Zompist Bboard Again

some kind of stress shift followed by i u > ɪ ʊ in (word-final?) closed syllables, then ɪ ʊ > e a (for ʊ > a, maybe ʊ > ɔ > a or ʊ > ʊə > oa > a)

How naturalistic/plausible could it be to have a morpheme that is /k/ when contained within a syllable containing a peripheral consonant but /ɣ/ elsewhen?
example:
mazûm > mazûğ
ḫuzum > ḫuzuğ
ḳurūśum > ḳurūśuğ
but:
rēbim > rēbik
surğum > surğuk

Ahzoh wrote: ↑Thu Oct 21, 2021 5:00 pm How naturalistic/plausible could it be to have a morpheme that is /k/ when contained within a syllable containing a peripheral consonant but /ɣ/ elsewhen?
example:
mazûm > mazûğ
ḫuzum > ḫuzuğ
ḳurūśum > ḳurūśuğ
but:
rēbim > rēbik
surğum > surğuk

Kalam /k/ is [ɣ] intervocalically (as distinguished from /ᵑɡ/), so this alternation is very nearly attested.

No I dont think so... I've never heard of a language that has differenbt forms for a consonant based on what the POA of a consonant in an adjoining syllable is. closest i can think of is Latin's /l/~/r/ allomorphy, but that is dissimilation, and is responsive only to one consonant: another /l/.

It’s for the same syllable though. The first three examples were all coronals in the onset; the last three were the peripherals.

Ahzoh wrote: ↑Thu Oct 21, 2021 5:00 pm How naturalistic/plausible could it be to have a morpheme that is /k/ when contained within a syllable containing a peripheral consonant but /ɣ/ elsewhen?
example:
mazûm > mazûğ
ḫuzum > ḫuzuğ
ḳurūśum > ḳurūśuğ
but:
rēbim > rēbik
surğum > surğuk

Seems unlikely as a straightforward phonetic development, but coronal vs. peripheral onset is an opposition that an analogical development could grow into, maybe from an earlier less predictable opposition in a language in which a similar but plausibly phonetically motivated alternation (w vs. j?) already

How do I create a situation where a consonant that cannot normally appear in the coda can still appear geminated across a syllable break (similarly to the sokuon in Japanese)?

You could generate the geminate by assimilation of an intervocal consonant cluster.

Polynesian went through a stage in which internal consonant clusters inherited from PMP still remained, but final consonants had dropped out. e,g, you could say "manta" but not "matan". this would be a fallback solution if nothing else feels right .

Pabappa wrote: ↑Sat Oct 23, 2021 6:18 pm Polynesian went through a stage in which internal consonant clusters inherited from PMP still remained, but final consonants had dropped out. e,g, you could say "manta" but not "matan". this would be a fallback solution if nothing else feels right .

Really cause i read proto oceanic lost word internal consonant clusters before it lost final codas. Have there been any recent papers in linguistics stating that developments from proto oceanic to proto polynesian could be better explained if word internal clusters were retained?

the paper i read could be eighty years old ... now that I think of it, im almost sure the only permissible clusters were nasal+stop, which we typically today think of as being not a cluster at all, even if it's bound to word-internal position. that said, there is a language in Polynesia, Tuvaluan, which allows geminate consonants and no other clusters, not even prenasals, so there's a precedent for consonant gemination based on vowel contraction. Tuvaluan's clusters only occur word-initially, however.

Pabappa wrote: ↑Sat Oct 23, 2021 8:20 pm the paper i read could be eighty years old ... now that I think of it, im almost sure the only permissible clusters were nasal+stop, which we typically today think of as being not a cluster at all, even if it's bound to word-internal position. that said, there is a language in Polynesia, Tuvaluan, which allows geminate consonants and no other clusters, not even prenasals, so there's a precedent for consonant gemination based on vowel contraction. Tuvaluan's clusters only occur word-initially, however.

Prenasal stops could occur in word initial position in proto oceanic so are best not analysed as clusters. Not sure about the exact developments from oceanic to polynesian since the paper (actually it's more like a guidebook) doesn't go into detail about them (only the beginning and final results) so i can't comment on their supposed retention. Anyways possibly the only way you could know they were retained word internally but not initially is if they had an effect on the following vowel or consonant most likely vowels since most austronesian roots are disyllabic.

Pabappa wrote: ↑Sat Oct 23, 2021 8:20 pm that said, there is a language in Polynesia, Tuvaluan, which allows geminate consonants and no other clusters, not even prenasals, so there's a precedent for consonant gemination based on vowel contraction.

This exact sound change is attested in most Micronesian languages, as well as others of the region: e.g. *tu-tuki → Kapingamarangi ttuki (Blust 2013 p657).

I recall someone said that if a sound change swept across an entire language that a pronoun would be the least likely to resist or be exempt from the change. However I wonder if they still could resist if influenced by other pronouns within the paradigm.

I have a change where all glides are elided intervocalically and the vowels around them coalesce. However this results in undesirable outcomes in my pronouns that I don't want.

the proto-language had these pronouns as stems: When they are cased, the resulting sound change would result in:
mi(y)-m > miyam > mīm (2ms-NOM)
ku(w)-k > kuwak > kūk (3fs-INS)
but the plurals would simply be:
mi(y)n-k > minak or mīnak (2mp-NOM)
ku(w)n-k > kunak or kūnak (3fp-INS)
And I don't like this outcome at all, even though they would not be particularly illegal phonotactically.

Ideally a outcome like this would be desired: mīyam (stress placement often lengthens vowels) and kūwam with the plurals mīnam and kūnak

I have a sound change whereby onset clusters of voiceless obstruent + stop give an aspirated stop, for example: So far so good. But I want a similar rule for onset clusters consisting of two voiced obstruents, such that the result be a voiced aspirate (or breathy voiced/murmured, phonetic details are not relevant here): I found that in the Index, but I still don't know if it goes well phonetically with the concurrent sound change with voiceless obstruents.

Just the other day I read that Old Khmer voiced onset stops became breathy: *kaa *ɡaa → *kaa *ɡe̤a, so now you have two data points.

Does the descendant case system look like it was naturally changed from the ancestor?

Ancestor: Note: C in above case both means "common gender" and "construct state"
Descendant:

Ahzoh wrote: ↑Thu Oct 21, 2021 5:00 pm How naturalistic/plausible could it be to have a morpheme that is /k/ when contained within a syllable containing a peripheral consonant but /ɣ/ elsewhen?
example:
mazûm > mazûğ
ḫuzum > ḫuzuğ
ḳurūśum > ḳurūśuğ
but:
rēbim > rēbik
surğum > surğuk

Perhaps a possible mechanism for this development:

Kusunda has two sets of vowels, e a o and i ə u. Most words can be pronounced with either set, you just can't mix the two (which is a pretty cool feature by itself), but some situations force certain pronunciations - for example, if a word has a uvular consonant in it, it must be pronounced with the e a o set .

I'm not sure if you can do anything with that, but that might be a possible way something like what you propose could develop. Maybe vowels are tense next to peripheral consonants and lax when they aren't, and this causes the k/ɣ alternation, or something. Of course, that might involve overhauling your entire vowel system, so maybe not ideal.

Is spirantisation triggered by fricatives or is it strictly a type of lenition? I.e. which of these is likely to occur:

P > F / s_V
P > F / V_V

If it's the latter, are there any other examples of lenition that doesn't include voicing?

Zju wrote: ↑Sat Jan 01, 2022 10:27 am Is spirantisation triggered by fricatives or is it strictly a type of lenition? I.e. which of these is likely to occur:

P > F / s_V
P > F / V_V

If it's the latter, are there any other examples of lenition that doesn't include voicing?

Both are likely to occur. The second one definitely happened in Goidelic.