Proto-Limjiang language

The Proto-Limjiang language is the common ancestor of all the Limjiang languages (namely Shár and Sak). The language was never attested - at least not in any form that indicates the pronounciation of words (proto-writing) -, all its speakers lived in an era where phonetic writing wasn't present.

Uniform changes
Arguably, the first of the steps proto-Limjiang took in its evolution from proto-Torgyrian was the merger of the front rounded vowels and the back unrounded vowels  into a single group of central vowels  undefined for rounding. This resulted in a 8-vowel system.

Likewise, the alveolar series and postalveolar series  were conflated into one series  which had palatal(ized) allophones before front vowels. The lateral affricates lost their lateral quality and became regular sibilant affricates. Clusters like were likewise also coalesced into affricates. The uvular stop was sprintalized to  - in sharp contrast with Proto-Elven, where it became a glottal stop  instead.

The last change that was truly uniformly applied to all variants and dialects of proto-Limjiang was the heavy limitation on syllable codas and onsets: syllable-initial clusters were simplified until a syllable onset could only consist of a single obstruent and either, or  - for instance,  were simplified into ,  respectively became , etc.

Codas likewise underwent heavy simplifications, until the only valid codas were - this was achieved by, among other things:
 * debuccalizing coda-position to
 * partially debuccalizing coda-positon clusters such as into
 * turning coda-position into preaspirated
 * breaking up the coda-position voiced stops into
 * any resulting became
 * the outcome of coda depended on the vowel that preceded it, with front vowels yielding, back and central vowels yielding
 * Vocalizing coda-position into either  or, depending on what kind of vowel was behind it.
 * turning the combination of with any kind of stop consonant into a cluster
 * Any combination of coda-position consonants that did not fit into the aforementioned mold was simply simplified, or completely removed altogether

Divergences
At some unknown time - probably some time around 6500 BEKE - proto-Limjiang diverged into three clearly distinguishable varieties demarked by geography.




 * Southern Limjiang metathesized the clusters to, e.g.  ➞ ,  ➞ , etc. Coda  became.
 * Any instance of two semivowels colliding - - resulted in the first semivowel being fortified into a stop or fricative.
 * Northern Limjiang underwent a shift in which the coda-position respectively became, e.g.  ➞ ,  ➞ ,  ➞ , etc. Coda  became.
 * Middle Limjiang combined the aforementioned two shifts: the Southern metathesis of semivowels when part of complex codas and the Northern debuccalization. Coda became, just like in Northern Limjiang.

The best way to demonstrate the difference between the three variants is with the proto-Torgyrian word, which became in early proto-Limjiang, then, it respectively became...
 * in Northern Limjiang, which eventually shifted to
 * in Middle Limjiang
 * in Southern Limjiang

Eventually, coda and  simplified to  in all three variants, and likewise,  also simplified into  in Northern Limjiang as well eventually, resulting in all three varieties only permitting simple syllable codas that can consist of.

Consonants

 * The dorsal consonants coalesced with following  to form labio-dorsal consonants.
 * The dental and alveolar consonants most definitely had palatalized allophones before front vowels and . There may or may not have been also a distinction between different degrees of palatalization.
 * The glottal consonants only existed in the syllable coda.

Vowels

 * The language also had diphthongs.
 * Every vowel other than could form a diphthong with, e.g.
 * Every vowel other than could form a diphthong with, e.g.