I've recently been perusing various articles about the historical development of the Algonquian languages, and there are enough strange sound changes floating around in the family that I thought it would be a good idea to start a thread to collect out-of-the-ordinary sound changes.

Shawnee had an unconditional (so far as I can tell; cf. Miller, 1959, An Outline of Shawnee Historical Phonology) change *s -> θ. (EDIT: H13 informs me over IRC that this isn't actually a weird change at all and happened in a few places in South-East Asia, but whatever, I'm leaving it here).

Arapaho has enough weird sound changes for an entire family. A rundown of the wackier stuff from Goddard 1974:






Blackfoot changed nasals to /ʔ/ before stops (with a couple of exceptions and caveats.)

Cheyenne inserted an epenthetic /h/ before stops between vowels: p t k -> hp ht hk / V_V.

(I can't find anything about Eastern Algonquian, but Eastern Algonquian is fairly conservative and most of the relevant papers are conference papers to which I don't have access. I'll probably add more later, but I need to go bake.

tS -> T unconditionally is also not a weird sound change, and from this we learn that making a weird sound changes thread is retarded because people will keep posting things that are not uncommon because they're not into sound changes enough.

I have therefore taken the liberty to make this just a "Sound Changes Thread", which is both conceptually more useful and practically less likely to embarrass all contributors.

s > T happened in Bashkir and Zhuang, but what is unweird about tS > T

tS > T happened from (post-?)Vulgar Latin to certain Castilian dialects of Spanish. So maybe not unweird, but definitely attested. I bet if we cast a wide net, we'd find a lot of tS > T.

[20:54] <H13> shit conlang is from
[20:54] <H13> < m n ŋ p t č k ʔ f s š h w r l y a ɨ ɔ u ə i o u
[20:54] <dhok> no *ɛ in proto-shit
[20:54] <dhok> ?
[20:54] <dhok> oh i see it is shit manchu
[20:55] <H13> yes
[20:55] <H13> voicing is from +ATR
[20:55] <H13> and the vowels got shifted around
[20:55] <H13> mostly by "everything is E or e now"
[20:56] <dhok> +ATR to voicing
[20:56] <dhok> that is a cool change
[20:56] <dhok> is that attested
[20:56] <H13> several times
[20:56] <H13> prominently in turkish
[20:56] <dhok> k00l
[20:56] <dhok> also in africa presumably?
[20:56] <H13> yes
[20:57] <dhok> hmmmmmm
[20:57] * dhok might steal this
[20:57] <dhok> do you actually need a +ATR harmony feature
[20:58] <H13> no
[20:58] <H13> it can be incidental
[20:58] <H13> proto-turkic had it incidentally to back/front
[20:58] <H13> and SW turkic went and voiced plosives before front vowels
[20:59] <dhok> fascinating
[20:59] <dhok> including rounded ones?
[20:59] <H13> yes
[20:59] <H13> *kör > gör
[21:00] <dhok> ooooh
[21:03] <H13> but I am suspecting that for this to work you need +front < +atr
[21:03] <H13> which is the fancy new altaicist hypothesis wrt turkic
[21:04] <H13> (this is attested in somali and mongolic)

...

[21:11] <H13> dhok: the *standard* hypothesis is that mongolic did it the other way around
[21:11] <H13> like
[21:11] <H13> +front > +ATR