Sound Change Quickie Thread

[StrangerCoug]

I'm good with that.

[CatDoom]

I can't point to a specific example at the moment, but liquids seem to be quite susceptible to elision, usually more so than nasals. Compensatory lengthening is outrageously common, and plenty of languages allow long vowels before coda consonants. It seems like a totally reasonable change to me.

Edit: A little poking around online indicates that the answer to your question might actually depend on what kind of r the language has. According to Compensatory Lengthening: Phonetics, Phonology, Diachrony by Darya Kavitskaya et al, trills aren't usually associated with a long transition after the vowel, as approximants are, and therefore wouldn't tend to affect vowel length when elided. That said, even the writer's own examples seem to put this assertion on shaky ground. For instance, he explains loss of final r with compensatory lengthening in western Anatolian Turkish dialects by asserting that rs in that region are "more approximant-like at least in some dialects," based on a personal communication. So... YMMV.

[jmcd]

How reasonable is Vrn > V:n?

I'm messing about with a non-rhotic (or less-rhotic) variety of Tirina and am trying to think of interesting things to do with the vowels; I don't want to just drop the /r/ and be done with it.

Have it turn into a diphthong like in Modern German.

[alynnidalar]

I can't point to a specific example at the moment, but liquids seem to be quite susceptible to elision, usually more so than nasals. Compensatory lengthening is outrageously common, and plenty of languages allow long vowels before coda consonants. It seems like a totally reasonable change to me.

Edit: A little poking around online indicates that the answer to your question might actually depend on what kind of r the language has. According to Compensatory Lengthening: Phonetics, Phonology, Diachrony by Darya Kavitskaya et al, trills aren't usually associated with a long transition after the vowel, as approximants are, and therefore wouldn't tend to affect vowel length when elided. That said, even the writer's own examples seem to put this assertion on shaky ground. For instance, he explains loss of final r with compensatory lengthening in western Anatolian Turkish dialects by asserting that rs in that region are "more approximant-like at least in some dialects," based on a personal communication. So... YMMV.

Interesting, thanks for the reference. It is trill-ish, but when I pronounce it I feel that it comes out "more approximant-like", so that fits the bill.

Have it turn into a diphthong like in Modern German.

I'll take a look at those sound changes, thanks for the suggestion! The "standard"/Elten dialect of Tirina doesn't have any diphthongs, but it could be a way to get them into a different dialect, whether it's the one I'm working on now or another.

[احمکي ارش-ھجن]

Could sound changes be restricted to word classes?

In example, i/y/_B in affixes only and where B=bilabial

[WeepingElf]

Could sound changes be restricted to word classes?

In example, i/y/_B in affixes only and where B=bilabial

Most historical linguists opine, AFAIK, that sound change is inattentive to morphological issues. Where it seems as if a sound change affected only certain classes of morphemes and others not, a morphological explanation is usually called for.

[hwhatting]

Could sound changes be restricted to word classes?

In example, i/y/_B in affixes only and where B=bilabial

Most historical linguists opine, AFAIK, that sound change is inattentive to morphological issues. Where it seems as if a sound change affected only certain classes of morphemes and others not, a morphological explanation is usually called for.

I think that holds true as a general rule. But I remember seeing descriptions of phonology where morpheme boundaries are conditioning factors; under the right circumstances, that may lead to word-class dependent change, if e.g. certain morpheme boundaries are limited to certain word classes.

[احمکي ارش-ھجن]

I mean like affix-only sound changes, like the example abovd.

[StrangerCoug]

I have what's supposed to be the protolang for a large language family with an inventory of /a aː e eː i iː o oː u uː p b̥ b̬ m t d̥ d̬ n s z̥ z̬ z̃ k ɡ̊ ɡ̌ ŋ/ and a syllable structure of (C)V(C). Coda consonants are permitted only after a short vowel and are limited to /p t s k/ there from a morphophonemic standpoint, but there are no other phonotactic restrictions right now. What are some interesting ways to play with the voiceless/slack/stiff contrast? (And what's the most plausible default voicing for the nasals in this kind of system?)

Anyone have any ideas? *taps foot impatiently*

[Travis B.]

I have what's supposed to be the protolang for a large language family with an inventory of /a aː e eː i iː o oː u uː p b̥ b̬ m t d̥ d̬ n s z̥ z̬ z̃ k ɡ̊ ɡ̌ ŋ/ and a syllable structure of (C)V(C). Coda consonants are permitted only after a short vowel and are limited to /p t s k/ there from a morphophonemic standpoint, but there are no other phonotactic restrictions right now. What are some interesting ways to play with the voiceless/slack/stiff contrast? (And what's the most plausible default voicing for the nasals in this kind of system?)

Anyone have any ideas? *taps foot impatiently*

Nasals I would assume would have stiff voice. Another thing to do is to have stiff voiced obstruents become slack voiced next to voiceless obstruents. Yet another thing would be to make the stiff voiced plosives become stiff voiced fricatives, where if combined with the previous thing would become slack voiced fricatives next to voiceless obstruents..

Also, is it intentional that you are omitting all approximants altogether?

[vokzhen]

I'd say nasals will probably belong to the least-marked category phonologically, which at a guess would be slack voice (an original modal voice weakened to slack voice, allowing room for or being pushed there by a more recent stiff voice), but it depends on how they came about, and either way they may be phonetically just modally voiced except in clusters. Over time as the other sounds shift, they may be realigned with a different set. Of course, depending on what rules you have in play, unclustered nasals might just be plain-voiced and not align either way.

Slack/stiff could become aspirate/voiced, or the plain set could aspirate making way for the stiff set to become plain. Either could become vocalic properties, and possibly then shift onto nearby consonants (plain-vowel-stiff > plain-creaky.vowel-plain > glottalized-vowel-plain), become tonal/register tone qualities, or cause vowel quality changes. Stiff voice could strengthen to a glottal stop that has further effects (tone, vowel length, rhinoglottophilia). Stiff could probably become nasalized the way implosives are wont to do, but I'm not sure.

The outcomes are likely position-dependent, compare Punjabi breathy voiced stops (stem-initial > plain, low tone on syllable; stem-final > voiced, high tone on previous two syllables; stem-medial between a short and long vowel > voiced, low tone on two following syllables), or Korean (original voiced set is aspirated initially and in clusters with /h/, voiced between sonorants, and voiceless elsewhere [including after a voiced stop, which becomes voiceless and unreleased], nasals are denasalized initially, and initial /l/ [n]).

[WeepingElf]

Could sound changes be restricted to word classes?

In example, i/y/_B in affixes only and where B=bilabial

Most historical linguists opine, AFAIK, that sound change is inattentive to morphological issues. Where it seems as if a sound change affected only certain classes of morphemes and others not, a morphological explanation is usually called for.

I think that holds true as a general rule. But I remember seeing descriptions of phonology where morpheme boundaries are conditioning factors; under the right circumstances, that may lead to word-class dependent change, if e.g. certain morpheme boundaries are limited to certain word classes.

I hit upon what seemed to be a morphology-sensitive sound change in my main conlang, Old Albic. Apparently, final vowels were lost everywhere except in the agentive case of animate nouns, and in the personal endings of the verbs. I first attempted to find a solution involving some weak consonant that was lost, but it did not work well. Later, I found a morphological solution. The forms which retained their final vowel were exactly those where the final vowel indicated number (in the other cases of the animate noun, a case ending followed that vowel, and inanimate nouns, where final vowels were lost, did not inflect for number in the early stage involved here), so this vowel was not just a final vowel but a morpheme in itself, and was retained.

[StrangerCoug]

Also, is it intentional that you are omitting all approximants altogether?

Yes.

[Max1461]

As crazy as this sounds, is /f/ > /sʰ/ plausible?

[----]

f > s has happened before but I don't know where the aspiration is coming from

[Max1461]

Analogy with other voiceless fricatives becoming aspirated.

[----]

Seems totally fair then.

[احمکي ارش-ھجن]

Could bilabial consonants cause labialization of some consonants and the round and backing of front vowel if they precede the bilabial consonants only?

[Travis B.]

Could bilabial consonants cause labialization of some consonants and the round and backing of front vowel if they precede the bilabial consonants only?

Typically I would assume not, as bilabials are typically unlabialized and have no velar or uvular component. To get the effect you want what you probably want is a rounded labiovelar.

[احمکي ارش-ھجن]

Could bilabial consonants cause labialization of some consonants and the round and backing of front vowel if they precede the bilabial consonants only?

Typically I would assume not, as bilabials are typically unlabialized and have no velar or uvular component. To get the effect you want what you probably want is a rounded labiovelar.

But I recalled somewhere that bilabials can cause preceding consonants to become labialized
The vowels however are an uncertainty, I assume they'd become rounded front vowels and then eventually start backing.

A rounded labiovelar is redundant, as anything involving the lips is necessarily labialized/rounded

[Zju]

I can't point to a specific example at the moment, but liquids seem to be quite susceptible to elision, usually more so than nasals. Compensatory lengthening is outrageously common,

English?

[vokzhen]

Those are pretty basic terms, and a quick google/wikipedia search would give you the meaning. r/l (liquids) disappear (elide), often to vowel or consonant length (compensatory lengthening), more often than m/n/ng (nasals) disappear (elide).

[----]

lol they were providing English as an example of liquid elision

[Travis B.]

Could bilabial consonants cause labialization of some consonants and the round and backing of front vowel if they precede the bilabial consonants only?

Typically I would assume not, as bilabials are typically unlabialized and have no velar or uvular component. To get the effect you want what you probably want is a rounded labiovelar.

But I recalled somewhere that bilabials can cause preceding consonants to become labialized
The vowels however are an uncertainty, I assume they'd become rounded front vowels and then eventually start backing.

I somehow suspect that bilabials that would cause preceding consonants to become labialized are themselves phonetically rounded, but this is not typically marked phonemically unless a contrast between unrounded and rounded labials exists.

Also, bilabials seem to be neutral with regard to adjacent vowels frontness/backness.

A rounded labiovelar is redundant, as anything involving the lips is necessarily labialized/rounded

No, since there are [kp gb]. (I cannot write the tie diacritic right now but these are single consonants not consonant pairs.) Also note that there exist contrasts between unrounded and rounded labials in some languages.

[احمکي ارش-ھجن]

I somehow suspect that bilabials that would cause preceding consonants to become labialized are themselves phonetically rounded, but this is not typically marked phonemically unless a contrast between unrounded and rounded labials exists.

I need to find a way to develop /Tw Dw kw gw xw Gw/ without /w/ and back vowels, because my ancestor lang does not have enough of the latter and too much of the former
Besides, labialization from following bilabials or such would justify why potential labiovelar-bilabial clusters get metathesized away.

Also, bilabials seem to be neutral with regard to adjacent vowels frontness/backness.

Clearly, but that doesn't mean the rounded front vowels won't back themselves; rounded front vowels already tend to be more back than their unrounded counterparts.

A rounded labiovelar is redundant, as anything involving the lips is necessarily labialized/rounded

No, since there are [kp gb]. (I cannot write the tie diacritic right now but these are single consonants not consonant pairs.) Also note that there exist contrasts between unrounded and rounded labials in some languages.

Well, [kp gb] are *labial-velar* not labiovelar, very different.