Proto Vauqun-Adzovъd

Summary
Descendant of Proto Csillan. Basically Csilla's equivalent of PIE. Unlike PCS, PČA 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, all vowel qualities distinguished length and some nasality. Though all vowel qualities are rather uncertain, the greatest debate surrounds the roundedness of the vowel , which does not seem to show any partiality to either rounded or unrounded in its descendants. Two pieces of evidence, obtained via internal reconstruction, seem to support either sides of the debate- on one hand, 's morphological alternation with  suggests that it was unrounded, whereas it's role as the product of umlaut for hints to a rounded quality. However, it should be taken into consideration that the product of an umlauted , which is indisputably rounded, is , which is indisputably unrounded. Regardless, the most common consensus is that the vowel  had various rounded and unrounded variants in specific phonetic, historical, situational and geographic contexts, which were stabilized by the various daughter languages. Thus, the preference over the grapheme  for this phoneme which itself is historically ambiguous to a rounded or unrounded vowel.

The stress was regularly placed on the second vowel of the root, unless that vowel was short and the following vowel was long (vowels lengthened via nasalization or other morphophonemics that were originally short do not affect this rule.)

Phonotactics
Legal Onsets: C, [+fricative][+plosive], Cj, [+obstruent][+fricative -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. Codas can occasionally contain up to three obstruents, particularly in verbs in the middle voice.

Legal Middle Clusters: any coda + onset, unless it breaks the following rules: 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
 * 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 plosives or fricatives are illegal (ie *æktti is illegal, but æksti, æktsi, ækþti, æktþi, æxtti, etc are all allowed, likewise with *æxssi)

Hiatus are generally avoided in standard productive morphology, but within roots and more often in derivation they are generally allowed.

Morphophonemics
Vowels underwent many shifts under certain conditions, with varying regularity: There were also a few grammaticalized consonant shifts which are discussed in the grammar section.
 * 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. This sometimes led to stem alternations, such as the noun kost ę (sword) gen sing kest eni, nom pl kosten ą , gen pl kesten 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  when stressed, long or at the end of a word. This is the only umlaut that can be regularly reconstructed to PPČ, but many daughter languages innovated new umlauts. It affected all applicable short vowels before the .
 * /a æ o u y/ > /æ e e y i/

Thematic Declension
The thematic declension contains entirely animate nouns ending in any short vowel except æ or u. 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 x 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 -i 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 gveną gen pl gvenæmi and tjeko (boar) is tjekǫ, gen pl tjekemi (i-stem endings are used in the plural.)

Yet another class of strangely declined nouns include those with a nominative singular ending with -āx, -ēx, -ōx, -ǣx, -ȳx. They decline like regular -a, -y, -o, -e and -i stems but form plurals with the infix -v, appended to the nominative singular w/out the -h (and associated lengthening), thus the nom plural of kamōx is kamov, gen pl kæmevi (same as above.)

Athematic Declension
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.

Some athematic nouns feature umlaut in the nominative singular, but not in any other case or number. Likely what happened (exactly what happened) is that these nouns derive from a diminutive suffix (every one of these nouns is a diminutive) that was originally -þi, which was reduced to -þ.

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

Articles
The only article reconstructable for Proto Čeuň-Adzovŭd 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-, x(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, at least not till much later.) Thus kēro (an animal) > s kēro (the animal) but dōfo (a life) > z dōfo? s tōfo? (the life). Likewise jīnoræþ (water) >       f jīnoræþ (the water) but bowæþ (a sight) > v bowæþ? f powæþ? (the sight).

Adjectives
The comparative degree is formed via the infix -įg- before the nominative singular (note that this causes umlaut.) Most native adjectives form the superlative via the infix -axm- before the nominative singular, whereas other adjectives (presumably those borrowed from other languages) formed it using the adverb taxmęk. Adverbs are regularly formed with the suffix -sų.

Adjectives with a nom. anim. sing in -ate have an irregular nom. inan. sing. in -aþ, all other forms are regular (effo/effu & ēkje/ēkjæ, but fate/faþ).

Verbs
The above endings are added to the appropriate stem of a verb (each tense has a unique stem, given in a verb's principle parts.) Only the active voice features distinct endings for the imperfect and aorist tenses, the middle and passive voices just use one set of "past" suffixes. Note that the (j) in the imperfect active endings serves to prevent hiatus if the imperfect stem ends in a vowel, likewise the parenthetical vowels in the passive endings are present when added to a principle part ending with a consonant. For the passive second person plural, the (w) prevents hiatus while the (j) appears only after a consonant. The middle & present active systems have no inherent epenthetic vowels, the consonant cluster simply tries to adapt to be pronounceable and legal (zrēd + ks > zrēþks), but if the result was still difficult to pronounce (ie zrēd + t, vemak + ks) speakers often insterted an epenthetic  (zrēdyt, vemakyks) or simply drop the first obstruent (zrēt, vemaks) In some languages, the  became standard when the stem ends in any obstruent. Note that this  does not become  in closed syllables; in all likelihood it might have been closer to a schwa [ә].

Note that the impersonal present can also serve as an gerund belonging to the athematic declension.

First Principle Part
The first principle part is usually the base root of the verb + the infinitive ending -ju. Exceptions to this include impersonal verbs (such as, where the first principle part is the present impersonal, and verbs were the base root is aoristic in nature and the present is formed via derivation (usually the inchoative suffix -đē), such as ęwæþk (I knew) pres act inf ęweđēju.

The first principle part is the source of all finite present forms of the verb, both active and mediopassive infinitives (the mediopassive is the active with the prefix kyz-) and both active and mediopassive participles of all tenses (note that the participle tenses differ from the finite verb tenses.)

Second Principle Part
The second principle part gives the finite aorist forms for all voices. In verbs that feature reduplication, the second principle part is formed by lengthening the stem vowel (riaju (to live) > riāþk (I lived), voju (to go) > vōþk (I went)). In most other verbs, it's formed with the suffix -se added to the root stem (togju (to say) > toksæþk (I said)). This occasionally eats the last vowel before the infinitive (aneju (to travel) > ąseþk (I traveled)) The most common exception to this rule, naturally, is aorist-root verbs, which lack any explicit morpheme for the aorist tense. Some verbs, though, form the 2nd pp via the prefix ki- (this usually appears in semantically passive verbs, such as sxenoju (to sleep) kisxenoþk (I slept), but not all; dvatyju (to sit) > dvatseþk (I sat).

Third Principle Part
The third principle part is only found in some archaic verbs, where it is formed via reduplication (ie riaju (to live) > reriajēk (I was living), voju (to go), vawojēk (I was going)). This root gives the imperfect forms of the verb. In every other verb the imperfect tense uses the 1st pp.

Subjunctive Mood
The subjunctive mood has reconstructed uses entailing desired, theoretical and possible actions. As far as subordination goes, the subjunctive was only used in conditional clauses, other secondary clauses were usually reserved for non-finite forms or the optative mood.

In archaic or "reduplicating" verbs (discussed above), the subjunctive in all tenses is formed through a non-productive ablaut: riat (I live) > riot (I should live). Standard verbs take an infix -wi- before the personal ending in the present: ōɣot (I converse) > ōɣowyt (I should converse). Root-aorist verbs take this infix in the aorist tense, since for those verbs that tense is the most basic form. The aorist and imperfect subjunctives are regularly formed via the prefix vo-. Note that this takes the place of any other tense marker: ąsæþk (I traveled) > voanæþk (if I traveled) (the -se- infix was removed), same with kiōttyþk (I witnessed) > voōttyþk (I should have witnessed). Note that hiatus is not blocked in this prefix, even when the vowels are identical. This and the regularity of both of these morphemes suggest that the subjunctive mood (outside of reduplicating verbs, at least) was a recent innovation at the time of late PČA.

Optative Mood
The optative mood is usually used in subordinate clauses of desire, command, purpose, etc. As a finite verb it functions similar to an imperative, which is otherwise lacking (or perhaps unreconstructable) in PČA.

It is regularly formed (yes, even in reduplicating verbs) by combining the appropriate participle with a clitcized auxiliary (which is only conjugated in the present active, thus the tense and voice is only marked on the participle. Note that this results in the optative mood having a different tense distinction from other verbs.) The auxiliary's conjugation is given below.