Proto Vauqun-Adzovъd

Summary
Descendant of Proto Csillan. Basically Csilla's equivalent of PIE. Unlike PCS, PČN was actually reconstructed. In certain places I'll use blue writing to highlight etymology information that couldn't be reconstructed yet is known to me 'cuz I made the lang :P

Vowels
As can be inferred from the graph above, most vowel qualities distinguished length and some nasality, while only two have only one form which makes sense when you consider that those two qualities only existed as the product of umlaut that did not affect long vowels (and all nasal vowels are long.)

Phonotactics
Legal Onsets: C, [+fricative][+plosive], Cj, [+obstruent][+fricative -dental -homorganic]

Legal Nuclei: any vowel of any quality or length- though nasal vowels cannot occur before a complex coda (2+ consonants)

Legal Codas: [+consonant -nasal]*, [r, w]C, [+fricative][+plosive],  [+fricative][+fricative] and of course nothing

Legal Middle Clusters: any coda + onset, unless it breaks the following rules: * see exception under morphophonemics
 * fricatives cannot follow nasals
 * the only plosive + plosive combinations allowed are geminates and those beginning with a velar, otherwise the initial plosive spirantizes: /p t b d/ > /f þ v đ/
 * chains of three or more obstruents are illegal

Any combination of obstruents must agree in voice: if a cluster is created that does not agree in voice, the second obstruent determines the voice for the whole cluster

Morphophonemics
Vowels underwent many shifts under certain conditions, with varying regularity: * in these dialects /æ/ is not affected by the umlaut
 * Vowels raise before r, w, n and m in closed syllables (-r also makes the vowel long if it isn't) /y e æ a o/ > /i y e æ u/ -highly regular in short vowels, less common in long vowels (particularly in declension & conjugation when one of these sounds is made to follow a long vowel). Also note u + w > ū.
 * Vowels of any length nasalize before n and m, after raising /æ e y i u/ > /ą ę ę į ų/, the only exceptions that can be found appear word-finally. Obviously this shift did occur when the nasal was in the word final position, as evidenced by words like sją (again) or the second declension dat/acc singular ending -tų, which likely had -m or -n after them in some form of PPČ. However, when there was pressure from other forms retaining the -m/n, the nasalization could fail to occur (or, perhaps, be undone), such as the noun kost en (sword) gen sing köst eni, nom pl kosten am , gen pl kösten æmi
 * Vowels lower in any other closed syllable /i y e æ u/ > /y e æ a o/, same regularity as first rule
 * Short vowels become long before the consonant  at the end of a word
 * Umlaut triggered by a long vowel (except ȳ). Fairly regular; does not occur, however, when the vowel was elongated due to nasalization at a morpheme boundary, ie fimæþ dat/acc sing fimetų, not fĭmĕtų, but gen/loc pl fĭmĕdō. The umlaut affects all short vowels preceding the long vowels (prefixes, though, were frequently excluded from this rule.)
 * Fronting umlaut (before ī, ē, į or ę): /a æ o u y/ > /æ e ö ü i/
 * Rounding umlaut (before ō, ū or ų): /a æ e i y/ > /o ö ö ü ü/
 * Open umlaut (before ā or ą- in some dialects ǣ*): /æ e i y u/ > /a æ e e o/
 * NB: /ö ü o/ are sometimes written /ĕ ĭ å~ů/ to make umlaut caused by certain declensions or conjugations more transparent, ie åtkåmū [otkomu:], 1st sing atkata, also đůzætā [ðozæta:] gen sing of đuzæþ


 * The vowel  seems to behave somewhat strangely at the end of a word: it is never long, as indicated by both languages that preserve the length distinction as well as those that replace it with a mobile accent system, yet it appears to always trigger an umlaut, no matter what. For example, the word for "house" is đejöþi-note the umlauted <ö>, not the regular .

There were also a few grammaticalized consonant shifts which are discussed in the grammar section.

First Declension
The first declension, also known as the thematic declension, contains entirely animate nouns ending in any short vowel except æ, u, ü or ö. It shows a great deal of variation depending on the root vowel, as well as a consonant shift to further mark the oblique plural cases. The consonant shifts are as follows: These shifts happen to the consonant right before the root vowel and they occur in every plural case except the nominative.
 * p t k b d g > f þ h v đ ɣ
 * f v s z > þ đ þ đ
 * V(n m) > [nasal vowel] + s
 * r > ɣ
 * [sonorant/hiatus] > [sonorant/hiatus] + s

I'll add some examples of declined nouns when I feel like it :þ

There are also some nouns belonging to this with a nominative singular ending in -o, -e or a consonant. These decline for the most part normally (nouns ending in a consonant act like an -e stem in all cases other than the nominative singular and plural, which are both null) except instead of a consonant shift the plural is marked in all cases (including the nominative) by the infix -am- (the  disappears when the stem ends with an -o or -e.) Thus the plural of gven (light) is gvenam and tjeko (boar) is tjekom.

Second Declension
The second declension, also known as the athematic declension, contains entirely inanimate nouns ending with the consonant -þ. The -þ can follow any vowel, but it is overwhelmingly tends to follow æ in the nom sing and e in all other cases (the alternation being the product of regular sound rules) ; this is so ubiquitous because the inanimate gender originally derives from the abstract/collective suffix -æþ, other vowels in the endings are the result of borrowing.

The t/d alternation is an unexplained aberration (it comes from an ancient consonant gradation that was only preserved in this suffix). Most Čeuɲ languages eliminate it in some way or another.

Third Declension
The third declension, also known as the H declension, contains both animate and inanimate nouns (the difference is based on semantics) and ends in ā, ō, or ȳ followed by -h. Nouns ending in -ȳh alternate the -ȳ with -i due to regular sound change, and those ending with -ōh alternate with either -o or -u, since both o and u become ō before -h word-finally. Nouns in -āh simply alternate with -a.

* The first ending is for animate nouns and the second for inanimate.

Articles
The only article reconstructable for Proto Čeuɲ is the definite article- or perhaps two definite articles, one for the animate gender and one for the inanimate.

The article seemed to have been able to be cliticized and prefixed to the noun it modifies, reduced to s-, h(t)-, or f-. This seemed to cause consonant mutations depending on the initial consonant of the noun, but the individual dialect groups do not seem to agree, therefore the language must not have settled on a stable system. The article could also be used as a personal pronoun (it was likely not cliticized in this usage.)

Adjectives
The inan ins pl suffix is -ĕþūɣ, not sure why it won't show :/