Sound Change Quickie Thread

[dɮ the phoneme]

After an unfortunately long period of being far too busy/tired to get any conlanging done, I finally have time again, but seem to be stuck with a case of "conlanger's block". I'm looking to derive more complex consonant clusters in a daughter language of Proto-Yonutian, whose syllables are entirely CV(C). Proto-Yonutian primary stress is always penultimate, with (maybe) secondary stress on heavy syllables (those with a coda consonant). The one change I'm more or less settled on is V > ∅ in syllables following the primary stress. But I'm struggling to think up other reasonable conditioning environments for vowel deletion, which are sufficiently "local" that they don't end up creating alternations whereby half the vowels randomly delete when a prefix gets added. That might be a mild exaggeration, but my general problem is that with Proto-Yonutian's extensive prefixal morphology, a lot of potential syncope rules end up creating too many vowel alternations for my liking, and these alternations are unpredictable enough that I'm not sure how analogy might resolve them. Are there examples of syncope, or other cluster-producing changes, from highly synthetic and mostly prefixing languages (Navajo, perhaps?) which might be helpful here?

[Xwtek]

After an unfortunately long period of being far too busy/tired to get any conlanging done, I finally have time again, but seem to be stuck with a case of "conlanger's block". I'm looking to derive more complex consonant clusters in a daughter language of Proto-Yonutian, whose syllables are entirely CV(C). Proto-Yonutian primary stress is always penultimate, with (maybe) secondary stress on heavy syllables (those with a coda consonant). The one change I'm more or less settled on is V > ∅ in syllables following the primary stress. But I'm struggling to think up other reasonable conditioning environments for vowel deletion, which are sufficiently "local" that they don't end up creating alternations whereby half the vowels randomly delete when a prefix gets added. That might be a mild exaggeration, but my general problem is that with Proto-Yonutian's extensive prefixal morphology, a lot of potential syncope rules end up creating too many vowel alternations for my liking, and these alternations are unpredictable enough that I'm not sure how analogy might resolve them. Are there examples of syncope, or other cluster-producing changes, from highly synthetic and mostly prefixing languages (Navajo, perhaps?) which might be helpful here?

Do that vowel deleting, and then run analogical change.

[akam chinjir]

Are there examples of syncope, or other cluster-producing changes, from highly synthetic and mostly prefixing languages (Navajo, perhaps?) which might be helpful here?

The Mayan language Mam has widespread deletion of pretonic vowels and reduction of posttonic vowels to schwa. It's not mostly prefixing, but there are enough stress shifts to give you surprising alternations.

[Xwtek]

Is this sound change realistic?

s > z / V_V or V_[+voice] or {i e}_#
ns> nz

The problem is the allowed final consonant cluster was only /n s ʔ h j w ns nʔ jʔ wʔ jn wn/

[mèþru]

Sounds realistic to me.

[dɮ the phoneme]

Per my previous post, I've settled on the following changes. In Proto-Yonutian, primary stress is always penultimate, and secondary stress falls on all heavy syllables except the ultima, since it immediately follows primary stress (additional question: is that a realistic stress system?).

V > ∅ / ˈσ_
σˈσ > ˈσσ /_#
unstressed V > ∅ / _ˈσ, except in initial syllables

Thus:

*ˌtemfesuˈtusoq > ˌtemfˈsutusq
*koˈdemas > ˈkodems

How realistic do these look? What about blocking the second change in the environment #_, so that instead I have *koˈdemas > koˈdems, for example? The justification for this being that initial syllables are almost always prefixes, and so might have a bias for remaining unstressed.

[Knit Tie]

How realistic would it be to have a sound shift were /pˤ fˤ bˤ vˤ/ become /kw xw ɡw ɣw/ intervocalically and word-initially, then the latter two clusters merɡe into /ɣw/ and subsequently become /ʕw/, which then simplifies into /w/? Word-initially, the Cw clusters develop a prosthetic /a/ before them to comply with the lanɡuaɡe's phonotactics, while word-finally and in clusters, the pharinɡealised labials simply lose the secondary articulation.

[Xwtek]

How realistic would it be to have a sound shift were /pˤ fˤ bˤ vˤ/ become /kw xw ɡw ɣw/ intervocalically and word-initially, then the latter two clusters merɡe into /ɣw/ and subsequently become /ʕw/, which then simplifies into /w/? Word-initially, the Cw clusters develop a prosthetic /a/ before them to comply with the lanɡuaɡe's phonotactics, while word-finally and in clusters, the pharinɡealised labials simply lose the secondary articulation.

About /ɣw/, it can be immediately merged into /w/, without intermediate step of /ʕw/. As gw -> ɣw, it's not possible without g -> ɣ. About /ˤ/->/w/, I don't know, but it looks rather implausible too.

[Tropylium]

As gw -> ɣw, it's not possible without g -> ɣ.

Latin shifts *gw to /w/ except after a nasal (e.g. veniō, vīvus, vorō) but retains /g/.

[Pabappa]

How realistic would it be to have a sound shift were /pˤ fˤ bˤ vˤ/ become /kw xw ɡw ɣw/ intervocalically and word-initially, then the latter two clusters merɡe into /ɣw/ and subsequently become /ʕw/, which then simplifies into /w/? Word-initially, the Cw clusters develop a prosthetic /a/ before them to comply with the lanɡuaɡe's phonotactics, while word-finally and in clusters, the pharingealised labials simply lose the secondary articulation.

whats up with your g key? do you have it set that way so you can type IPA? just curious.

Regarding the question, I'd recommend a shift of /pʕ bʕ/ > /kp gb/, though Im not sure it's attested. If you can at least get there, you can peel away the second stop in steps like /kp/ > /kɸ/ > /kw/. /fʕ vʕ/ could perhaps just shift directly to /xʷ w/. Also, Im not sure why your language would need to prohibit initial /kw/ ... it's common for that to pattern as /kʷ/, i.e., as a single phoneme. but if your intermediate step is /kp/ you might want to set up a rule like that even so.

[mèþru]

Probably a prohibition on initial clusters, and it being /kw/ instead of /kʷ/

[Knit Tie]

About /ɣw/, it can be immediately merged into /w/, without intermediate step of /ʕw/. As gw -> ɣw, it's not possible without g -> ɣ. About /ˤ/->/w/, I don't know, but it looks rather implausible too.

Ah, I should explain the context! The language does, in fact, transition all /ɡ/ to /ɣ/, which later become /ʕ/ like in Hebrew, and those /ʕ/s later become /ŋ/ throuɡh rhinoɡlottophilia before vowels and are simply elided in all other positions, so I thouɡht that they would elide before /w/ as well.

whats up with your g key? do you have it set that way so you can type IPA? just curious.

I use the IPA keyboard app, it's ɡenerally amazinɡ.

And transitioninɡ to /kp/ and /ɡb/ does sound nice and appropriate for the lanɡuaɡe's African substrate, thanks!

Probably a prohibition on initial clusters, and it being /kw/ instead of /kʷ/

Exactly this. The language only ever allows VCCV clusters and uses prothesis to eliminate all other ones.

[mèþru]

I would think that /ʕw ʕj/ would turn into /ŋʷ ɲ/ instead. which could then either merge back into /ŋ/ or get reanalised as /ŋw nj/. You could even have one go one way and the other go the other way.

[bbbosborne]

can glottalized nasals turn into implosives?
/mˤ/ -> /ɓ/
/nˤ/ -> /ɗ/
etc.

[Whimemsz]

I don't know if there's a real-world precedent but I don't see why not. Go for it

[Xwtek]

How about
d -> r / #_T
t -> s / #_T
k -> h / #_T

But
p -> p / #_T
b -> b / #_T

where T is stop.

[k1234567890y]

How about
d -> r / #_T
t -> s / #_T
k -> h / #_T

But
p -> p / #_T
b -> b / #_T

where T is stop.

looks fine to me

[Raphael]

How plausible would it be to have a consonant appear out of nowhere after word-final /u/?

[WeepingElf]

How plausible would it be to have a consonant appear out of nowhere after word-final /u/?

Not very plausible, I think. Depends on what consonant appears "out of nowhere". Something like u > ou > ov is perhaps marginally possible.

[Man in Space]

I think it was on the ZBB that I remember someone mentioning an attestation of u i > uk ic / _#.