Sound Change Quickie Thread

[Knit Tie]

As I've probably said my conlanɡ will have a transition from /ʶ/ - i.e. uvularization - to vowel quality: /ɛ i u/ become /æ ɪ~e o~ʊ/ adjacent to uvilarised consonants and /ɑ/ will raise to /ɔ/ everywhere except when adjacent to uvularised consonants. I'm wonderinɡ if there are any interestinɡ ways I can make uvularisation jump about or delete itself to have the resultinɡ /i ɪ~e ɛ æ ɑ ɔ o~ʊ u/ vowel system have some sort of harmony in it?

[Zaarin]

As I've probably said my conlanɡ will have a transition from /ʶ/ - i.e. uvularization - to vowel quality: /ɛ i u/ become /æ ɪ~e o~ʊ/ adjacent to uvilarised consonants and /ɑ/ will raise to /ɔ/ everywhere except when adjacent to uvularised consonants. I'm wonderinɡ if there are any interestinɡ ways I can make uvularisation jump about or delete itself to have the resultinɡ /i ɪ~e ɛ æ ɑ ɔ o~ʊ u/ vowel system have some sort of harmony in it?

Pharyngealization has a habit of spreading, usually blocked by high vowels, sibilants, and sometimes palatals. Since pharyngealization and velarization are hard enough to distinguish, never mind uvularization, I would expect it could behave in the same manner. And of course you could have straight up consonant harmony, wherein a word can only have uvularized or non-uvularized consonants (perhaps with certain consonants being neutral).

[Tropylium]

The problem is, palatalization is still only marginally contrastive in the initial consonant, owing to a couple of vowel shifts, which seems like an odd situation to me.

This is exactly how palatalization works in Veps, for example. Estonian is even more limited and only has contrastive palatalization word-finally (in standard Estonian only for coronals, in dialects also for labials and velars).

Aside from some widespread front-back merger of vowels, which I don't really want to do, I'm not quite sure how to make palatalization contrastive root-initially.

*j > dʲ is one common way to bring around more initial palatals; so is long-range assimilation: C…Cʲ > Cʲ…Cʲ.

If you don't like front/back mergers, another option for getting palatalization out of vowel contrasts are close/mid mergers: Ci Ce > Cʲi Ci (or, in your case, probably rather Cy Cø > Cʲy Cy). Or even in inverse: Ce Cø > Cie Cyø > Cje Cjø > Cʲe Cʲø (or Cʲi Cʲy), versus Ci Cy remaining unpalatalized (perhaps even partly getting retracted to Cɨ Cu).

[Pabappa]

So, I have a proto-lang with a very constrained root structure of CV(C)CV, and I'm trying to develop contrastive palatalization.

How constrained is the proto-proto-language? I have a very similar situation in my language family, and I decided to solve it by setting up vowel-initial words, and having all of the unstrssed initial vowels delete , leaving their effects on the secondarily initial consonants. This led to a language with almost all of the words being consonant-initial, and having a contrast that previoudly did not exist. (I used labialization and a disappearing /u-/, but of course it would work just as well with palatalization and a disappearing /i-/, or perhaps other vowels as well.)

If not that, then how about proposing that the first V in each root can have come from an earlier diphthong? Or even just reconstructing palatalized initials for the proto-language itself.

[Knit Tie]

I know there are languages where the voicing of stops and fricatives is dependent on their position - i.e. intervocalic -, but can a language develop positional voicing even for its sonorants, such as liquids and nasals? If we assume that the lanɡuaɡe has /j w ɾ l/ for liquids and /m n ɲ ŋ/ for nasals and starts devoicinɡ them in certain positions, what can those voiceless sonorants then turn into?


My end ɡoal is to create a very unusual allophony, you see, somethinɡ that is downriɡht weird typoloɡically.

[mèþru]

but can a language develop positional voicing even for its sonorants

yes

Voiceless /j l/ will probably shift to fricatives, the rest might just stay the same except voiceless. Another idea for the nasals is that they become pre/postnaslised voiceless stops. /ʍ/ could break into /hw/ or /xʷ/.

[Man in Space]

can a language develop positional voicing even for its sonorants, such as liquids and nasals?

Yes, this appears to have happened to a very small degree during the development of sonorant allophony in Wichita:

The allophones [ɾ] and [n] are in complementary distribution: It is [n] before alveolars (/t, ts, s/ and in geminate [nn]) and initially before a vowel, and [ɾ] elsewhere. Thus its initial consonant clusters are [n] and [ɾ̥h], and its medial & final clusters are [nts], [nt], [ns], [nn], [ɾʔ], [ɾh].

[Esneirra973]

Is /sP/ > /Pʼ/ a plausible sound change, with P being a voiceless plosive or affricate?

[mèþru]

yes

[Esneirra973]

yes

Thanks.

[Nortaneous]

I know there are languages where the voicing of stops and fricatives is dependent on their position - i.e. intervocalic -, but can a language develop positional voicing even for its sonorants, such as liquids and nasals? If we assume that the lanɡuaɡe has /j w ɾ l/ for liquids and /m n ɲ ŋ/ for nasals and starts devoicinɡ them in certain positions, what can those voiceless sonorants then turn into?


My end ɡoal is to create a very unusual allophony, you see, somethinɡ that is downriɡht weird typoloɡically.

Old Chinese had voiceless nasals > fricatives.

Is /sP/ > /Pʼ/ a plausible sound change, with P being a voiceless plosive or affricate?

Is the outcome an aspirate series or an ejective one? Aspirate would be more common - see Tsakonian and some Spanish dialects. For ejectives, the closest I know of is Korean.

[Esneirra973]

I know there are languages where the voicing of stops and fricatives is dependent on their position - i.e. intervocalic -, but can a language develop positional voicing even for its sonorants, such as liquids and nasals? If we assume that the lanɡuaɡe has /j w ɾ l/ for liquids and /m n ɲ ŋ/ for nasals and starts devoicinɡ them in certain positions, what can those voiceless sonorants then turn into?


My end ɡoal is to create a very unusual allophony, you see, somethinɡ that is downriɡht weird typoloɡically.

Old Chinese had voiceless nasals > fricatives.

Is /sP/ > /Pʼ/ a plausible sound change, with P being a voiceless plosive or affricate?

Is the outcome an aspirate series or an ejective one? Aspirate would be more common - see Tsakonian and some Spanish dialects. For ejectives, the closest I know of is Korean.

The outcome would be an ejective one.

[Zaarin]

I'd expect something more along the lines of /sP/ > /Pʰ/, /P/ > /Pʼ/.

[Pabappa]

I'd expect something more along the lines of /sP/ > /Pʰ/, /P/ > /Pʼ/.

I second this. Chain shifts like /p ph/ > /p' p/ are much more common than /ph/>/p'/ while bypassing the plain /p/. However it gets tricky because you may want the change to be conditional so that the plain form can be the most common in both the proto and the daughter language. (I.e. you probbly won't want ejectives to be more common than plain stops, although some langs do that)

[Tropylium]

*sP > Pʼ is attested across Sino-Tibetan (I'd have to look up the exact details), but I believe the usual route there is something like *sP > *əsP > *əP > *ʔəP > *ʔP > Pʼ.

Word-medially a more likely mechanism might be *sP (> *Pʰː) > *Pː > *ʔP > Pʼ (and I guess an epenthetic schwa intermediate would allow word-initial cases to go through this too).

[Zaarin]

If you're set on /sP/ and willing to take the long way around, what Tropylium describes makes sense to me.

[Nortaneous]

*sP > Pʼ is attested across Sino-Tibetan (I'd have to look up the exact details), but I believe the usual route there is something like *sP > *əsP > *əP > *ʔəP > *ʔP > Pʼ.

Is it? Are the intermediate stages attested?

Korean tense consonants come from pC- sC- -ʔC- (at least in MK -rq.p- > -r.pp-). The intermediate stage in this case might have been gemination. Tangut had *sCV > CṾ, probably via a tense onset *C̣V; it isn't known how the Ṿ series was pronounced, but for whatever reason they're generally called 'tense vowels'.

[Dē Graut Bʉr]

Alternatively, you can have P > P^h followed by sP > P > P'.

[mèþru]

All of this because I didn't notice the apostrophe and that it was a question about simple elision.

[missals]

I would think you could get P' from sP extremely simply, just by sP > hP > PP > P'

I mean, I think hP > PP is plausible. It could go hP > ʔP > PP, and that would be plausible for sure, right?