Sound Change Quickie Thread

[akam chinjir]

Thank you!

[Zju]

Is there any language that contrasts front and back palatal stops? Presumably from earlier /tʲ kʲ/.

[Nortaneous]

Is there any language that contrasts front and back palatal stops? Presumably from earlier /tʲ kʲ/.

Yanyuwa

if you accept affricates as close enough to stops, Japhug and some other languages in that area

[holbuzvala]

I heard it said (I believe on this thread) that /w/ > /v/ unconditionally is so common as to be unremarkable.

Is /Cʷ/ > /Cv/ likewise common unconditionally? And if not, what sorts of things might condition it?

My thoughts on possible environments to encourage /Cʷ/ > /Cv/:

More: show

1. dissimulation: Cʷ/Cv/_[ouw]
2. before any vowel: Cʷ/Cv/_V

[Nortaneous]

I heard it said (I believe on this thread) that /w/ > /v/ unconditionally is so common as to be unremarkable.

Is /Cʷ/ > /Cv/ likewise common unconditionally? And if not, what sorts of things might condition it?

My thoughts on possible environments to encourage /Cʷ/ > /Cv/:

More: show

1. dissimulation: Cʷ/Cv/_[ouw]
2. before any vowel: Cʷ/Cv/_V

in some languages' transcriptional conventions ʷ is used very broadly - certain NWC 'rounded' consonant series are actually labiodentalized (and others can be coarticulated, e.g. /tʷ/ [t͡ʙ̥ ~ t͡p])

[dɮ the phoneme]

Does any language (that allows word final consonants) delete vowels in final syllables across the board, regardless of whether the syllable is open or closed?

[Pabappa]

French did it at some point .... "all post-tonic vowels were deleted" ... but there is often a remnant schwa to break up imperssible consonant clusters. I think that, because this schwa is predictable, it can be analyzed as /Ø/ as are the schwas of Georgian and some other languages with rich consonant clusters.

[Richard W]

French did it at some point .... "all post-tonic vowels were deleted" ... but there is often a remnant schwa to break up imperssible consonant clusters. I think that, because this schwa is predictable, it can be analyzed as /Ø/ as are the schwas of Georgian and some other languages with rich consonant clusters.

I beg to differ. My recollection is that from Latin to Old French, "all post-tonic vowels were deleted except /a/". The excepted /a/ does seem to have become schwa, and the 'remnant' schwa is word final even if the deleted vowel was not. Most of the excepted /a/ and some of the remnant schwas were then dropped in Modern French - I believe the rules for the retention were the same in the two categories. That's two widely separated points.

[StrangerCoug]

So in an attempt to break Standard Average Me, I came up with this kitchen sink phonology of sorts in the scratchpad for my protolang (it doesn't have its own thread yet):

Phonology
⟨ʼ ġ q qʼ g k kʼ ŋ d t tʼ n b p m⟩ /ʔ q qʰ qʼ k kʰ kʼ ŋ t tʰ tʼ n p pʰ m/
⟨ġẋ qẋ qẋʼ gx kx kxʼ dth tth tthʼ bf pf⟩ /q͡χ q͡χʰ q͡χʼ k͡x k͡xʰ k͡xʼ t͡θ t͡θʰ t͡θʼ p͡f p͡fʰ/
⟨z c cʼ j ch chʼ dl tl tlʼ⟩ /t͡s t͡sʰ t͡sʼ t͡ʃ t͡ʃʰ t͡ʃʼ t͡ɬ t͡ɬʰ t͡ɬʼ/
⟨h ẋ x sh s th f⟩ /h χ x ʃ s θ f/
⟨hl l hr r hy y hw w⟩ /ɬ l ɾ̥ ɾ ç j ʍ w/
⟨a ā i ī u ū⟩ /a aː i iː u uː/

The diphthongs /ai au/ are allowed as well, written ⟨ai au⟩.

The onset is mandatory and can be any consonant; by convention, ⟨ʼ⟩ is not written word-initially. /ɾ j w/ can follow any non-affricate occlusive except /ʔ/, /l/ can follow any non-affricate occlusive except /ʔ/ and coronals. The contrasts /ɬ l/, /ɾ̥ ɾ/, /ç j/, and /ʍ w/ exist only word-initially and after vowels; the voiced version is always written after a consonant for simplicity even though phonetically voiceless.

Stress
Stress is on the last long vowel or diphthong if there is one; if there are no long vowels or diphthongs, then stress is on the penultimate syllable.

Not in my scratchpad, but worth noting: I write /ç/ instead of /j̊/ pretty much because I expect the latter to be pretty unstable and change quickly to the former—same logic behind which I write /ɬ/ and not /l̥/. Further, ⟨ẋ⟩ is supposed to represent a uvular fricative, not a velar fricative.

Since this is a protolang and I have 50 consonants but only 6 vowels, I'd like the consonants to wreak some interesting havoc on the vowels in my daughterlangs. I know velars I meant uvulars tend to retract vowels; anything else I can do? (I don't want tone at this point.)

Also, how stable do you expect the given consonant inventory to be? I stole some ideas from Tlingit to design it, so I feel I have at least partial justification for it, but I'd expect daughterlangs to pare it down a little.

[Pabappa]

Im guessing /ġ/ is also uvular? If so, /ġẋ/ is an odd choice .... if it starts voiced and ends voiceless .... but no weirder than /gh/, i suppose. Or is the fricative allophonically voiced here?

[StrangerCoug]

Im guessing /ġ/ is also uvular? If so, /ġẋ/ is an odd choice .... if it starts voiced and ends voiceless .... but no weirder than /gh/, i suppose. Or is the fricative allophonically voiced here?

There are no voiced stops. ⟨ġ⟩ represents /q/ and ⟨q⟩ represents /qʰ/.

[Pabappa]

Ah, sorry, I misread. Now that I see it clearly I think the inventory could be quite stable, and you'd have a free hand to do what you want. Still, syllable structure matters .... the closer to CV the language is, the less likely consonants are going to start creating new vowels out of old ones. You seem to list /l/ twice in the phonotactics .... apologies if Im misreading again ... but if /C/ + l is the most common cluster in the language, i'd expect it to be a fertile source of sound change.

[StrangerCoug]

You seem to list /l/ twice in the phonotactics .... apologies if Im misreading again ...

That one is a legitimate error on my part; will go back and fix so it isn't contradictory.

ETA: Not so much an error as an implausibility, but the liquids and semivowels should be voiced after nasals, not voiceless after all consonants. I'll leave the debate open, though.

[Nila_MadhaVa]

So I’m messing around with a consonantal root type proto lang and I’ve run into something that I’d really appreciate some feedback on. From what I can tell, a fair number of sound changes that it’s daughters go through will only end up phonemic in affixes, particles, determiners etc. while only being allophonic in roots/stems. Having had a quick look at how begadkefat occurs in Hebrew, this outcome seems plausible to me, but since I don’t have a particularly in-depth understanding of any of the Semitic languages, any input would be very welcome.

A question that occurs to me in regards to begadkefat is why the lenition hasn’t spread to the rest of the paradigm by analogy? Is it because the affected forms are in the minority, statistically speaking? Or is it perhaps because they are less ‘basic’, or less commonly used?

On that note, which forms would lead to a change in the entire paradigm by analogy? The statistical majority, the most basic, the most commonly used?

[Knit Tie]

tɹ dɹ → tʃ dʒ → kʲ gʲ

And then the preexistent velars back so there's a distinction between /kʲ gʲ/ and /q ʁ/, even though though the latter two are post-velar and not fully uvular. /ŋ/ is fully velar and unpaired.

[Pabappa]

A question that occurs to me in regards to begadkefat is why the lenition hasn’t spread to the rest of the paradigm by analogy?

Well, that would give us a system entirely without singleton stops, if I read you right. We'd have the geminates, but they would only occur between vowels, so that would be typologically unusual. I think if something like that were to happen it would likely affect only a subset of the stop inventory, not the whole thing. e.g. Arabic did /p/ > /f/ without /b/ > /v/, and did /g/ > /dž/ without /k/ > /tš/.

tɹ dɹ → tʃ dʒ → kʲ gʲ

And then the preexistent velars back so there's a distinction between /kʲ gʲ/ and /q ʁ/, even though though the latter two are post-velar and not fully uvular. /ŋ/ is fully velar and unpaired.

Fronting is always more common than backing, and I'd be especially hesitant to say yes if this shift is unconditional such that you would end up with /qi/ from earlier /ki/. tɹ dɹ → tʃ dʒ is certainly fine though.

[Knit Tie]

tɹ dɹ → tʃ dʒ → kʲ gʲ

And then the preexistent velars back so there's a distinction between /kʲ gʲ/ and /q ʁ/, even though though the latter two are post-velar and not fully uvular. /ŋ/ is fully velar and unpaired.

Fronting is always more common than backing, and I'd be especially hesitant to say yes if this shift is unconditional such that you would end up with /qi/ from earlier /ki/. tɹ dɹ → tʃ dʒ is certainly fine though.

I'd say it's less backing and more of a fusion of two series - after the alveolar approximant became /j/, there were a lot of velar+yod clusters, so I was thinking if it would be possible that the velar+yod and alveolar+yod series merge into something in the middle between then. And, of course, /ki/ will always become /kʲi/.

[Nila_MadhaVa]

Well, that would give us a system entirely without singleton stops, if I read you right. We'd have the geminates, but they would only occur between vowels, so that would be typologically unusual. I think if something like that were to happen it would likely affect only a subset of the stop inventory, not the whole thing. e.g. Arabic did /p/ > /f/ without /b/ > /v/, and did /g/ > /dž/ without /k/ > /tš/.

It would be unusual, but would that, in and of itself, stop it spreading to the rest of the paradigm if the conditions for analogy were right? The reason I asked was to find out what conditions would/could cause the entire paradigm to change by analogy since I'm trying to figure out when changes would become phonemic in my own roots.

[linguistcat]

For the purposes of sound changes, is /ɨ/ more likely to pattern with front or back vowels?

[bradrn]

For the purposes of sound changes, is /ɨ/ more likely to pattern with front or back vowels?

Based on the changes from the Index Diachronica, it looks like /ɨ/ most often changes to /i/, suggesting that it patterns with front vowels. But as with everything with sound changes, there’s lots of uncertainty: almost no language contrasts /ɨ/ and /ɯ/, and /ɨ/ → /ɯ/ is a very plausible sound change, so if you want /ɨ/ to pattern with back vowels you could easily do /ɨ/ → /ɯ/ → [sound change involving back vowels].