Sound Change Quickie Thread

[Nate]

*p' → b pː w /_

I don't understand the thought behind turning voiceless plosives into approximants. :/

One of my iffies. /p'/ isn't a voiceless plosive, but an ejective. Is *p' → *p → w not possible? Both are bilabial at least. I probably should have said so, but C is a younger language than A or B, or at least attested in writing more recently. It's gone through a bit more development than the other two.

Whoops, I didn't pay enough attention. Well, as a one step change it would've still been odd.

Yeah, C (now named Abeda) sound changes are sometimes more than one step. For p', I think the best way to explain p' → w is p' → ph → w. Or perhaps the ph becomes f, then w?

*t' → d tː h /_

/h/? What?

Now that I think about it, that doesn't make much sense. I wanted the ejectives in C to turn into something much more different than the other two. What would be better reflexes of them?

Hmm, the ones I know about are voiceless plosive, voiced plosive, voiceless geminate plosive. :/

I'm aware as well (now). The steps for s' → h are s' → sh → h. An improvement?

Show us some sample words and how they'd evolve. It would be interesting to see the sound correspondences between the three languages. B seems conservative.

Yeah, I made sure that B seems more rigid. I guess the best way I can compare them to natlang history, is that A is Spanish, B is Italian (don't quote me on that; I don't speak enough Italian to be 100% sure), ad C is French.

Example words would be useful. After I transfered the sound changes to a forum-friendly format (they were made in a Word document table. maybe I'll use that next time), I got too lazy to do so.

For reference, here's the basic phonology of the proto-lang. Feel free to rip into as well:

Code: Select all

Consonants

Labials: [p b p']
Dentals: [t d t']
Velars: [k g k']

Fricatives: [s h s']

Nasals: [n m]

Liquids: [r l]

Semivowels: [j w]

Vowels
Pure: [a e i u aː eː iː uː]
Diphthongs: [ja je ju wa we wi ai ai ei ui au eu iu] and long equivalents, id est [jː aːu], etc.
Syllabic consonants: r̥, l̥, m̥, n̥

Here's some examples I've whipped up.

*bwir, 'person'
pīr bwir byr

*kjut, 'man'
hīs kjut çus

*s'aut, 'woman'
sāt sːaut ʃaus

*k'meil, 'body'
gəmīl kemēl imel

*gelt, 'tear'
kels get get

*m̥gāk-r̥, 'tongue'
mukāra emgaker ugaxa

*t'mjak, 'blood'
dəmē temjak miaç

*dāp-m̥, 'chin'
tafum dāpem dau

*gub-r̥, 'leg'
kopra guber guva

*makt-m̥, 'tooth'
māsum maktem maxt

*sag-m̥, 'bone'
sakum sagem saʁ

*k'nup-m̥, 'ear'
gənofum kenupem inū


That's as much as I can muster at the moment, but I may add later.

Okay. It seems like B changes mostly when the protolang had clusters. And it looks like a middle way between A and C. Btw, those syllabic consonants were not supposed to be voiceless, right?

I've since made some changes to the sound changes, and added a bit more, so the above are probably out of date. B (Geder) has changed a tad in that t → tɕ before a front vowel (d → dʑ as well) and a thing or two more. I feel it's far too close to Proto-Anrulin, so I may end up changing it much more.

Uh... *looks at IPA* Ah. Looks like I had IPA and PIE orthography mixed up. They are syllabic consonants, not voiceless voiced consonants. I'll change that, then.

After I finish up the basics of the PAn nouns, I plan to make a thread of my own for the Anrulin family. I'll try to include sound changes as well

Request, though. For my notes and stuff, I'm using a sound change notation method of my own, since I'm not 100% on how it's done (and what I DO know about it seems way too long and takes up too much space). Does anyone have a link that describes how sound change notation works in text format? Or is it possible for me to upload my stuff with a guide to my own method (it's not complicated or long, just different a tad)?

[cybrxkhan]

How plausible is an affricate aspirated consonant? (i.e. t͡ʃ > tʰ, t͡ʃ > kʰ, etc.)

I can't think of a natlang precedent offhand, but based on the fact that all fricatives can easily debuccalise to [h], shifts like p͡f > pʰ, t͡s > tʰ, and k͡x > kʰ are fairly plausible, and also t͡ʃ > tʰ or cʰ. But a change of affricate > fricative seems more common, especially for non-coronal affricates; note e.g. that German /p͡f/ has merged with /f/ in most northern variants except when preceded by a vowel. Also, t͡ʃ > kʰ is not very plausible; you'd expect a reflex of *t͡ʃ to end up further forward in the mouth (see above).

Thank you!

[----]

Vietnamese had ʂ > tʰ and s > t in Sinitic words, which I would say is a pretty similar set of changes. It definitely makes the idea plausible.

[Vuvuzela]

Would n/l/_S make sense, where "S" represents a stop?

[cerealbox]

sorry if this has been asked before, but how do you derive complex phonotactics from a simple open-syllable structre without being too contrived?

[Herr Dunkel]

sorry if this has been asked before, but how do you derive complex phonotactics from a simple open-syllable structre without being too contrived?

Vowel contraction and umlaut.

Good example: Proto-Germanic > Any Gmc. lang other than Gothic

(PGmc. *habukaz > ONrs. haukr (three syllables into one) )

Edit: Also PSl. (protoslavic) > Russian - remember PSl was open-syllable

[Cedh]

Would n/l/_S make sense, where "S" represents a stop?

Yes.

[Melteor]

f>ɸ
v>p

[Whimemsz]

You probably need some intermediates for the second change (v > b > p works).

sorry if this has been asked before, but how do you derive complex phonotactics from a simple open-syllable structre without being too contrived?

Lose unstressed vowels. Voila: lots of consonant clusters. Or lose some intervocalic consonants: voila, V+V sequences! (which can be collapsed into diphthongs or whatever -- or even affect neighboring consonants and disappear). Or lose final vowels. And combine some of these together, e.g.: *tibanáhuto > tianáhuto > tjanáhto > tʃanaht, or *kutáti > ktatʃ.

[cerealbox]

You probably need some intermediates for the second change (v > b > p works).

sorry if this has been asked before, but how do you derive complex phonotactics from a simple open-syllable structre without being too contrived?

Lose unstressed vowels. Voila: lots of consonant clusters. Or lose some intervocalic consonants: voila, V+V sequences! (which can be collapsed into diphthongs or whatever -- or even affect neighboring consonants and disappear). Or lose final vowels. And combine some of these together, e.g.: *tibanáhuto > tianáhuto > tjanáhto > tʃanaht, or *kutáti > ktatʃ.

my proto language doesn't have stress. Also am I supposed to say something like CVC structure reduce to CC when CC is the following: sp, sk, pr, etc...? Seems a bit contrived to me. Losing final vowels I've tried, but it only gets me so far.

[Whimemsz]

That's not necessarily contrived. Spoken Latin lost (unstressed) /u/ between certain consonants only (such as /k/ and /l/; so Classical Latin oculus "eye" was pronounced something like *[oklos] in spoken Latin). You can always have your proto-lang's system change to one of stress at some point in its evolution (this happened in the history of Latin and most other IE languages as well, since PIE was based on pitch-accent rather than on stress per se). Even with no stress system, you can have loss of some intervocalic consonants and then the conversion of some of the resulting vowel sequences to sequences of vowel and glide (so, voiced stops are lost between vowels and /ia/ becomes /ja/, etc., as in my examples in my previous post)

[Solarius]

sorry if this has been asked before, but how do you derive complex phonotactics from a simple open-syllable structre without being too contrived?

You could always devoice and then drop vowels between voiceless consonants, like Oogami.

[Nortaneous]

sorry if this has been asked before, but how do you derive complex phonotactics from a simple open-syllable structre without being too contrived?

You could always devoice and then drop vowels between voiceless consonants, like Oogami.

That's not quite what Oogami did, is it? I thought it was just i u > s f.

[Solarius]

sorry if this has been asked before, but how do you derive complex phonotactics from a simple open-syllable structre without being too contrived?

You could always devoice and then drop vowels between voiceless consonants, like Oogami.

That's not quite what Oogami did, is it? I thought it was just i u > s f.

Oh, is it? I guess I just assumed it did that. Devoicing and then dropping is still a pretty good Idea though.

[Nate]

Are the following sound changes realistic?

t /_[e i] > tj > tç > tʃ
d /_[e i] > dj >dz > z > s

[Whimemsz]

Yes. Palatalization and assibilation of dentals/alveolars before front vowels is one of the most common kinds of sound change (you don't really even need intermediate steps between t > tʃ and d > z if you don't want them, though in this case they do help to make the initial step in the chain of changes more "symmetrical"). Fricative devoicing also has precedent (e.g. Spanish). So you're good.

[Nate]

Yes. Palatalization and assibilation of dentals/alveolars before front vowels is one of the most common kinds of sound change (you don't really even need intermediate steps between t > tʃ and d > z if you don't want them, though in this case they do help to make the initial step in the chain of changes more "symmetrical"). Fricative devoicing also has precedent (e.g. Spanish). So you're good.

Sweet. Thank you. My main issue was that I doubted that they could happen in the same language.

[----]

Lhasa Tibetan has devoiced all voiced obstruents, and many phonemes have undergone heavy palatalization, so a similar thing has happened in that language.

[Kaenif]

Was trying to derive lateral obstruents and remove l in consonant clusters from an earlier language. Are the following changes plausible?

ɸl, sl, hl > ɬ / _{e,i}
hʷl > ɬʷ / _{e,i}
dl, tl, gl, kl > tɬ / _{e,i}
gʷl, kʷl > tɬʷ / _{e,i}
l > ʷ / C_{a, ə, o, u}
bl, pl > bʲ, pʲ

[----]

Does pi > t͡ɕʷi make sense?

[Nortaneous]

Was trying to derive lateral obstruents and remove l in consonant clusters from an earlier language. Are the following changes plausible?

ɸl, sl, hl > ɬ / _{e,i}
hʷl > ɬʷ / _{e,i}
dl, tl, gl, kl > tɬ / _{e,i}
gʷl, kʷl > tɬʷ / _{e,i}
l > ʷ / C_{a, ə, o, u}
bl, pl > bʲ, pʲ

Looks fine to me, although it'd make much more sense with the order reversed. I'm also not sure about /bl pl/ palatalizing even in back-vowel environments when nothing else does, but that could come before l-velarization before back vowels.

Does pi > t͡ɕʷi make sense?

It's attested all over Europe without the rounding, so if your labials have more rounding than the average Eurolang labial, sure.

[WeepingElf]

Was trying to derive lateral obstruents and remove l in consonant clusters from an earlier language. Are the following changes plausible?

ɸl, sl, hl > ɬ / _{e,i}
hʷl > ɬʷ / _{e,i}
dl, tl, gl, kl > tɬ / _{e,i}
gʷl, kʷl > tɬʷ / _{e,i}
l > ʷ / C_{a, ə, o, u}
bl, pl > bʲ, pʲ

Looks fine to me, although it'd make much more sense with the order reversed. I'm also not sure about /bl pl/ palatalizing even in back-vowel environments when nothing else does, but that could come before l-velarization before back vowels.

Happened in Italian: platea > piazza.

Does pi > t͡ɕʷi make sense?

It's attested all over Europe without the rounding, so if your labials have more rounding than the average Eurolang labial, sure.

All over Europe? I cannot think of any example.

[Whimemsz]

Well [labial]+i/j > (some sort of palatal thing) is a very common change (there's a clear preference crosslinguistically to avoid labial+[j] clusters or palatalized labials). A number of Romance languages are an example (e.g. French sache < *sapja). Though I remember reading a thing once arguing that these were by an intermediate step whereby the glide was devoiced and spirantized to [ç] or the like and then the labial was lost.

[Herr Dunkel]

Let's say there's free stress that is mobile and moves during agreement.
After some time, stress moves to the initial syllable.

Is it possible to have the primodial initially-stressed syllables acquire high tone, and the new ones a low tone (I think it is, but better be safe, not sure if this happened anywhere)

[Nortaneous]

Perhaps a slight exaggeration. French, Romanian, and Tsakonian are the ones I can think of without going through and looking, and some Slavic langs come close, with j > J / m_.