The Proto-Mocha language is the hypothetical ancestor of all Mochan languages. It was spoken along the river Aksora (PM *Ak-sōra) around 3000 BAE.
Proto-Mochan was an agglutinating, mainly prefixing, split-ergative SOV language.
Based on the correspondances between Proto-Mocha's daughter languages, an inventory of 31 consonants and 8 vowels can be reconstructed.
The following consonants can be reconstructed for Proto-Mocha:
|Plosive||*p *b||*t *d||*ts *dz||*tʃ||*k *g||*kʷ *gʷ|
For convenience, glottalization will be represented by an underdot. It has been suggested, that */tʃ/ */tʃʼ/ */ʃ/ may have been /tʲ tʲʼ sʲ/. They will be written *č *č̣ *š and their alveolar counterparts *c *c̣, respectively.
The contrast between *kʷ *gʷ *kʷʼ and *k *g *kʼ was neutralised in front of *u, same with *x vs *xʷ and *ŋ vs *ŋʷ.
There is also evidence for the existence of a glottal stop *ʔ which occured in front of every word starting with a vowel.
Eight oral vowels can be reconstructed for Proto-Mocha:
|Mid||*e *ē||*o *ō|
The original quality of the mid vowels *ē *ō is part of a dispute, some suggest they where the long counterparts of *e *o, some others suggest they where short close-mid vowels /e o/, with *e *o being short open-mid vowels /ɛ ɔ/.
There is no evidence for secondary articulation on vowels, neithar can diphtongs be reconstructed for PM.
Generally speaking, the structure can be described as (C)CV(C)(C). The occuring clusters however were limited and most roots were of CVC(C)V shape.
Evidence shows, that syllables that started with a vowel probably had a /ʔ/ inserted in front of them. This becomes apparent in the reduplication process of nouns, where vowel initial words are treated like consonants (see below).
Initial clusters were limited to a few roots and mostly came from prefixes or vowel deletion: *xʷba "bean" from *xub- "seed", kʷsamt- "Kosamt" from kos- "Horn". There was also a small group of native vocabulary of CCV shape: *psa "fish", *sxʷa "knife", tpi "eye". The origin of those clusters remains obscure.
Another group of words with initial clusters are nouns formed with the nominalizing prefix m- : m-štaw "illness", from štaw- "to be sick".
Medial clusters occured very frequently.
Final clusters occured, but were limited to two consonants. There were also no nouns ending with a cluster. If a noun root ended with a cluster, a vowel was added to it.
PM stress was word-final.
PM nouns were marked for case and number. Each nouns also had an inherited gender, or noun class, not marked on the noun itself, but on verbs or adjectives. Nouns were an open class, and many of them were derived through affixes or, less frequently, from compounding.
There was an definitve article, *a-, which always preceded all of the noun's other prefixes.
The structure of PM was was more or less [Article][Posession][Derivation][Reduplication][ROOT][Derivation][Case].
PM distinguished between singular and plural. Most nouns two different stems for their number. Usually, the singular was the bare stem, and the plural form was derived via reduplication. This reduplication process was a bit complicated.
The most simple nouns were the ones of CV or CCV shape. Here, the syllable was simply doubled for the plural form. In case of a cluster, only the first consonant was doubled:
*ma "breast" > *mama "breasts"
*tpi "eye" > *titpi "eyes"
Nouns with two or more syllables are more complex. First, the leftmost syllable is reduplicated, then, the vowel of the second syllable drops out:
*leča "tongue" > **le-leča > *lelča "tongues"
*kōč̣i "male, man" > **kō-koč̣i > *kōkči "men"
This, however, did not apply to words containing a medial cluster:
*šompa "foot, leg" > *šo-šompa "leg", not **šošmpa
Vowel initial nouns always started with a *ʔ and are reduplicated like every other noun:
*ala "mother" > *aʔala > aʔla "mothers"
*iškē "worm, snake" > *iʔiškē "worms, snakes"
In all cases, labialized velars lose their rounding when reduplicated:
*ŋʷɨ "flower" > *ŋɨŋʷɨ "flowers", not **ŋʷɨŋʷɨ
Derivational prefixes, like the nominalizer *m- always precede the root and are not part of the reduplicated stem:
*m-sōre "river" > *m-sōsre "rivers"
It is proposed, that the two directional cases are a very recent development, and that they came from an earlier locative case combined with the directional particles xʷ- and š-, also found in pronouns (*mi-xʷ "that one there", mi-š "that one here") and verbs (*xʷ-ał- "go, go away", *š-ał- "come").
The absolutive had a zero ending, but an empenthetic vowel was added if the root ended in a consonant cluster: *walt-i "hair-ABS".
The plural case ending were directly attached to the reduplicated plural form of the noun: *pati-r "women-DAT", *papti-n-er "women.PL-DAT.PL".
As previously mentioned, PM had a quite rich system of derivation for its nouns. One of the most productive prefixes was the nominalizer *m(e)- forming nouns out of verbs or adjectives:
*cox- "breathe, Breath" > *m-cox-e "Human, Animal", *štaw- "to be sick" > m-štaw "illness", *kɨŋʷ- "high" > *m-kɨŋʷ-o "Mountain"
Compounds of noun + verb, or, less frequently, noun + noun or verb + verb, also occured:
*šuriŋʷi "soldier" < *šura "weapon" + iŋʷ- "to bear", *štawoš "witch, wizard" < *štaw- "sick" + *eš- "to do", *ḳʷapoča "body" < ḳʷab "bone" + *poča "house", *kʷa-čiča "clan leader, authority" < *kʷa "head" + *čiča "maternal grandmother".
PM first and second person pronouns had a free and a clitic form and were distinguished for number, but not for gender or clusivity:
The clitic pronouns were used for possesive means or for emphasis, since person was not marked on verb.