Sound Change Quickie Thread

[Nortaneous]

Is it at all plausible to have a series of palatals but no velars (except allophonically)? If so, how might it come about? Fronting of velars?

Yes. Vanimo and Xavante. Not sure how it'd come about, but fronting of velars sounds plausible.

[ná'oolkiłí]

Do you have uvulars? For a lot of PNW and Caucasian languages that have velar and uvular series, the former is often phonetically prevelar if not postpalatal. So, for instance, **/kʰ k k' (qʰ q q')/ */k̟ʰ k̟ k̟'/ /cʰ c c'/ could be pretty plausible. And if you didn't want any uvulars, a coinciding **/qʰ q q'/ */χ χ ʔ/ /h h ʔ/ or something could leave you without dorsals except for palatals. I phonology without velars is very typologically strange, and I'd think very unstable, so if you wanted to make daughter language(s) from it you could get some interesting results =)

[Jetboy]

Oh, thanks, that was a lot easier than I'd expected it to be. Isn't there some sort of universal that says that a language has to have at least three of /ptkʔ/?

EDIT: No, I don't have any uvulars, I'm afraid; they intimidate me.

[Nortaneous]

I phonology without velars is very typologically strange, and I'd think very unstable, so if you wanted to make daughter language(s) from it you could get some interesting results =)

How so? I don't really see any good way for velars to develop except through borrowing, unless maybe something weird happens with /t/.

Isn't there some sort of universal that says that a language has to have at least three of /ptkʔ/?

Not really. I think it's just at least three plosives.

[ná'oolkiłí]

No, I don't have any uvulars, I'm afraid; they intimidate me.

Haha, don't be— uvulars are fun

How so? I don't really see any good way for velars to develop except through borrowing, unless maybe something weird happens with /t/.

Hmm, I suppose it would really depend on the rest of the inventory, but maybe:
*p t c m n kʷ k k ŋʷ ŋ / __ [+vowel +back]
*l w ɰ~ɣ (/ __ [+vowel +back])
*C{w, l} k(ʷ)
*Ø ŋ / #__V (I think this happened in some Samoyedic languages, right? Or am I making that up...)
or even (and this is probably stretching things)
*Ø ʔ ʡ q͡ʡ (lol Somali) q k | #__V

[Tropylium⁺]

How so? I don't really see any good way for velars to develop except through borrowing, unless maybe something weird happens with /t/.

/t/ has a high likelihood of backing if there are no coronal stops of any sort in the system. There are at least 20 branches of Oceanic [pdf] that have more or less independantly done k ʔ, t k. Hawai'ian, Samoan and dialectal Tahitian are just the most famous examples. Sometimes it's conditional rather than unconditional, with eg *t_t > *k_t.

I suppose it would really depend on the rest of the inventory, but maybe:
*p t c m n kʷ k k ŋʷ ŋ / __ [+vowel +back]
*l w ɰ~ɣ (/ __ [+vowel +back])
*C{w, l} k(ʷ)
*Ø ŋ / #__V (I think this happened in some Samoyedic languages, right? Or am I making that up...)

Yeah, they did (the Northern group, to be specific). Also some others had *w kʷ / #_.
All the changes here seem plausible, but having them all occur simultaneously would probably be overkill.

[Nortaneous]

How plausible is this?

ʒ (phonetically rounded) ɥ (also z weakens to some sort of rhotic which eventually develops into ɾ and y develops at around this time also)

ʃ ʂ x χ / [V -front]_
ʃ s (phonetically palatalized) / [V +front]

[Jetboy]

ʃ ʂ x χ / [V -front]_

This change doesn't sound too implausible; Spanish had something similar, though it only went to /x/.

[Acid Badger]

I've got a proto-language with a series of palatalised consonants and I want the daughter languages to lose them.
So, any nice ideas what to do with /xʲ/ and /ŋʲ/?

[Bob Johnson]

I've got a proto-language with a series of palatalised consonants and I want the daughter languages to lose them.
So, any nice ideas what to do with /xʲ/ and /ŋʲ/?

What's wrong with [x] and [ŋ]?

Or V i / Cʲ_ and then ʲ ∅ ... or just ʲ i

(Yes, it's boring.)

[jmcd]

I've got a proto-language with a series of palatalised consonants and I want the daughter languages to lose them.
So, any nice ideas what to do with /xʲ/ and /ŋʲ/?

[C] and [J] would be ideas as would [xj] and [Nj] / [nj]. Other ideas would be [j_0] and [j~]. Or you could shift the original [x] and [N] to [X] and [N\] before turning the [x_j] and [N_j] into plain velars.

[Atom]

What can one do with ejectives? I have created myself a series of ejectives ( tS_>, k_>, k_w_>) which I would like to remove.

[Jetboy]

I know that sometimes palatalized rhotics will shift from /rʲ/ to /jr/; if such a change happened in a language that also had a palatalized /n/, would the change /nʲ/ to /jn/ be possible?

[Astraios]

What can one do with ejectives? I have created myself a series of ejectives ( tS_>, k_>, k_w_>) which I would like to remove.

Voice them? I don't know how attested that is, but it doesn't seem a huge jump to me - ejective > voiced implosive > ordinary voiced plosive.

[jmcd]

I know that sometimes palatalized rhotics will shift from /rʲ/ to /jr/; if such a change happened in a language that also had a palatalized /n/, would the change /nʲ/ to /jn/ be possible?

Most certainly possible I'd say. Simialr to what happened in French IIRC.

[WeepingElf]

What can one do with ejectives? I have created myself a series of ejectives ( tS_>, k_>, k_w_>) which I would like to remove.

Voice them? I don't know how attested that is, but it doesn't seem a huge jump to me - ejective > voiced implosive > ordinary voiced plosive.

Some people say that exactly this happened in the prehistory of Proto-Indo-European.

[personak]

You know what should be done? A dictionary of phonemes, and how they can originate and disappear. Though people would still have questions about a specific sound-change, there would no longer be anything about "i have inventory blank, how do I get /phoneme/."

What do y'all think?

[Dewrad]

I know that sometimes palatalized rhotics will shift from /rʲ/ to /jr/; if such a change happened in a language that also had a palatalized /n/, would the change /nʲ/ to /jn/ be possible?

Most certainly possible I'd say. Simialr to what happened in French IIRC.

YDRC. VL *ponju > OF poing (something like /pojJ/ or /pojn/).

[Atom]

Interesting on the ejectives -> voiced plosives, especially since I already intended on keeping the bilabial and dental forms as implosives. I can just have the rest of them become normal voiced stops, which then fricativize (the conlang in question has no voiced stops). Thanks!

[Nortaneous]

VL *ponju > OF poing (something like /pojJ/ or /pojn/).

But was there an intermediary stage of /poJ/, or was that just metathesis?

[jmcd]

VL *ponju > OF poing (something like /pojJ/ or /pojn/).

But was there an intermediary stage of /poJ/, or was that just metathesis?

Would've been metathesis I'm sure. That's why the spelling has the i before the ng when other instances of /J/ are represented gn.

[Soap]

I dont know. I think Old French had different rules for spelling. Final -e was still pronounced, so they wouldnt represent final /ñ/ by {gne} as is done in modern French (though even today there's a very short off-glide schwa in most contexts). There doesnt seem to be much easily available online for guides about the spelling and pronunciation.

[jmcd]

I dont know. I think Old French had different rules for spelling. Final -e was still pronounced, so they wouldnt represent final /ñ/ by {gne} as is done in modern French (though even today there's a very short off-glide schwa in most contexts). There doesnt seem to be much easily available online for guides about the spelling and pronunciation.

In Modern French /J/ is represented by <gn>. Examples where it isn't the last phoneme are oignon /OJO~/ and magnetoscope /maJetoskop/. Some things did change, for example <z> no longer repesents /ts/, but <gn> I'm pretty sure is about the same it always was.

[cybrxkhan]

I'm starting to seriously dabble into sound changes for the first time, so I hopefully have some relatively easy to answer questions.

Anyhow, how plausible would the following scenario be:

What would be plausible ways for a voiced stop to become voiceless? I'm thinking of having an allophonic contrast between aspirated and nonaspirated voiced stops in the protolanguage, and then having the aspirated voiced stops become nonaspirated voiced stops, and the nonaspirated voiced stops become voiceless stops. I.e.:

/tʰ/ > /t/
/kʰ/ > /k/
/pʰ/ > /p/
/t/ > /d/
/k/ > /g/
/p/ > /b/


However, in my case, in the proto-language, there is a phonemic differentiation between unrounded and rounded vowels. Consonants would be aspirated before rounded vowels, but as the consonants go their changes as shown above, the rounded vowels simultaneously become unrounded, so thus, in the daughter language, there are no rounded vowels and aspirated stops.

---

Also, what would be a possible way for an affricate to develop in the daughter language? For instance, how would I get /t̠ʃ/? Is it as simple as saying the original /t/ and /ʃ/ combined under certain conditions?

---

Thank you in advance!

[finlay]

Yes, those work.

With tʃ: for one, you don't need to write the retracted diacritic, because this is implicit in the fact that ʃ is coming after (it's only really important to put this in if there's a contrast, in all but the narrowest of transcriptions). A better way to make it explicit is to use the tie bar, so t͡ʃ rather than tʃ. Even then there isn't an awful lot of phonetic differentiation between the two.

having t + ʃ → t͡ʃ is one way to do it, yeah, although you then have to have the two next to each other. Palatalisation is a common way to introduce it; this happens before high front vowels, so you could have [k]→[tʃ]/_i,e or [t]→[tʃ]/_i,e , and then you could lose the i if it comes before another vowel; the first happened in Italian, so [tʃao] rather than [kiao] in "ciao", and the second can be seen in words like 'question' [kwɛstʃən]. Similarly, you could have [tj]→[tʃ] or [kj]→[tʃ]. Commonly they go through the intermediate stage of [c] and/or [cç]. Or possibly [ts]. It's also possible to go straight from [ts] to [tʃ].

I could also take as an example my conlect Mybutan, which gets [tʃ] (and [dʒ]) from a bunch of original sounds... first you have Sentalian's [ɕ] turning to [ʃ] and [j] to [ʒ], so [tj] and [tɕ] become [tʃ] ([tʒ] isn't allowed and becomes [tʃ] here), although [tɕ] is only allowed medially in Sentalian. Mybutan also deletes schwas, so Sentalian's [təɕ] also becomes [tʃ]. [kj] also becomes [tʃ] and [gj] becomes [dʒ], but [kɕ] becomes [kʃ]. The other place it gets it from is Sentalian's /tl/, which is realised as the lateral affricate [tɬ]; here it just loses the lateral part of it and becomes the affricate [tʃ]. So it comes from a variety of places in the end (and there are a couple of things that make it possible in initial, medial and final position, too).