	poma	taifai
flow	maoi	--	maoi	apa
squirt; dart	misra	mosai	missoi	ioga
roll	iami	soomi	simi	hami
turn, spin	gii	ioo	tigra	homo
climb	jaaga	jaaio	riono	--

On top of this system, nuclear affixes on the verb root add further distinctions of manner, means, and path of motion, among numerous other things.

numbers

The numeric system of Jamna Kopiai is complex. Due to a history of scholarly tinkering, it now uses a mixed system in which bases of 3, 10, and 30 are variously employed.

The first ten numbers are original, native number words, untransparent except that "five" and "ten" are etymologically related to a now-archaic term tianna for "thumb", and the dual of this word. Eleven through thirty are represented by names of old gods; however, there are additional numerals for each multiple of three, based on the thirty roots (original ten plus gods), to which the number one be added or subtracted - as it can to the main roots. In Jamna culture, objects are typically counted in groups of three, and have to be if they are in the range of 31-91 (91 is thirty threes plus one). Numbers beyond this range are rarely employed in counting, only calculation, and are represented with a directly base-30 system.

Somewhat interestingly, the language also employs the collective form "gods" (haaii) as a non-numeric quantifier meaning "roughly in the ten to thirty range" or, more loosely, "many but easily countable".

English	root term	counted by threes
1	oi
2	ato
3	pai	paioi
4	koi
5	tian
6	aman	paiato
7	soota
8	hai
9	faaji	paippai
10	tinan
11	tikassa
12	hanikoki	paikkoi
13	fiijoata
14	ooma
15	tafotir	paittian
16	siminaokan
17	gihooaga
18	joopoo	paiaman
19	hangima
20	miso
21	moiinika	paissoota
22	kropopao
23	rihag
24	jiisoso	paihhai
25	oniaia
26	panikopia
27	sinja	paiffaaji
28	porros
29	sirangahirij
30	nama	paittinan
31		paittinan ni oi or nama ni oi
32		paittikassa sam oi
33		paittikassa
34		paittikassa ni oi
...		...
91		painnama ni oi