Sound Change Quickie Thread

[sangi39]

However, it may depend on what other series the language had.

It's literally just voiceless vs. voiced at the moment. I'm just trying to figure out a way of getting from /p/ vs. /b/ > /p/ vs. /p'/ without relying on the usual "use the glottal stop" method but I'm not sure if I'm thinking that it's easier said than done

Then again, I could just not worry about it and have /p/ vs. /p'/ to begin with and leave it at that But that's no fun

[Basilius]

It's literally just voiceless vs. voiced at the moment. I'm just trying to figure out a way of getting from /p/ vs. /b/ > /p/ vs. /p'/ without relying on the usual "use the glottal stop" method but I'm not sure if I'm thinking that it's easier said than done

You may try something like:

(1) p :: b -> (2) pʰ :: b -> (3) pʰ :: p -> (4) pʰ :: pˀ

Some problems with it:

[*] (2) -> (3): it appears that /b/ used to be more marked than the other series, and now it turns into something that looks less marked than the aspirate;
[*] (3) - > (4): the other way round, the apparently less marked plain voiceless plosive turns into the more marked ejective.

This can be cured by additional sound changes altering the relative frequencies of the two series (may be easier if the language had a lot of clusters).

But besides, it can be questioned if "plain" is the less marked series at (3); for example, in languages of Caucasus it is common that aspirates are lenes and "plain voiceless" plosives are fortes (thus potentially more marked). However, those languages usually also have ejectives and plain voiced plosives, so the whole set of contrasts is quite different; and then there's a problem with /b/ becoming a *fortis* /p/ (contrasting with *lenis* /pʰ/) - it is hardly the case that the tenseness relations were the same at the precious stage.

[Hallow XIII]

This is extremely simple.

b d g > pʼ tʼ kʼ

Don't ask me why, but these two series often have each other as allophones and IIRC are documented to have shifted into one another multiple times, a change which by some people is also assumed to have happened in PIE (although I think ejective to voiced is more common, I am very sure I have seen the other way around already). Presumably it has something to do with implosives, but that does not matter here.

EDIT: In other words, your original proposal seems v. acceptable to me.

[Basilius]

<...> I am very sure I have seen the other way around already<...>

Made me curious. Don't hesitate to PM me when you recall where.

<...>a change which by some people is also assumed to have happened in PIE.<...>

I understand, for you this counts as attested? (Also, you could notice that I consider it relevant if the language had more than two plosive series.)

[Hallow XIII]

No, this does not count as attested. But it was a relevant little fact that popped into my mind.

[Nortaneous]

p sure i can have, and have heard, allophonic glottalization of lenis plosives in english

[KathTheDragon]

I have a vowel inventory of /ɪ i y y: u e o ɛ ɜ ʌ ɔ a ɐ ɑ/, and I need to mess around with the unrounded vowels. Are there any general sound changes I can use on them?

[finlay]

chain shift, vowel harmony, mergers, take your pick. it's kind of an open ended question.

why is y: the only long vowel though?

also ɜ, ʌ, ɐ are all pretty close together so i'd expect some kind of sound change to either merge them or push each other apart to make them more distinct – if they're less common sounds I'd expect the merger.

[KathTheDragon]

why is y: the only long vowel though?

There were once no contrasting long vowels, then /ɪy/ > /y:/ which contrasts with /y/. I do plan on creating more long-short distinctions.

[äreo]

I have a vowel inventory of /ɪ i y y: u e o ɛ ɜ ʌ ɔ a ɐ ɑ/, and I need to mess around with the unrounded vowels. Are there any general sound changes I can use on them?

general? well, uhh
ɪ i y y: u > ɪ iː ʏ yː uː
o > ʊ _C
o > oː
e ɛ ɜ > ɛ _C
e > eː
a ɑ > a _C
a > æː
ɑ > ɑː
ʌ ɐ > ʌ
ɜ ʌ > ʌː except _C
this leaves you with /a ɛ ɪ ʏ ʌ ɔ ʊ æː ɑː eː iː yː ʌː oː uː/ which looks rather like Estonian, tho I don't know your phonotactics so.

[KathTheDragon]

(C)V(C)

By 'long-short' distinctions, I meant pairs like /y y:/ where length is the sole distinguishing factor.

[Tropylium]

<...> I am very sure I have seen the other way around already<...>

Made me curious. Don't hesitate to PM me when you recall where.

Possibly relevant: Shona appears to have juggled the three proto-Bantu stop series around as thus:
*p *t *k → f ʀ h (the 2nd went thru an *r stage)
*mp *nt *ŋk → pʰ tʰ kx
*mb *nd *ŋg → pʼ tʼ kʼ
This looks rather like it went thru a *pʰ *tʰ *kʰ, *p *t *k, *b *d *g stage. After the lenition of series 1 and the aspiration of series 2, I suppose series 3 might have first become plain voiceless, then spontaneously acquired glottalization. I'm guessing though.

why is y: the only long vowel though?

There were once no contrasting long vowels, then /ɪy/ > /y:/ which contrasts with /y/.

Yeah no this kind of a thing does not happen. Among your close vowels, you have a tense : lax distinction among the front unrounded ones (/i ɪ/), a length distinction among the front rounded ones (/y yː/), and nothing among the back vowels (just /u/). I suggest starting by fixing this: y yː → ʏ y.

I also suppose that your contrasts like /e ɛ/, /o ɔ/ would be based on tenseness more so than height? "5-height" systems /i ɪ e ɛ a/ also don't really happen.

After that, here's a little vowel ballet in three stages:
1.1) The tense non-close vowel diphthongize: e o a ɑ → ie uo ɛa ɔɑ
1.2) Two of the lax central vowels become fully open: ɐ ʌ → a ɒ
Current system: tense /i y u ie uo ɛa ɔɑ/, lax /ɪ ʏ ɛ ɜ ɔ a ɒ/. Note the absense of an /yø/ to go with /y/, or /ʊ/ to go with /u/.
2.1) Fronting: uo → yø
2.2) Back vowel raising: ɔɑ ɔ ɒ → uo ʊ ɔ
2.3) Elimination of central vowels as a category: ɜ → ɑ
Current system: tense /i y u ie yø uo ɛa/, lax /ɪ ʏ ʊ ɛ ɔ a ɑ/. Rather less "holey", though /ɛa/ is kind of a loner.
3.1) Low vowels become tense: a ɑ → aː ɔɑ
3.2) Diphthong breaking: ie yø uo ɛa ɔɑ → ʲeː ʲoː ʷoː ʲaː ʷaː. Probably would also imply palatalization as i y → ʲi ʲy.
3.3) Lowering of lax vowels: ɪ ʏ ʊ → ʲe ʲo o. You could well add extra conditions on *ʏ, like → e / _Cʲ, or → o / _velar. Or keep it as ʲø.
3.4) Long vowel raising: eː aː oː → iː ai uː / _Cʲ; eː aː oː → yː au uː / _Cʷ. Just some ideas, but you'd need to introduce /iː yː uː/ somehow by this point.
3.5) Low mid merger: ɛ ɔ → a. Possibly under some conditions these could also become e o, or eː oː, etc.
3.6) Palatalization and labialization drop on some (all?) consonants. This might happen earlier, in which case you could block long vowel raising.
Resultant system: /i e a o u y/ plus length, /ai au/.

Example words:
*baka → *bɛakɛa → *bʲaːkʲaː → /baikaː/ (or maybe something like /baicaː/)
*sʌka → *sɒkɛa → *sɔkɛa → *sakʲaː → /sakaː/
*tɐsɜ → *tasɜ → *tasɑ → *taːsɔɑ → /tausaː/
*mɔlʏ → *mʊlʏ → *molʲo → /molo/
*xebɪ → *xiebɪ → *xʲeːbʲe → /xiːbe/
*pɛɾogi → *pɛɾuogi → *paɾʷoːgʲi → /paɾuːgi/

[KathTheDragon]

Ok, lemme just say that this is derived from English, and that ɪy > y: is one of the latest sound changes amongst the vowels before this snapshot. It is, however, accompanied by ɪ: > i and ɛ: > e, which I feel should let me change whatever English really does into short/long without much pain. While I mess around with the unrounded vowels, mind. I just want ideas as to what I could plausibly do to it.

[finlay]

English (at least BrE/RP) is usually written as ɪ i:, etc, with a long/short pair which also vary by their tongue position. Mostly this is to facilitate EFL. Writing them as ɪ i or i i: is also valid, but it's worth considering that if you've got those pairs you can write them that way instead.

[Basilius]

<...> I am very sure I have seen the other way around already<...>

Made me curious. Don't hesitate to PM me when you recall where.

Possibly relevant: Shona appears to have juggled the three proto-Bantu stop series around as thus:
*p *t *k → f ʀ h (the 2nd went thru an *r stage)
*mp *nt *ŋk → pʰ tʰ kx
*mb *nd *ŋg → pʼ tʼ kʼ
This looks rather like it went thru a *pʰ *tʰ *kʰ, *p *t *k, *b *d *g stage. After the lenition of series 1 and the aspiration of series 2, I suppose series 3 might have first become plain voiceless, then spontaneously acquired glottalization. I'm guessing though.

Thank you for the example! It's certainly relevant, and your guess looks very plausible.

Also, it's similar to what I proposed above:

You may try something like:

(1) p :: b -> (2) pʰ :: b -> (3) pʰ :: p -> (4) pʰ :: pˀ

Some problems with it:

[*] (2) -> (3): it appears that /b/ used to be more marked than the other series, and now it turns into something that looks less marked than the aspirate;
[*] (3) - > (4): the other way round, the apparently less marked plain voiceless plosive turns into the more marked ejective. <...>

<...> it can be questioned if "plain" is the less marked series at (3); for example, in languages of Caucasus it is common that aspirates are lenes and "plain voiceless" plosives are fortes (thus potentially more marked). However, those languages usually also have ejectives and plain voiced plosives, so the whole set of contrasts is quite different; and then there's a problem with /b/ becoming a *fortis* /p/ (contrasting with *lenis* /pʰ/) - it is hardly the case that the tenseness relations were the same at the precious stage.

What looked problematic was the evolution of markedness relations; it is interesting what it was like in Shona at the stage immediately preceding the glottalization (and, among other things, if the would-be glottalized series could be fortis at that moment).

EDIT: looking at the Shona/Sesotho inventory some more, I recall reading somewhere that prenasalized *voiceless* plosives in a couple Bantu languages can be allophonically realized as voiceless nasals; besides, the immediate result of denasalization with the *voiced* prenasalized series could be implosive rather than plain voiced.

That is, an intermediate stage could look like, for example, *pʰ *tʰ *kʰ, *m̥ *n̥ *ŋ̥, *ɓ *ɗ *ɠ, where the implosive series is the only voiced one and looks kinda doubly marked (voiced + glottalized), perhaps wanting a shift to ejective as the first subsequent change.

This is irrelevant for the plausibility of the voiced -> ejective change in general, but may be important for assessing the scenario which sangi39 is considering.

[KathTheDragon]

English (at least BrE/RP) is usually written as ɪ i:, etc, with a long/short pair which also vary by their tongue position. Mostly this is to facilitate EFL. Writing them as ɪ i or i i: is also valid, but it's worth considering that if you've got those pairs you can write them that way instead.

Ok, my specific dialect of English, and I generally don't have long vowels, apart from ɪ: and ʊ:, which develop from ɪə and ʊə. Anyway, I've done what I've done, and I'm not going to change it 'cos you don't like there being a long/short variation amid a mass of tense-long/lax-short variations.

Edit: I fudged one of the changes in the consonants, and I have created a mass of diphthongs ending in /ɯ/ as a result. Most of these should drift to other diphthongs, but there'll be a few left over (/ʌɯ/ for example) that I can collapse into long vowels.

[Basilius]

Anyway, I've done what I've done, and I'm not going to change it 'cos you don't like there being a long/short variation amid a mass of tense-long/lax-short variations.

Why asking about plausible SC's etc., then? Just keep doing what you're doing. Call it creativity, too.

But if you're still interested in plausible things that may happen to your vowel inventory, do consider either eliminating that isolated length contrast or obtaining more contrasting pairs.

[KathTheDragon]

I want more contrasting pairs. I said so. The problem is (and why I came to this thread) is that I have run out of good ideas.

[KathTheDragon]

I've eliminated /ʌ ɪ/, and added /ɔ: o: u: i: ɯ ɯ:/. Feeling better?

[Genome]

Yeah.
Does β ɣ seem realistic?

[äreo]

Yeah.
Does β ɣ seem realistic?

β > w > ɰ > ɣ

[Click]

Does β → ɣ seem realistic?

It does.

[Nortaneous]

When consonants are prenasalised, their pronunciation and spelling may change. Murmur no longer shifts to the following vowel. Fricatives become affricates, and if voiceless, become ejectives as well, at least with some speakers: mf is pronounced [ɱp̪fʼ]; ndl is pronounced [ndɮ];n+hl becomes ntl [ntɬʼ]; n+z becomes ndz [ndz], etc. The orthographic b in mb is a voiced plosive, [mb].

[Dē Graut Bʉr]

Could anyone explain how /z/ possibly can have become /k’/ in one of my conlangs?

[Drydic]

You screwed up? That's a pretty unlikely sound change, just by itself. Also how big of a time period are you thinking for it?