Qori languages

Qori is a family of languages spoken primarily by the Qori people living in the Qoriyan Mountains. The language family is subdivided into four major groups (Northwest, Northeast, Southwest, Southeast), a division based on two isoglosses regarding pronounciation (one isogloss being South-North division, the other being an East-West division). These languages are to a large extent mutually intelligible, to the point where some consider them dialects of the same language (of the one Qori language), rather than separate languages.

Isoglosses
The Qori languages are divided into four groups: Northwestern, Southwestern, Northeastern, Southeastern. These four groups are created by two major divisions, two major isoglosses, or rather, two major optional sound shifts from Proto-Qori. One of these divisions is a North-South division that splits the Qori languages into Southern and Northern groups mainly regarding their treatment of the dorsal consonants of Proto-Qori; the other is the East-West division that splits the Qori languages into Eastern and Western groups mainly regarding their treatment of the rounded vowels of Proto-Qori. When combined, these two divisions divide the Qori languages into four groups: Northwestern, Southwestern, Northeastern, Southeastern.

To demonstrate, the Proto-Qori word became  in Northeastern Qori,  in Northwestern Qori, remained  in Southeastern Qori, became  in Southwestern Qori - at least before every other sound shift that affected individual languages/dialects.

It's important to note, that ever since the two sound shifts that created these four groups, there has been a lot of migrations, exchange of loanwords, and hybridization between languages/dialects, as well as new sound shifts that transcended traditional dialectal boundaries.

The North-South division
Proto-Qori had a consonantal system in which consonants - with the exception of - had two variants: a light and dark variant. Proto-Qori also had a rule that dark consonants could only appear before back vowels and other dark consonants, while light consonants could appear only before front vowels and other light consonants.

One pecularity was the fact that the dark counterpart of was in fact a fricative  rather than a stop - or, that the light counterpart of  was a stop  rather than a fricative. The light dorsal consonants may or may not have been palatalized like other light consonants (or had a more palatal, post-palatal / pre-velar or palato-velar place of articulation, rather than fully velar), but the dark dorsal consonants were uvular for certain.

This Northern-Southern division concerns the treatment of the Proto-Qori dorsal consonants and optionally the light coronal series  as well - the labial consonants  and the dark coronal consonants  were not affected by this isogloss.

The Northern languages
The Northwestern and Northeastern languages/dialects had a shift where the original light dorsal became palatal, while the original dark dorsal  became velar.

The new palatal stops sounded similar enough to the original light coronal  to create a crowding effect, where the original light dorsal series  had to shift somewhere to maintain a distinction.

In most of these languages/dialects, became shibilants, while  either lost its palatalization (causing a merger with ) or merged with the new. In some languages/dialects, the original light-dorsal non-sibilants became retroflex, while the shibilants  either remained as they were, or became shibilants. In languages/dialects that introduced a new class of retroflex consonants, usually got affricated.

In pretty much all of these languages/dialects, got debuccalized to.

The Southern languages
In the Southwestern and Southeastern languages/dialects, the dorsal consonants retained their original pronounciations, meaning that the dark dorsal consonants remained uvular, the light dorsal consonants remained velar. Most of these languages/dialects just straight-up shifted the light coronal series to alveolo-palatal, with the optional affrication of  to.

Some Southern languages/dialects merged the uvular stop with the corresponding fricative  or turned it into a glottal stop.

The East-West division
Proto-Qori had an 8-vowel system, in which there were three front unrounded vowels, two front rounded vowels, two back rounded vowels, and one back unrounded vowel. The rule was that non-initial front vowels could only appear after light consonants, and that non-initial back vowels could only appear after dark consonants.

The East-West divide concerns the treatment of the rounded vowels of Proto-Qori - the unrounded vowels  weren't affected by this isogloss.

The Western languages
The Northwestern and Southwestern languages/dialects replaced the front-back distinction for rounded vowels with a tongue-root distinction. The original back rounded vowels became back rounded vowels with a retracted tongue root, while the original front rounded vowels  beame back rounded vowels with an advanced tongue root.

The Eastern languages
The Northeastern and Southeastern languages/dialects preserved the original vowel system, keeping the front-back distinction for rounded vowels... at least at first. Some languages/dialects shifted the front rounded vowels into rising diphthongs, fronting diphthongs , rounding diphthongs , or straight-up unrounded them , but the majority clearly preserved them as they originally were.