| Cəssın |
[ ˈʈɤs.sɯn ]
|Period||c. 1100 YP|
|Spoken in||Isthmus south coast|
|Total speakers||0.8 million|
|Writing system||adapted Fáralo script|
|Classification|| Edastean |
|Basic word order||SOV|
|Morphology|| agglutinating |
w/ some fusion
Cəssın [ˈʈɤs.sɯn], or Cəssın löŋive [ˈʈɤs.sɯn løˌŋi.ve] is an Edastean language spoken along the south coast of the isthmus which connects the Siixtaguna subcontinent with the main portion of northern Peilaš. The language is a direct descendant of Fáralo, and thus a member of the northeastern subbranch of Edastean. Its closest relatives (named in the order of increasing geographical and linguistic distance) are Woltu Falla, Namɨdu, and Puoni, as well as a number of smaller and lesser-known Fáralo varieties.
The Cəssın-speaking area is centered on the city of Cəssa (Fáralo Čisse); it encompasses roughly the northeastern quarter of classical Huyfárah. The region is not, however, linguistically homogenous: While Cəssın is the majority language overall, quite a number of predominantly Doroh- and Feråjin-speaking villages are found in the countryside, and the city itself is home to a sizeable community of Affanons. The languages of all these people have been highly influential in the linguistic development of Cəssın; some of the characteristic features that separate it from other Fáralo languages are retroflex obstruents, vowel harmony, and a fairly thorough shift to head-final word order that has given rise to postpositions, suffixed locative cases, and a mostly agglutinative verbal complex in which earlier auxiliaries are suffixed to the root verb. The major isoglosses for all these features are found roughly 200 miles west of Cəssa; for instance, the dialect of Ultumosu (F. Oltumosou) lacks vowel harmony but is otherwise very similar to the variety described here, whereas the speech of Sacak (F. Sertek, local Šêrtak) is much closer to the language of Ussor and the Oltu valley.
- 1 Phonology
- 2 Morphology
- 2.1 Nominal morphology
- 2.2 Verbal morphology
- 3 Derivation
- 4 Syntax
- 5 Samples
- 6 Lexicon
- 7 See also
|plosive||p · b||t̪ · d̪||ʈ · ɖ||c · ɟ||k · ɡ|
|liquid||ɾ · l||ɻ · ʎ|
- /p b v m/ are written p b v m.
- /t̪ d̪ s n ɾ l/ are written t d s n r l.
- /ɲ ɻ ʎ j/ are written ny ry ly y.
- /k ɡ ŋ x/ are written k g ŋ h.
- The retroflex obstruents /ʈ ɖ ʂ/ only occur adjacent to back vowels, while their palatal counterparts /c ɟ ɕ/ usually appear adjacent to front vowels. However, the two series are contrastive in word-initial position before a å o u, and sometimes medially adjacent to at least one back vowel. The practical orthography does not normally distinguish between them, using c z ş for both /ʈ ɖ ʂ/ and /c ɟ ɕ/. Where it is necessary to make a distinction, palatals may be unambiguously indicated by using cy zy ç instead.
The Cəssın vowel inventory is:
- /i e æ ø y/ are transcribed i e ä ö ü.
- /ɯ ɤ ɑ o u/ are represented by ı ə a o u.
- /ɒ/ is written å. It occurs only in tonic and pretonic syllables.
- c z ş n l ry vary between [ʈ ɖ ʂ ɳ ɭ ɻ] in words with back harmony and [c ɟ ɕ n l ɹʲ] in words with front harmony. In casual speech the plosives c z are often slightly affricated, so that any of [ʈʂ ɖʐ], [tʃ dʒ], [tɕ dʑ], or [cç ɟʝ] may actually be heard.
- The non-alternating palatals cy zy ç are always "bright" in sound, i.e. [tɕ dʑ ɕ] or [c ɟ ç]. Among the nobility, [kʲ ɡʲ] may be found for the plosives.
- s ş become voiced in word-initial and intervocalic position and when adjacent to a liquid consonant.
- g lenites to [ɣ] or even [ɰ] between vowels. Some speakers extend this to word-initial position and/or to clusters with a liquid.
- h is pronounced [ç] in coda position after front vowels, and [x] in coda position after back vowels. Word-initially, most speakers have [x] generally, but [ç] before front vowels is not unheard of. Between vowels the most common realisation is [h]; however, among lower-class urban speakers h often merges with g intervocalically, being realised as [j] or [ɰ] depending on the frontness of the preceding vowel.
- The mid vowels e ö ə o vary in height or tenseness for most speakers depending on whether they stand in an open or closed syllable. In open syllables they are usually quite close to the cardinal values [e ø ɤ o]; in closed syllables they are pronounced [ɪ ʏ ʊ˒ ʊ] in rural dialects, and [ɛ œ ʌ ɔ] in urban dialects.
- In rural dialects the peripheral low vowels ä å are often raised to [ɛ ɔ].
- In urban dialects the low unrounded vowels ä a fall together as [a] or [ɑ] in open syllables and [ɜ] or [ʌ] in closed syllables. For the same speakers, å may become an unrounded [ɑ].
- The unrounded back vowels ı ə are often not fully back [ɯ ɤ], but central [ɨ ə]. This is most frequently found in the western dialect areas.
A prominent feature of Cəssın is vowel harmony, whereby vowels in the same word will agree with each other in terms of frontness and/or roundedness. On closer inspection, the language actually has two distinct harmony systems:
Word stem harmony
Lexical word stems in Cəssın may be constructed with either of the four overlapping sets of vowels listed below:
- front vowels only: i e ä ö ü.
- back vowels only: ı ə a å o u.
- unrounded vowels only: i e ä a ə ı.
- rounded vowels only: ü ö å o u.
Another way to phrase this would be that no lexical word is allowed to contain both front unrounded vowels i e ä and back rounded vowels å o u, and no word is allowed to contain both front rounded vowels ü ö and back unrounded vowels a ə ı either.
All native lexical words conform to this harmony system, and borrowings are usually adapted to fit. Compound lexemes, however, are not always resolved and may thus feature mixed harmony. Due to the nature of the historical processes by which vowel harmony came into force, it is not always predictable which harmony set a given lexeme belongs to. What is more, several of those ambiguous stems actually occur in doublets that differ in both harmony and semantics. In order to deal with this, harmony values for all entries are given in the lexicon.
Prefixes must agree with the harmony group of the stem they are added to according to the system described above. As the total number of prefixes is limited, and many of them also undergo certain consonant sandhi processes, the alternations will be covered for each individual prefix where it is introduced.
Grammatical and derivational suffixes are also subject to vowel harmony. However, the rules for suffixes (which make up the major part of Cəssın morphology) are slightly different from those for stems and prefixes. In particular, the rounding opposition is underlyingly neutralized for post-tonic vowels; suffixes thus harmonize with the stem in frontness only. The five harmonizing vowels are designated in their underlying forms with the cardinal vowel letters i e a o u, which give the following sets of surface vowels:
In mixed-harmony compounds, suffixes harmonize with the last morpheme in the stem.
Sound changes from Fáralo
Cəssın nouns inflect for case, number, deixis, and possession. Aside from the oblique cases, nominal morphology is exclusively prefixing.
The morphological template for nouns in Cəssın looks as follows:
| primary case & number
nominative sg. +, pl. -k-
accusative sg. -, pl. -s-
|ROOT|| secondary case|
1st person ac-/ce-
2nd person aŋ-/oŋ-
3rd person ag-/ge-
Primary case and number
The two "primary cases", nominative and accusative, are morphologically identical in the singular for most nouns. However, some words exhibit mutation of their initial consonant in the prefixed nominative singular forms; see below for details. The plural forms of both cases use prefixes; underlyingly k- in the nominative and s- in the accusative.
There are five levels of deixis in Cəssın, which are marked with prefixes placed immediately before the case/number markers described above.
- The unmarked absolute form is used for nominal predicates, for nouns immediately preceded by a number or quantifier, and in partitive or appositive constructions.
- The indefinite refers to a specific but unspecified instance of the noun in question, similar to the indefinite article in other languages. It is marked with ro- before consonant-initial stems and with or- before vowel-initial stems.
- The definite refers to an instance of the noun in question that has already been mentioned. It is marked with a prefix whose underlying form is lo-. Proper names are usually considered definite in their unmarked form.
- The proximate and obviative are mostly used to introduce new instances of the noun, although they can both also be used anaphorically. The proximate form marks its referent as spatially or emotionally close to the speaker, or indicates that it is an expected participant of the action or predication. In contrast, the obviative form marks the referent as further away, unfamiliar, or unexpected. Both the proximate and obviative can function as a topicalizer. The proximate is marked with å- before consonants and v- before vowels. The obviative is marked with se- before consonants and ş- before vowels.
When not appearing immediately next to a full noun phrase referring to their possessor (which would be cast in the genitive-benefactive case), possessed nouns do not take deictic prefixes, but are instead obligatorily marked with possessive prefixes that agree with the person of the possessor. The number of the possessor is not indicated by these prefixes; number marking on possessed nouns refers to the number of the possessum.
- 1st person possession is marked with underlying ac-, which surfaces as ce- in the plural before stem-initial consonants.
- 2nd person possession is marked with aŋ- in the singular, and with oŋ- in the plural.
- 3rd person possession is marked with ag-, or ge- in the plural of consonant-initial nouns
Combined nominal prefixes
The basic forms of the combined deixis/possession/case/number prefixes are shown in the table below. Many of these prefixes have different forms when used with consonant- or vowel-initial stems respectively.
|1st person possessive||aş+||aş-||cek-||ces-||ac-||ac-||aşk-||ass-|
|2nd person possessive||aŋ+||aŋ-||oŋk-||ŋos-||aŋ-||aŋ-||oŋŋ-||oŋgr-|
|3rd person possessive||ag+||ag-||gek-||ges-||ag-||ag-||agg-||agr-|
- The prefix-final k of the nominative plural becomes g before underlying voiced obstruents and non-palatal liquids, and surfaces as gemination of a following nasal or ly. It also causes devoicing of a following g, which results in a geminate kk, and lenition of following t c cy to s ş ç. Stem-initial p is reduced to zero with compensatory rounding on a subsequent unrounded vowel.
- The prefix-final s of the accusative plural lenites to r before nasals and v, and it assimilates completely into following liquids, forming a geminate. Before palatal and retroflex obstruents it undergoes backing to ş, and with stem-initial y it coalesces into ry.
- The prefix-final g in the 3rd person possessive singular undergoes palatalisation to y when it ends up in intervocalic position between two front vowels, and coalesces with stem-initial ly y into lly y. Before consonant clusters, an epenthetic i or ı (depending on the frontness of the vowel in the next syllable) is inserted where phonotactically necessary.
- Stem-initial b d z zy g coalesce with preceding ŋ into a geminate nasal at the POA of the plosive, giving mm nn nny nny ŋŋ.
- Stem-initial b d regularly become v r after g and vowels, and p t after ş.
- Stem-initial z zy become ş after k, ry after vowels, and coalesce with preceding g into geminate zz zzy.
- In indefinite singular nouns, the prefix-final r coalesces with stem-initial h s ş into t ts c, and surfaces as gemination on coronal plosives.
- Nouns beginning in an unrounded vowel undergo assimilatory rounding of that vowel after the indefinite plural prefix.
- All nominative singular prefixes cause mutation of stem-initial p t k h (indicated by +):
- In the indefinite, stem-initial p t k change to v r g, and h coalesces with the r of the prefix into ts.
- In the definite, proximate, and obviative, stem-initial p t k h change into v r g s.
- In the 1p possessive, h becomes ş, but p t k do not mutate.
- In the 2p possessive, h becomes s, and p t k coalesce with the ŋ of the prefix into the geminated nasals mm nn ŋŋ.
- In the 3p possessive, p t k h become v r g s.
- Instances of h that derive from Fáralo *f mutate differently in the indefinite and 3p possessive, where they become v instead.
In addition to the above, all noun prefixes are subject to vowel harmony, assimilating to the word stem in frontness, roundedness, and/or height. The results are not easily summarized into a coherent system, but the number of nominal prefixes is small enough that all individual rules can be listed:
- The prothetic i in the absolute accusative plural becomes u if the next vowel is rounded, and ı if the next vowel is back unrounded.
- Any vowel designated as o in the above table becomes ö if the next vowel is front. In the definite singular, the vowel in the prefix is raised to u ü if the stem begins with a nasal or palatal consonant.
- The å in the proximate does not exhibit any alternations.
- The e in the obviative becomes ö if the next vowel is rounded. Before a nasal or a palatal consonant, the vowel in the prefix is raised to i or ü.
- The e in the 1st person possessive plural also becomes ö if the next vowel is rounded, but does not undergo raising. Also, it surfaces as ə if the next vowel is back unrounded.
- The e in the 3rd person possessive plural becomes o if the next vowel is rounded. Note also that the prefix-initial consonant g undergoes palatalisation if the next vowel is front unrounded, giving ye-.
- Any vowel designated as a in a possessive prefix may behave differently depending on whether the first syllable of the word stem etymologically had a cardinal vowel *a *i *u or a non-cardinal vowel *æ *ɛ *ɛi *ɔ *ɔu. In the former case (which is treated as an irregularity), the prefix vowel surfaces as o or ö if the next vowel is back rounded, and as e otherwise. In the latter case, the vowel in the prefix becomes ä before front vowels, å before back rounded vowels, and a before back unrounded vowels.
Cəssın has six oblique or secondary cases, which are marked with suffixes that attach to the accusative stem in the singular, and to the nominative stem in the plural.
- The genitive-benefactive is marked with the suffix -(a)m. It can indicate a wide range of relationships; among the most common are possession, qualitative attribution, and association with a beneficiary.
- The comitative is marked with the suffix -(u)n. It marks a noun which is associated with the head noun in a non-directional or reciprocal relationship, for instance accompaniment or circumstance.
- The instrumental is marked with the suffix -(i)ta. It is used to signify the instrument, manner, or method by which an action is performed, and to mark inanimate agents of passive sentences.
- The locative is marked with the suffix -na. It is used to specify the location of a noun, and in conjunction with most non-dynamic prepositions. With temporal nouns it signifies "while" or "during".
- The lative is marked with the suffix -sa. It marks the goal of motion or the recipient in a dative clause. With temporal nouns it signifies "before".
- The ablative is marked with the suffix -(o)ŋ. It marks origin or source of motion, and animate agents of a passive construction. With temporal nouns it signifies "after" or "since".
The vowel-harmonic variants of the case suffixes are as follows:
- All suffix-initial vowels are dropped after another vowel.
- Stems whose last consonant is a single intervocalic liquid or nasal usually delete their final vowel in the locative case.
- Stem-final m n ŋ often (but not always) change to v r g with the addition of a vowel-initial suffix.
- Instances of stem-final m n ŋ that alternate with v r g in other cases combine with the lative suffix into -psa -tsa -ksa. Regular stem-final r also gives -tsa.
- Non-alternating stem-final nasals surface as n in the instrumental, causing the suffix-initial vowel to be deleted.
- Stem-final obstruents and liquids also cause deletion of the suffix-initial vowel in the instrumental. c cy lenite to ş before the instrumental suffix; p t k coalesce with it into -tta. When the last consonant in the stem is a coronal plosive or affricate, this contraction also happens if a stem-final unstressed vowel will be deleted, even where the plosive or affricate is preceded by a liquid, nasal, or fricative (e.g. bünte 'navigation' → bünttä 'by navigation').
- Stem-final c cy s ş ç h l combine with the lative ending into -ssa. This may also trigger deletion of a following vowel, but not after l or consonant clusters (mosu 'harbor' → mossa 'to the harbor', but iştälö 'kitchen' → iştälösä 'to the kitchen').
- Stem-final sibilants lenite to r before the locative ending. All other obstruents and nasals assimilate completely, forming a geminate nn.
Due to the powerful pronominal agreement system on verbs, free-standing nominative and accusative pronouns are not used very frequently. They mostly serve to emphasize a participant, to name the subject of a predicative clause or the direct object of ditransitive verbs, or to clarify reference in situations where verbal agreement is ambiguous.
- Unlike full nouns, pronouns form their instrumental, locative, lative, and ablative cases with prefixes instead of suffixes. In fact, the pronominal variants are etymologically older; they result from phonological fusion of Fáralo preposition + pronoun combinations. The nominal case suffixes were subsequently formed via topicalisation of nouns in an oblique role, with a following resumptive 3sg pronoun in the appropriate case eventually agglutinating to its referent.
- In the 2pl, a k has been added in several forms by analogy with the other plural pronouns.
- Conversely, the inherited /k/ in the 2sg pronouns (< lewku-) has been dropped where it ended up in word-final position, making those forms more obviously singular.
|this one||these||that one||those|
- If required, a distinct plural can be formed in the oblique cases of the distal demonstratives by suffixing -k, but this is nonstandard and occurs mostly in legal documents and scholarly writing.
|who, what||which one|
There are also four more general interrogative pronouns, which do not inflect for case:
- mala 'where'
- holo 'when'
- örge 'why'
- eggä 'how'
- me 'no, none'
- näm 'some, any'
- mas 'a few'
- uvı 'many, much'
- aşan 'very much, abundant; in excess'
- eŋ 'all, every'
When these quantifiers immediately precede a noun in the absolute state, they cliticize to it, undergoing vowel harmony and some limited consonant sandhi.
Complex numbers can be built with the conjunction å, which is left out after another vowel: 458 is boyam å duro hor.
Ordinal numbers are formed with the adjectival suffix -(i)n, which causes vowel alternations in most words for numbers lower than 10, and lenites the final consonant of multiples of 100. Only the last word in a complex ordinal number takes the suffixed form.
Under the influence of neighbouring SOV languages such as Doroh, Affanonic, and Lotoka, Cəssın mostly dropped the basic conjugation of Fáralo (which was used with VSO word order), extending the use of the auxiliary conjugation instead (which followed AuxSOV order). Over time, five major developments occurred:
- Constructions with more than one auxiliary became increasingly common; any additional auxes were placed after the main verb at the end of the clause.
- The lexical verb (together with any added uninflected auxiliaries) came to be seen as a kind of adverbial modifier to the inflected aux. With this reanalysis, it became possible for the main verb to be topic-fronted like other adverbs, ending up immediately before the auxiliary.
- Pronominal clitics came to be increasingly used even when overt lexical arguments where present, and eventually turned into polypersonal agreement suffixes.
- The inflected auxiliary was moved to the end of the clause in order to create a basic SOV sentence layout, carrying any pronominal clitics with it. At this stage, main verbs were almost always directly followed by an auxiliary, whether in sentence-initial or sentence-final position. In the end this development resulted in a unitary verbal complex with the structure [root]-[auxes]-[inflection].
- When vowel harmony became productive, the verbal complex as a whole was subject to harmonic assimilation rules.
Synchronically, it seems most useful to treat the Cəssın verb in terms of mostly agglutinative template morphology. The surface forms are sometimes sensitive to prosody, and the set of agreement morphemes that evolved from Fáralo's pronominal clitic is fusional in nature. The negative and irrealis affixes pose some difficulties for templatic analysis (firstly, they do not have a fixed slot but always occur immediately before the last former auxiliary, and secondly they can also appear preceding the root); however, a template-based approach still offers the most easily understandable description and shall therefore be used in this grammar sketch.
The basic structure of the Cəssın verb can be summarized as follows:
| truth value
telic passive -im-
atelic passive -oll-
direct causative -su-
indirect causative -tiş-
non-past sg. -Ø-, pl. -k-
past sg. -(e)n-, pl. -i-
transitive: subj > obj
ditransitive: subj >> dat
The first slot of the Cəssın verb is normally used to mark the epistemic truth value of the action. The realis is unmarked; the irrealis is marked with a prefix that is underlyingly og-, and the negative is marked with the prefix m(a)-. It should be noted that the irrealis and the negative are conceptualized on the same scale by Cəssın speakers: the irrealis signifies an event whose truth is doubted or unknown, and the negative signifies an event about which the speaker is certain that it did not/will not happen.
- The negative prefix reduces to m- before a vowel.
- The prefix-final g in the irrealis surfaces as gemination on a following nasal or ly, as k before voiceless consonants (triggering lenition of following t c cy to s ş ç), and as y before front vowels.
- Stem-initial p is reduced to zero in the irrealis, with compensatory rounding on a subsequent unrounded vowel.
- Stem-initial b d lenite to v r in the irrealis, and stem-initial z zy coalesce with the prefix-final g into geminate zz zzy.
- Stem-initial prevocalic b d z zy lenite to v r ry ry in the negative.
As with nominal prefixes, vowels in these verbal prefixes are affected by the quality of the first vowel in the stem.
- The o in the irrealis becomes ö when the next vowel is front. Since the prefix-final g undergoes palatalisation, the surface form of the morpheme in this environment is öy-.
- The a in the negative becomes ä if the next vowel is front, and å if the next vowel is back rounded.
When any overt inflection for voice, aspect, mood, or emphasis is present, the truth value of the action is not expressed word-initially, but within the auxiliary complex. The morphemes used there derive from the same etymological forms as the prefixes described here, but several of their surface forms are quite fusional, and so they will be considered separately.
The auxiliary complex
The middle portion of the Cəssın verb consists of former auxiliaries, which have evolved into an agglutinative complex marking voice, aspect, mood, and polarity. In terms of template morphology, the auxiliary complex must be considered to have at least five distinct slots, one of which is variable in position. In practice, it is rare for more than two non-variable slots to be filled at all, and many of the options available in one slot are incompatible with some of the options in another one.
For ease of understanding, the various auxiliary slots and their possible combinations will be treated one by one in this description, with the irrealis and negative inflections, which are always prefixed to the last overt auxiliary morpheme only, shown on each auxiliary to account for morphological variation.
Another thing to take into account is prosodically-conditioned variation in some auxiliaries, which results from elision of unstressed vowels in certain environments.
The first auxiliary slot may contain one of two markers for the passive voice, which differ in their associated aspectual value. The active voice is formally unmarked.
- The telic passive is used to describe perfective or inchoative actions from the point of view of the patient or theme. It is semantically similar to English passive constructions with the auxiliaries 'get' or 'become'. It is marked with the morpheme -im- (from Fáralo eim 'turns'); the morpheme-final m assimilates in POA to a following obstruent or nasal.
- The atelic passive has a habitual or stative connotation, similar to English passives with 'be'. It is marked with the morpheme -oll- (< ezul 'takes'). Before a consonant, the suffix has a short final l.
|back harmony||front harmony|
The second auxiliary slot can be occupied by one of seven aspect markers. All of them can appear in the active voice; however, only the progressive and egressive aspects may co-occur with the atelic passive, and only the inchoative, progressive, egressive, and iterative aspects may co-occur with the telic passive.
- The aorist, which is used to simply refer to an action without specifying its internal structure, is formally unmarked.
- The habitual, which indicates that an action or state occurs regularly and/or persistently, is marked with the morpheme -ar(e)- (from Fáralo æde 'stays').
- The prospective, which indicates that an action is about to occur, is marked with the morpheme -nun- (< noun 'goes'). This reduces to -nn- when it ends up between two vowels.
- The inchoative, which indicates the beginning of an action or state, is marked with the morpheme -(h)epp(e)- (< hæpabe 'starts'). The morpheme-initial -h- is dropped after consonants, turning instances of m n ny ny ŋ that derive from earlier *b *d *ɖ *ɟ *ɡ into p t c cy k.
- The progressive, which indicates that an action is underway or that a state is currently valid, is marked with the morpheme -ur(o)- (< odu 'comes'). When followed by a consonant cluster that begins with a nasal, the morpheme-final vowel is raised to u.
- The egressive, which indicates the end of an action or state, is marked with the morpheme -(e)ta- (< etá 'stop' and ta 'just did'). The suffix-initial vowel is only present after consonant clusters or geminates.
- The retrospective, which indicates that an action has occurred, resulting in a currently relevant state, is marked with the morpheme -pel(a)- (< pila 'has done').
- The iterative, which indicates that an action occurs several times, is marked with the morpheme -kam(a)- (< kuəma 'repeats').
|back harmony||front harmony|
The third auxiliary slot indicates mood. Most of these markers may co-occur with all possible voices and aspects; however, the only common combinations of mood and aspect are obligative or desiderative mood with inchoative, egressive, or iterative aspect. Of note are the two causatives, which share a morphological slot with the other moods even though they are arguably more voice-like derivational suffixes themselves. Both causatives normally increase the valency of the verb by promoting the causer to subject position.
- The indicative is unmarked.
- The potential, which indicates that an action or state is possible, permitted, or likely, is marked with the morpheme -ba- (from Fáralo boa 'can'). After a vowel or underlying *ɖ *ɟ *ɡ *k, this suffix lenites to -va-; after s ş ç it devoices to -pa-.
- The obligative, which indicates that an action or state is necessary, compulsory, or promised, is marked with the morpheme -(e)s(o)- (< is 'should' and so 'must'). Either vowel is only present when phonologically necessary.
- The desiderative, which indicates that an action or state is intended, hoped for, or recommended, is marked with the morpheme -al- (< al 'wants').
- The deductive, which indicates that the occurrance of an action or state has been deduced from context, is marked with the morpheme -err- in the present tense and -ez- in the past tense (< idər 'seems'). Before any consonant, and always in the intensive, this suffix simplifies to -er(e)-.
- The direct causative, which indicates that the action is the result of direct causation, is marked with the morpheme -su- (< soy 'causes').
- The indirect causative, which indicates that the action is the result of indirect causation, is marked with the morpheme -tiş- (< teiš 'bends').
|back harmony||front harmony|
The fourth and last positional auxiliary slot can contain only one morpheme, the intensive marker -şş- (from the Fáralo null auxiliary siš 'do'), which reduces to -ş- when adjacent to a consonant, except that after morpheme-final nasals or liquids an epenthetic i ı is inserted. If no epenthesis occurs, any e ə in the preceding syllable is raised to i ı, except when part of the negative affix.
|back harmony||front harmony|
Tense and number
The morphological markers for tense and number in Cəssın are fusional. The nonpast singular is unmarked, the nonpast plural takes the suffix -k-, the past singular takes the suffix -(e)n-, and the past plural takes the suffix -i-. These markers are partly fused with the pronominal agreement suffixes; the combined forms are therefore given in the following section.
Cəssın verbs exhibit agreement with up to two core participants of the action. In intransitive verbs, only the subject is cross-referenced on the verb. Monotransitive verbs agree with both the subject and the direct object, while ditransitive and causative verbs agree with the subject and the indirect object. Ditransitive suffixes of same-person subject and object are also used with a number of semantically monotransitive verbs to indicate a reflexive action.
It should be noted that the set of agreement suffixes is somewhat defective. For instance, number is not distinguished for indirect objects at all, nor for first person direct objects when the subject is plural; the latter set of affixes is also used for plural subject > 3sg direct object. What is more, the non-past tense does not distinguish between transitive first and third person subjects. Some of the other suffixes show a degree of syncretism as well; for instance, the singular subject ditransitive affixes in the non-past tense are mostly the same as the plural subject ditransitive forms in the past tense.
Because of a certain amount of fusion at the morpheme boundary, pronominal agreement is best considered together with tense marking. The combined tense/person suffixes are given below; note that all vowels are underlying vowel-harmonic morphophonemes that alternate between i e ä ö ü in front-harmony words and ı ə a o u in back-harmony words.
|subject||no object||direct object||indirect object|
|2sg||-ro||-ri, -(o)rve||-roŋ||-ris, -(o)rves||-rus||-ras, -(o)rgas||-rim, -(o)rvim||-rum||-ram, -(o)rvam|
- Stem-final vowels are dropped with the addition of a suffix that begins with a non-bracketed vowel. The preceding consonant usually undergoes palatalization if the dropped stem-final vowel was one of i ı e ə.
- The bracketed -g- in the 1sg>1pl and 3sg>3pl suffixes appears only after a vowel or liquid. After front vowels it palatalizes to -y-.
- Some of the 2sg suffixes have two variants. The forms with -rv- or -rg- appear after stem-final vowels (with e a being replaced by the suffix-initial -(o)-); the forms with single r appear after consonants.
- This r in the 2sg suffixes becomes n after nasals, forming a geminate nn (however, nasals that derive from earlier *b *d *ɟ *ɡ surface in their original form, giving br dd zz gr). After p k it becomes s, and after other voiceless consonants it fortifies to t (with lenition of preceding c cy to ş ç). With l r it usually combines into ld dd; occasionally rr also occurs, the most prominent example of the latter is when it immediately follows the retrospective aspect morpheme.
- After a nasal or liquid, the 1sg>3pl and 1sg>>3 suffixes surface as -zas and -zam respectively.
- In the rare case that a stem-final consonant and a suffix would form an illicit cluster, the vowels given in brackets are inserted at the morpheme boundary.
|subject||no object||direct object||indirect object|
- The suffix-initial vowel in the past singular is dropped after vowel-final stems, with the exception that 2sg forms beginning with -(o)- override stem-final e a.
- In the 3sg (except intransitive, 3sg>3pl, and 3sg>>1), the suffix-initial vowel is also dropped when preceded by a single liquid or nasal.
- Conversely, before the past plural suffixes, all stem-final vowels are deleted.
Cəssın forms predicative clauses with a stand-alone version of auxiliaries that have otherwise become incorporated into the verbal complex. The inflection of these copular auxiliaries is strongly reduced compared to full verbs: they cannot combine with other elements of the auxiliary complex, and they do not exhibit detailed subject agreement either, requiring overt pronouns instead.
In theory, all grammaticalised auxiliaries can function as a copula. In practice, most speakers only use four of them:
- äre, which corresponds to the habitual aspect, denotes a permanent state.
- uro, which corresponds to the progressive aspect, denotes a temporary state that may end soon.
- pelä, which corresponds to the retrospective aspect, denotes a state resulting from a recent change.
- im, which corresponds to the telic passive voice, denotes transition to a new state.
All of these copulae inflect for tense, number, and truth value; in a fairly regular manner. The complete paradigm of all four main copulae is shown below:
|permanent (cop)||temporary (pred)||resultative (res)||dynamic (dyn)|
Most Edastean languages employ zero-derivation quite liberally, and Cəssın is no exception. Verbs and adjectives can freely be used as nouns, and adjectives can freely be used as verbs, both simply by appending the appropriate inflectional morphology. The meaning of these derivations is generally transparent: Nominalized verbs denote an instance of the action described by the verb root. Nominalized root adjectives denote the abstract quality described by the adjective; nominalized derived adjectives denote a person or object carrying this quality. Verbalized adjectives always denote the state of having the associated quality (with targets or beneficiaries optionally referenced as indirect objects).
It is also possible for many nouns to be used as verbs, but this is done less often, and the semantics are not as straightforward. With nouns denoting people or animals, the verbalized form usually describes an action typically done by the referent, and with nouns denoting locations it describes an action typically performed at that place. With nouns denoting inanimate objects, verbalizations may refer (a) to an action performed with the object as an instrument, (b) to an action that affects or produces the object, or rarely (c) to an action that moves something else to a position associated with the object. All of these have in common that they normally produce monotransitive verbs that take a direct object. However, since zero-verbalizations from nominal stems can be quite idiosyncratic, such derivations will receive separate entries in the lexicon.
Another notable phenomenon is that the valency of many verbs can be adjusted without any morphological marking other than the appropriate pronominal agreement. In particular, direct and indirect objects may often be dropped, both with or without adjustment of verbal participant marking. Conversely, most lexically monotransitive verbs can take an optional target or beneficiary in the lative or genitive-benefactive case, which is then cross-referenced by agreement suffixes. However, lexically intransitive verbs (especially semantically stative ones) cannot be converted into transitives quite as easily; a derivational or morphological causative is usually required.
Derivation of new lexical items can also be accomplished morphologically. This is especially prevalent for forming adjectives (more than two thirds of all adjectives in Cəssın carry one of the major adjectival suffixes -(i)n or -ga), and for word-formation within the same part of speech. Derivations which are semantically more specific than those listed under 'zero-derivation' above are also carried out with affixes.
|affix||part of speech||meaning||example|
|-(e)l||noun/adj → noun||diminutive|| gu 'road' → gul 'alleyway'|
nälör 'horse' → nårrəl 'pony'
|-(o)ş||noun/adj → noun||primary instance; augmentative|| leyä 'word' → leyäş 'decree, verdict'|
caŋ 'specific' → cagoş 'unique specimen'
|-(v)aş||adj → noun||object with quality|| iryün 'aesthetically valuable' → iryünäş 'work of art'|
råga 'clever' → rågvaş 'gadget, tool'
|-lo||noun/adj/pp → noun||location|| nivi 'plan, intend' → nivlö 'headquarters'|
muzo 'near, next to' → muzolo 'surroundings'
|-eç||noun/adj/verb → noun||associated process|| hossa 'trade route' → hosseç 'long-distance trade'|
suncə 'kneel' → suncəç 'oath of allegiance'
|-run||noun/adj → noun||person (usually male); male animal|| gibrä 'sheep' → gibrün 'ram'|
cyoto 'benevolent, generous' → cyotorun 'patron, sponsor'
|-sa||noun/adj → noun||female|| cuk 'king' → cuksa 'queen'|
ökö 'blessing' → ökösä 'priestess'
|-vo||verb/adj → noun||person, agent|| roşka 'creative' → roşkavo 'inventor; genius'|
egrän 'fulfill (a task)' → egrämmö 'servant; assistant'
|-o||verb → noun||affected object, result|| şor 'pour' → şoro 'drink' (n.)|
tirnä 'have an affair with' → tirnü 'bastard child'
|-(s)u||noun/adj → verb||causative, factitive|| ının 'holy, sacred' → ınnu 'consecrate'|
dännye 'fortress' → dännyesü 'fortify'
|-ce||verb/adj → verb||reflexive, reciprocal; atelic|| nuno 'remove' → nunzə 'leave, withdraw; abdicate'|
iste 'correct, true' → istece 'correct oneself'
|-se||noun → verb||process of losing/emitting sth.|| åta 'humour, wit' → åtasə 'entertain, amuse; be eloquent'|
mütsi 'purpose' → mütsise 'lose track of'
|-(i)n||noun/verb/pp → adj||adjectivizer|| dul 'fall asleep' → dulın 'exhausted'|
tam 'south' → temin 'southern'
|-ga||noun → adj/adv||adjectivizer, adverbializer|| öyol 'foot, base' → öyolga 'stable; supportive'|
nara 'east' → narga 'eastward'
|m(a)-||adj/noun → no change||negative, opposite|| ittän 'friendly, polite' → mittän 'bold, brazen, rude'|
- All derivational suffixes are subject to vowel harmony.
- Segments in brackets are only present when their omission would cause vowel hiatus or an illegal consonant cluster.
- Suffixes beginning with a consonant often have special allomorphs adjacent to certain consonants. This is mostly lexicalized in older formations; however, the voiced obstruents v b d z g predictably coalesce with preceding non-alternating nasals into the geminates mm mm nn nny ŋŋ, and with so-called "alternating nasals" into the geminates bb bb dd zz gg, even in newly-formed derivations.
- Suffixes beginning with a vowel cause lenition of stem-final alternating nasals m n ŋ to v r g after back vowels, and to v r y after front vowels.
- In many inherited suffixal derivations, a medial vowel is dropped and the resulting cluster is simplified, sometimes accompanied by a shift in vowel harmony (e.g. nälör 'horse' → nårrəl 'pony'). Such alternations do not usually occur in new formations; however, when a vowel-initial suffix is added to a disyllabic stem of the shape (C)(C)VC₁VC₂ where C₁ is a nasal and C₂ is a nasal or plosive, or where at least one of C₁ and C₂ is a liquid, the second vowel of the stem will normally be syncopated by analogy with older words. If the resulting cluster contains two different consonants with the same manner of articulation, regressive assimilation will occur.
| To Be Continued...|
Cedh is still working on this section. The contents are incomplete and likely to undergo changes.
The noun phrase
Basic noun phrase syntax
The canonical order for noun phrase constituents is (loc₁) (gen) (num) (adj) det-N (app) (loc₂) (rel), where loc₁ is a preposed locative phrase, gen is a genitive phrase, num is a quantifier, adj is an adjective or participle, det- is the deixis/possession prefix, N is the head noun, app is an appositional noun phrase, loc₂ is a postposed locative phrase, and rel is a relative clause. Only det-N is required. The slots for locative phrases, adjectives, and appositional elements may each hold multiple items.
Noun phrases headed by a personal pronoun have the simpler structure P (app) (loc₂) (rel). Unlike ordinary noun phrases they may not contain any of the preposed constituents, only appositional noun phrases, postposed locative phrases, and relative clauses.
Possession is indicated by casting the possessor in the genitive-benefactive case and placing it before the possessed noun. If any quantifiers and/or adjectives referring to the possessed item are present (positioned between the possessor and the possessum), the latter's deictic prefix will be replaced by a possessive prefix. This is optional if the possessor immediately precedes. Possessive prefixes, especially in the 1st and 2nd person, may also refer to an implicit possessor that is not overtly specified.
There are three distinct types of locative phrases in Cəssın. (Not all of them are actually locative in meaning, but indicating a position in space and time is the prototypical function of these constructions, at least when used within a noun phrase.)
The simplest variant, traditionally called a cased locative, consists simply of an ordinary noun phrase in the locative, ablative, lative, or instrumental case (listed in decreasing order of frequency), and it is used to indicate relatively basic relational concepts more or less corresponding to the English prepositions 'on, at, in' (loc), 'from' (abl), 'to' (lat), and 'with, by' (instr). Cased locatives are normally placed before their head noun, preceding any quantifiers, adjectives and genitive phrases, but they may also appear in the postposed locative slot after the head, especially in complex noun phrases or with pronouns.
More specific relationships are indicated with adpositional phrases. Many adpositions in Cəssın are ambipositional, i.e. they can appear both before their object (forming a prepositional phrase) or after it (forming a postpositional phrase). When used attributively within a noun phrase, prepositional phrases follow the head noun, and postpositional phrases precede it. The difference between these constructions is mostly stylistic; however, it should be noted that several adpositions display a semantic distinction between prepositional and postpositional usage, and that some adpositions can only be used in one type of adpositional phrase. There is also a register difference: Prepositions are most common in the more conservative higher registers, while an extended use of postpositions is typical for the vernacular of the urban lower classes.
The object of pre- and ambipositional adpositions is cast in the nominative case by default. However, it is often possible to use other cases instead, with a roughly aspect-like shift in semantics: The locative may replace the nominative to indicate a static but less permanent state (similar to a progressive aspect), the lative can be used to indicate the beginning of a relationship (similar to an inchoative aspect) or a movement towards the referent, and the ablative can be used to indicate an ending relationship (similar to an egressive aspect) or a movement away from the referent.
Most of those adpositions which can only be used postpositionally always govern the genitive-benefactive case for their complement.
A list of the most common adpositions follows. Note that some of the ambipositions are slightly different in form when used as a pre- or postposition. The prepositional variants are for the most part direct reflexes of the Fáralo words; postpositional variants etymologically derive from prepositions with a fused 3rd person pronoun as their object. Postpositions without a prepositional counterpart mostly derive from recent grammaticalisation of locative noun phrases; incidentally, this is the only productive adposition-forming process in the language.
- nerre 'inside of'
- with lat: 'into'
- with abl: 'out of'
- herre 'outside of'
- with abl: 'away from'
- muzo 'near, next to'
- taş 'around; throughout'
- taŋ 'behind'
- with abl: 'beyond'
- büvö 'between, among'
- mävin 'without' (only with nom)
- pinnä 'about, with regard to'
- with abl: 'judging from'
- nin 'as, like, similar to'
- mete (prepositional), mıca (postpositional): 'failing' (only with nom)
- giret (prepositional), gıc (postpositional): 'towards, up to' (telic)
- with gen: 'in the direction of' (atelic)
- räve (prepositional), rivlä (postpositional): 'along, through, by way of'
- with gen: 'across'
- vol 'under, below'
- with lat: 'down towards'
- with abl: 'up from'
- yomo 'over, above'
- with lat: 'up towards'
- with abl: 'down from'
- volam 'belonging to' (only with nom)
POSTPOSITIONS (only with gen):
- yolna 'in front of'
- bunna 'in back of'
- hurna 'to the left of'
- ånna 'to the right of'
- yonna 'on top of'
- panna 'at the bottom of'
- tunna 'at the disposal of; in reach of'
Basic clause syntax
|å||'and' (used within syntactic constituents)|
|ova||'and' (used to link clauses and phrases)|
|ållo||(subclause is core argument of verb)|
|gåto||purposive ('so that, in order to')|
|töl||instrumental ('by means of')|
|uŋglo||alternative ('instead of, rather than')|
|ellö||anterior ('before, until')|
|oglo||posterior ('after, since')|
The complementizer ållo is interesting etymologically since it results from a merger of no less than five different constructions with similar sound and meaning, three of which could be used to topicalize an action when preceding a verbal noun, and the other two indicating intentions: wa lu (lit. 'this, the X'), ouwa lu ('and the X'), al lu ('want that X'), we lu ('will X'), and wætə lu ('regarding X'). The regular phonetic outcomes of these would have been *olo, *ålo, *ollo, *ålo, and *åtlo respectively.
The legend of Emperor Tsinakan
- Main article: Tsinakan text
Sep Şancan, Kargaram mlava kån aksacuk, ova loyorolam å lunyamam agımlokat, oşuŋ ivişşeneş:
Ellö epennälin eşperyöm löyöpelnä, eŋ herrin löglinni yam coskıvo urvı. Herrin löglinni oşuŋ ivyec: "Egveryö, rum üvislinni höpespeläŋŋäs, mlava kån cuk urun. Dara zärye üşö naŋ pelä. Ova åminyöl rum epeyürä egveryöm löyöpelnä, åşə həcon uro."
Səşum epeheppin e, loyorolam å lonyamam agımlokat, eşperyöm loyöpelnä, ellö zumannınzas herrin löslinni rum hölkünäryec, lumam Ovasam lögremässä tücünişşin. Kecempelänzäs, ova ŋåtol lumosa cöstun olpəlanzas. Oşuŋ iveniş: "Aştas, lögvüm lotol! Suşın seglinni rum otəcon sekärkim, zıpuroşkəc! Üşö ak ının anzallaşam löskäglö porohəppəşkas! Söröponım üköşşörgäs!"
Kün lumo cesleyä rotoŋŋas. Olını ova essörvä kåmmossunəş. Tåşə, ro glinnä herrin löglinni rum zıpuryəc höpeşşinzäs å giveşşinzäs. Rormocon usbusun isgibrän cyotanzas, ova lölinnisä Kargatsa såppåvıntışşınzas.
The shamaness and the bear
Nol epeheppen loyollogaŋ, lorånassa ornogom loyoko şötepelneş. Lunogo dirin eri, löricöyäm geşin occuk mölö, å cılcın å cyoc ugu urvı. Örän pirgä, gåto pospıntışən öröllö, vogga lunyolom lökäglönä hölküŋŋä lilmeheppen. Lunogo melve ogluryasa gıc olcın ova eŋ suşın lormozo örvöbrörötä tånuttışəŋŋas. Dara kellärişşeneş zyoryuŋ zosıŋŋa lököyosommo.
Zyunıŋŋa lorånassa lonogosa tücüneppen, lunyolom rivlä rum örkiretä tömmöllürün. A porat å sossın erren; nevin rörninyöŋ örnyömmi nin, am löbrinne setin urvı. Lara lunogo kelläreneş akarga, å lut, zärye üşö ayana mereneş. Vogga rolga, lorånassa åksun olınəş ova pupsun lotohunam lopovı seşpeheppişşeneş, ŋötommaş kilin öllyomota.
(This text was written on Feb 6, 2011 as part of Conlang Relay 18. Translated from Roger Mills' Prevli.)
- Main article: Cəssın/Lexicon