Atypical vowel system changes

[Travis B.]

Does anyone know of any atypical vowel system changes in dialects they are familiar with?

The reason why I ask is around here there are some, well interesting vowel system changes that merge some cases of LOT and PALM with THOUGHT and CLOTH in different ways.

E.g. My father and I have three low vowel phones, [a ɑ ɒ], where [a] is from LOT and PALM except adjacent to /r w h kw gw/*, [ɑ] is from LOT and PALM adjacent to /r w h kw gw/*, and [ɒ] is from THOUGHT and CLOTH.

However my mother and my daughter have two different systems that have less low vowel phones. In the case of my mother she has [a ɑ], where my [a] is LOT and PALM except adjacent to /r w h kw gw/* and [ɑ] is from LOT and PALM adjacent to /r w h kw gw/* and from THOUGHT and CLOTH. My daughter has the same system except where my mother has [ɑ] she has [ɒ].

* except, in most cases, where before /r/ before a fortis obstruent, where then LOT and PALM are [ʌ]

The thing is I have not heard of these systems anywhere, whether the three-way system my father and I have, or either two-way system that partially merges cot and caught that my mother or my daugter have.

[Sumelic]

E.g. My father and I have three low vowel phones, [a ɑ ɒ], where [a] is from LOT and PALM except adjacent to /r w h kw gw/*, [ɑ] is from LOT and PALM adjacent to /r w h kw gw/*

That's an interesting description. What are some example words where /kw gw/ conditions [ɑ]? I can't think of any examples of words with /ɑkw/ or /ɑgw/, and wouldn't it be possible to just view /kwɑ/ and /gwɑ/ as being conditioned by the preceding single segment /w/?

[KathTheDragon]

wouldn't it be possible to just view /kwɑ/ and /gwɑ/ as being conditioned by the preceding single segment /w/?

Not necessarily, since /wɑ/ could behave differently word-initially or after s

[Travis B.]

Two words that show this are [ɑ]qua and Aconc[ɑ]gua.