Sound Change Quickie Thread

[Tropylium]

Having a common five vowel systems, what can I do with vowels that involves stress? Is there anything besides laxing unstressed vowels, or tensing stressed vowels?

Well, you can of course start by laxing, then do all sorts of stuff with the laxness contrast. Just look at the English GVS. Also note that there is no obligation to phonemicize the laxness contrast! You can well rotate things around a bit while keeping them within a 5-phoneme set. Say, unstressed *i *u > *ɪ *ʊ > *ʲə *ʷə > /ja wa/; then unstressed *e *o > /i u/. Perhaps *a can further re-develop to unstressed /e o/ in some positions.

What are the typical conditions for palatalisation, labialisation, and pharyngealisation to happen?

Adjacency to something that is palatal, labial, or pharyngeal, simply put.

Note also that these are in no way limited to "a vowel causes a consonant change". *lɟ > ʎɟ would also be palatalization; *wa > wɔ (as in English) would also be labialization; *iCɑ > ɪ̙Cɑ would also be pharyngealization. (Though when limited to vowels only, these are usually called "fronting", "rounding" and "retraction".)

[Qwynegold]

Umlaut most frequently involves features spreading from unstressed vowels to stressed vowels, and is often associated with vowel reduction and syncope.

Hmm? I thought it just involved features spreading to adjacent vowels, regardless of stress.

Impressionistically, umlaut tends to happen in languages with strong (and often initial) stress though. Old Irish, Germanic, Chechen, and it appears a similar correlation was mentioned by Tropylium over in this thread concerning Finnic. I wouldn't be too surprised to learn if it existed in a language with weak or no stress, but I think it's safe to say they would be in the minority.

Another option might be breaking, stressed vowels gain onglides as Latin e/o did in Spanish.

Hmm... So it's only a hypothesis? But interesting nonetheless. Thanks!

[Qwynegold]

Having a common five vowel systems, what can I do with vowels that involves stress? Is there anything besides laxing unstressed vowels, or tensing stressed vowels?

Well, you can of course start by laxing, then do all sorts of stuff with the laxness contrast. Just look at the English GVS. Also note that there is no obligation to phonemicize the laxness contrast! You can well rotate things around a bit while keeping them within a 5-phoneme set. Say, unstressed *i *u > *ɪ *ʊ > *ʲə *ʷə > /ja wa/; then unstressed *e *o > /i u/. Perhaps *a can further re-develop to unstressed /e o/ in some positions.

Thanks! I'll think about this too.

[8Deer]

What are some plausible ways, preferably conditional, to have codas of /k kʰ ɡ ŋ t tʰ d n p pʰ b m s r/ alter the vowel height or backness of /a aː i iː u uː/? The vowels /i iː/ are occuring after /tʃ tʃʰ dʒ nʒ̃/ with too high a frequency to my liking and I want to decrease the relative frequency of those two vowels occuring after affricates relative to other vowels, especially back vowels.

A few Austronesian languages do funky things to vowels before coda /k ŋ/ including diphthongization (i u > iə uə), lowering (i u > e o) or a combination of both (i u > eə oə).

[Qwynegold]

Is it plausible to have allophonic nasalization of vowels before /n/, but not before /m/?

[StrangerCoug]

I don't think so. Normally that happens regardless of the place of articulation of the nasal consonant.

[Qwynegold]

I don't think so. Normally that happens regardless of the place of articulation of the nasal consonant.

OK, thanks! I'll change it then.

New question! I have [ʰp ʰt ʰc ʰk] (as allophones of /h/ + plosive). What can I do with preaspiration? I don't mind keeping it, since it's only allophonic, but I need to get rid of [ʰc] at least. That's because I'm trying to get rid of /c/, at least as a phoneme. At one point [c] becomes [ɟ] in intervocalic position. What can I do with [ʰc] in that environment?

[sirdanilot]

I don't think so. Normally that happens regardless of the place of articulation of the nasal consonant.

OK, thanks! I'll change it then.

New question! I have [ʰp ʰt ʰc ʰk] (as allophones of /h/ + plosive). What can I do with preaspiration? I don't mind keeping it, since it's only allophonic, but I need to get rid of [ʰc] at least. That's because I'm trying to get rid of /c/, at least as a phoneme. At one point [c] becomes [ɟ] in intervocalic position. What can I do with [ʰc] in that environment?

Urm, what? What is even the difference between a h stop cluster and a preaspirated stop (never even heard of that before) phonetically? Sure, in your PHONOLOGY it can mean something different, but phonetic realization is also a thing.

One thing I could come up with was a change to a palatal fricative ç but I'd thin you'd want lenition like this for all your stops then.

[Qwynegold]

I don't think so. Normally that happens regardless of the place of articulation of the nasal consonant.

OK, thanks! I'll change it then.

New question! I have [ʰp ʰt ʰc ʰk] (as allophones of /h/ + plosive). What can I do with preaspiration? I don't mind keeping it, since it's only allophonic, but I need to get rid of [ʰc] at least. That's because I'm trying to get rid of /c/, at least as a phoneme. At one point [c] becomes [ɟ] in intervocalic position. What can I do with [ʰc] in that environment?

Urm, what? What is even the difference between a h stop cluster and a preaspirated stop (never even heard of that before) phonetically? Sure, in your PHONOLOGY it can mean something different, but phonetic realization is also a thing.

One thing I could come up with was a change to a palatal fricative ç but I'd thin you'd want lenition like this for all your stops then.

http://en.wikipedia.org/wiki/Preaspiration
So it is a bit unclear. :/ But see the last subheading in that article (called h-clusters).

Hmm, I don't actually need to do anything to [ʰc] until that one sound change that affects some intervocalic consonants, including [c]. But I think it should become something voiced. Maybe [ʝ]?, and then [ʒ] because [ʒ] already exists in the language.

[sirdanilot]

It is perfectly possible to have both the j thing and ʒ as separate phonemes, why get rid of the former? In fact I think a merger would be a bit odd since ʒ is a sibilant nd the voiced palatal fricative is not a sibilant.

Preaspiration is comparatively uncommon across languages of the world,[4] and is claimed by some to not be phonemically contrastive in any language.[5] Ladefoged and Maddieson[3] note that, at least in the case of Icelandic, preaspirated stops have a longer duration of aspiration than normally aspirated (post-aspirated) stops, comparable to clusters of [h]+consonant in languages with such clusters. As a result, they view preaspiration as purely a distributional feature, indistinguishable phonetically and phonologically from clusters with /h/, and prefer to notate preaspirated stops as clusters, e.g. Icelandic kappi /ˈkʰahpi/ "hero" rather than /ˈkʰaʰpi/.

In short preaspiration basically does not exist. It's simply a cluster of h and consonant, often arising from fricative and consonant clusters. I do not think your voicing thing is a good idea. Lenition (p - f, t - s etc.) seems like a much more realistic idea for your sound change. Your pesky palatal stop would then become ç (and maybe x around back vowels, which is also a cool merger ey !)

But hey it's your lang so whatever floats your boat

[Nortaneous]

ʰc ʰk > hp ht çc xk
çc > ʃtʃ > ʃ:

if you don't have contrastive gemination elsewhere then ʃ: > ʃ, then intervocalic lenition

also
ʰp ʰt > xp xt
or
ʰp ʰt > xp st
or
ʰp ʰt > xp xt > fp xt
or just
ʰp ʰt > hp ht

[Hallow XIII]

Preaspiration is exactly what it says on the tin: a stop preceded by an interval of devoicing. As for "it's just a hC cluster", well, so is aspiration, really.

As for things you can do with it: Scottish Gaelic, at least some dialects of it, basically turned C_h$ into xC$. So, for instance, breacan [ˈpɾɛxk.ən] but breagan [ˑpɾɛk.ən] [ʰc] becomes [çc]: mhic [viçc].

If you do that then you can have xp xt çc xk > fp xt ʃt xk or you can do a Donegal Irish and > wp ɹt ʃt/jt wk

[Nortaneous]

or do a middle english and have /x/ diphthongize preceding vowels

[Click]

Is [œ ø ʏ y] backing to [ɔ o ʊ u], perhaps via [ʌ ɤ ɯ̽ ɯ] plausible? How else to get rid of front rounded vowels without simply unrounding them or them leaving semivowels everywhere?

[Hallow XIII]

Short answer, if you want to do anything interesting with that you've got to mix it up. Mergers, diphthongizations, you can do all sorts of things. Simple unconditional changes are usually not the best way to do it, though.

[WeepingElf]

Is [œ ø ʏ y] backing to [ɔ o ʊ u], perhaps via [ʌ ɤ ɯ̽ ɯ] plausible? How else to get rid of front rounded vowels without simply unrounding them or them leaving semivowels everywhere?

I can't think of a natlang precedent - the usual fate of front rounded vowels seems to be unrounding - but I think what you are thinking of should work.

[Nortaneous]

One of my conlangs had y œ > ɯ ɤ. No idea if it's attested, but acoustic similarity.

Mongolian is usually said to have had y ø u o > u o ʊ ɔ but I've seen that be contested.

[Clearsand]

I want to know about options for spontaneous word initial consonants after vowels. I know about the glo'al stop, but I wonder if an initial velar nasal or dark L has ever been attested. Perhaps other nasals as well?

[Hallow XIII]

I read in one place that some variety of cantonese redistributes initial /0 ŋ/ so that the former occurs with high tone and the latter with low tone initials. It looks a bit suspect, though; usually the quoted sound change is that ŋ- is simply deleted.

[----]

I want to know about options for spontaneous word initial consonants after vowels. I know about the glo'al stop, but I wonder if an initial velar nasal or dark L has ever been attested. Perhaps other nasals as well?

0 > velar nasal is attested in Palauan and one or two Samoyedic languages. Arapaho and Cheyenne have 0 > h word initially.
A couple scattered languages will spontaneously gain word initial glides before certain vowels. You could do something like 0 > j /#_{a, e, i} and get away with it.

[vokzhen]

I want to know about options for spontaneous word initial consonants after vowels. I know about the glo'al stop, but I wonder if an initial velar nasal or dark L has ever been attested. Perhaps other nasals as well?

Presumably you mean before vowels? The ones I've heard are various glides and /ŋ/, which seems likely to have come from a glottal stop with rhinoglottophilia. Initial glides could be fortified to make it appear as though word once took an initial /dʒ b/ or /gʲ g gʷ/ or similar, maybe even have a brief period of time where they genuinely are inserted in, though I kind of doubt glide-insertion and glide-fortition would be operating at the same time. You could probably get the insertion of nasals in there if your glides are nasalized or fully nasal stops before nasal vowels, as some South American languages have (i.e. /j/ is [j ~ ɲ] depending on whether the vowel is nasal or not).

There is h-insertion in some English speech, but it's pretty clearly triggered from hypercorrection. If you weren't needing a complete sound change, that might means you could get things like /q/ or /l/ hypercorrected from /ʔ/ or /w/.

[sirdanilot]

As far as creating things out of thin air is concerned, some West-Flemish (and West-Zeelandic-Flemish) dialects of dutch switch around what are in standard Dutch the phonemes /h/ and /x/ to effectively /x/ and /h/. However the change was a bit more complicated than that, not just 'hey guys let's switch around these phonemes'.

I had a Math teacher who would say [ˈheo drixuk] for normal [ˈxeo drihuk] 'geo-driehoek', so even for loanwords it's productive. He was from [ˈmɑldəhəm] (which in standard Dutch would be [ˈmɑldəxɛm]. (at least that was his surname and Maldegem is in that region and he had a matching accent, so...) A geo-driehoek is one of these things.

[gach]

0 > velar nasal is attested in Palauan and one or two Samoyedic languages.

I think Nganasan has Ø > ŋ / #_V and Nenets slightly more complex Ø > ŋ / #_V[back] and Ø > ɲ / #_V[front]. Not so sure about Enets. I'm now about 4000 km from home so someone with more flexible access to literature can perhaps check this.

[jmcd]

Is [œ ø ʏ y] backing to [ɔ o ʊ u], perhaps via [ʌ ɤ ɯ̽ ɯ] plausible? How else to get rid of front rounded vowels without simply unrounding them or them leaving semivowels everywhere?

I think [œ ø ʏ y] > [ɔ o ʊ u] (> [ʌ ɤ ɯ̽ ɯ]) would be more reasonable. Or diphthongisationas Hallow suggests.

[Qwynegold]

Oh, sorry. I realized that the distribution of phonemes in the conlang is such that preaspirates other than [ʰp, ʰc] never happen.

It is perfectly possible to have both the j thing and ʒ as separate phonemes, why get rid of the former? In fact I think a merger would be a bit odd since ʒ is a sibilant nd the voiced palatal fricative is not a sibilant.

Eh? Surely [j] has turned into [ʒ] in a bunch of languages, and that one's not even a fricative.