Sound Change Quickie Thread

[----]

Well, the characters used in the native script actually are the palatals. Maybe it is though. There's a separate sequence spelled <-it> and it's pronounced as [eɪʔ]. I'm really not sure now.

---

Dbe has intervocalic voicing for plain stops, but there is a separate phoneme /d/. Could I have it be realized as an implosive [ɗ ] as a way for speakers to distinguish it from intervocalic /t/, even if there's no glottal stuff going on otherwise in that environment?

[Dē Graut Bʉr]

What do you think of aspiration before long vowels?

[Pogostick Man]

What do you think of aspiration before long vowels?

Given the conditioning this doesn't seem too beyond the pale for me. Stressed-syllable-initial voiceless stops tend to aspirate in English—do the long vowels in your lang attract stress?

[Dē Graut Bʉr]

What do you think of aspiration before long vowels?

Given the conditioning this doesn't seem too beyond the pale for me. Stressed-syllable-initial voiceless stops tend to aspirate in English—do the long vowels in your lang attract stress?

Yes, they do attract stress, so then I suppose it's possible.

[Porphyrogenitos]

I'm trying to figure out what to do with a palatal and a labiovelar series /pʲ tʲ kʲ pʷ tʷ kʷ/. I've decided on /tʲ/ > /t͡s/ and /kʲ/ > /t͡ʃ/, but I'm unsure of what to do with /pʲ/ and the labiovelars. Any ideas? Are there any plausible options for /kʷ/ aside from /p/ or /w/ or perhaps /g/?

[WeepingElf]

I'm trying to figure out what to do with a palatal and a labiovelar series /pʲ tʲ kʲ pʷ tʷ kʷ/. I've decided on /tʲ/ > /t͡s/ and /kʲ/ > /t͡ʃ/, but I'm unsure of what to do with /pʲ/ and the labiovelars. Any ideas? Are there any plausible options for /kʷ/ aside from /p/ or /w/ or perhaps /g/?

This is similar to a problem I encountered in East Hesperic (Hesperic is my main conlang family which Old Albic belongs to). In East Hesperic, I have /tj/ > /ts/ and /kj/ > /tʃ/, which begs the question what happened to /pj/. I decided that it became /f/. The Proto-Hesperic labiovelars simply lost their labialization and merged with the plain velars (after a wave of palatalization that turned some of the plain velars to postalveolar affricates).

[Daedolon]

is this realistic?

p t k s > p_h t_h k_h s_h
b d g z > p t k s > p_> t_> k_> s_>

[WeepingElf]

is this realistic?

p t k s > p_h t_h k_h s_h
b d g z > p t k s > p_' t_' k_' s_'

Yes. It happened in some Eastern Armenian dialects (with the stops, at least; I think not with /s/).

[Pole, the]

is this realistic?

p t k s > p_h t_h k_h s_h
b d g z > p t k s > p_' t_' k_' s_'

Yes. It happened in some Eastern Armenian dialects (with the stops, at least; I think not with /s/).

Has it produced ejectives in Armenian, too?

[Chagen]

For my PIE conlang family I want to keep the laryngeals but only in certain positions. Are the following plaisible, taking <L> to mean a laryngeal and <#> to mean the beginning or end of a word:

VL#: kept
CLC: becomes a vowel
CLV: dropped
#LC: dropped (unless C is a non-nasal resonant in which case it stays)
VLC: kept
VLCC: dropped, vowel is colored/lengthened (i.e the hypothetical word *deHmos would become <dehma> but *deHpros would become <dēpra>
LV: kept
VLV: kept

Note that all laryngeals would collapse into one /x~h/. Also they would still color even if they don't drop meaning that <ah ēh> would be a lot more common than <eh> (no <oh> because o/a/_ in this lang) if I'm understanding them right.

[WeepingElf]

is this realistic?

p t k s > p_h t_h k_h s_h
b d g z > p t k s > p_' t_' k_' s_'

Yes. It happened in some Eastern Armenian dialects (with the stops, at least; I think not with /s/).

Has it produced ejectives in Armenian, too?

Yes, but only in some eastern dialects. The first step (aspiration of old voiceless stops, devoicing of old voiced stops) is common to all of Armenian, and had already happened in Old Armenian. From there, Western Armenian underwent another shift (about which I have seen contradictory statements), and Eastern Armenian mostly kept the Old Armenian system intact. But some Eastern Armenian dialects developed ejectives from the voiceless unaspirated stops.

[احمکي ارش-ھجن]

are preceeding front vowels a good environment for stop consonant lenition?
And if I choose one stop consonant to lenit would the other have to lenit in that environment too?
I want that /t/ and /d/ lenit to dental fricatives word finally if there is a preceeding front vowel, but I do not want the other stops to do this in the same environment.

[Pole, the]

is this realistic?

p t k s > p_h t_h k_h s_h
b d g z > p t k s > p_' t_' k_' s_'

Yes. It happened in some Eastern Armenian dialects (with the stops, at least; I think not with /s/).

Has it produced ejectives in Armenian, too?

Yes, but only in some eastern dialects. The first step (aspiration of old voiceless stops, devoicing of old voiced stops) is common to all of Armenian, and had already happened in Old Armenian. From there, Western Armenian underwent another shift (about which I have seen contradictory statements), and Eastern Armenian mostly kept the Old Armenian system intact. But some Eastern Armenian dialects developed ejectives from the voiceless unaspirated stops.

Oh right, thanks.

[WeepingElf]

are preceeding front vowels a good environment for stop consonant lenition?
And if I choose one stop consonant to lenit would the other have to lenit in that environment too?
I want that /t/ and /d/ lenit to dental fricatives word finally if there is a preceeding front vowel, but I do not want the other stops to do this in the same environment.

Hmm, at first glance this does not seem very plausible, but the front vowel may palatalize the stop, and the resulting palatal affricate may later become a fricative, as in /t/ > /tʃ/ > /tθ/ > /θ/ or similar.

[KathTheDragon]

However, if dental consonants are palatalising here, is it not also highly likely that velars would palatalise too?

[WeepingElf]

However, if dental consonants are palatalising here, is it not also highly likely that velars would palatalise too?

True, that is highly likely. So my suggestion may just not cut it.

EDIT: Found a solution in another thread. First you palatalize: /t/ > /ts/, /k/ > /tʃ/. Second, you get /ts/ > /tθ/ > /θ/ and /tʃ/ > /ʃ/. Now you have /ʃ/ > /x/, as in Spanish, and then /x/ > /k/. Result: /t/ is lenited to /θ/ and /k/ comes full circle to /k/. Accordingly for the voiced counterparts.

[sangi39]

However, if dental consonants are palatalising here, is it not also highly likely that velars would palatalise too?

True, that is highly likely. So my suggestion may just not cut it.

EDIT: Found a solution in another thread. First you palatalize: /t/ > /ts/, /k/ > /tʃ/. Second, you get /ts/ > /tθ/ > /θ/ and /tʃ/ > /ʃ/. Now you have /ʃ/ > /x/, as in Spanish, and then /x/ > /k/. Result: /t/ is lenited to /θ/ and /k/ comes full circle to /k/. Accordingly for the voiced counterparts.

I posted this over on the CBB, but I was wondering if Sardinian could be used as a natlang precedent here, where /tj dj/ underwent palatalisation but /kj gj/ did not, unlike the rest of the Romance languages.

[Dewrad]

However, if dental consonants are palatalising here, is it not also highly likely that velars would palatalise too?

True, that is highly likely. So my suggestion may just not cut it.

EDIT: Found a solution in another thread. First you palatalize: /t/ > /ts/, /k/ > /tʃ/. Second, you get /ts/ > /tθ/ > /θ/ and /tʃ/ > /ʃ/. Now you have /ʃ/ > /x/, as in Spanish, and then /x/ > /k/. Result: /t/ is lenited to /θ/ and /k/ comes full circle to /k/. Accordingly for the voiced counterparts.

I posted this over on the CBB, but I was wondering if Sardinian could be used as a natlang precedent here, where /tj dj/ underwent palatalisation but /kj gj/ did not, unlike the rest of the Romance languages.

Equally, some varieties of English? For me tube and chew begin with [tʃ], June and dune with [dʒ], but cute retains [kj] (can't think of a good word with /gj/...)

[احمکي ارش-ھجن]

However, if dental consonants are palatalising here, is it not also highly likely that velars would palatalise too?

True, that is highly likely. So my suggestion may just not cut it.

EDIT: Found a solution in another thread. First you palatalize: /t/ > /ts/, /k/ > /tʃ/. Second, you get /ts/ > /tθ/ > /θ/ and /tʃ/ > /ʃ/. Now you have /ʃ/ > /x/, as in Spanish, and then /x/ > /k/. Result: /t/ is lenited to /θ/ and /k/ comes full circle to /k/. Accordingly for the voiced counterparts.

I posted this over on the CBB, but I was wondering if Sardinian could be used as a natlang precedent here, where /tj dj/ underwent palatalisation but /kj gj/ did not, unlike the rest of the Romance languages.

Equally, some varieties of English? For me tube and chew begin with [tʃ], June and dune with [dʒ], but cute retains [kj] (can't think of a good word with /gj/...)

I think I'll go the Sardinian path.
Also, how many years would these sound changes take?

[Dewrad]

However, if dental consonants are palatalising here, is it not also highly likely that velars would palatalise too?

True, that is highly likely. So my suggestion may just not cut it.

EDIT: Found a solution in another thread. First you palatalize: /t/ > /ts/, /k/ > /tʃ/. Second, you get /ts/ > /tθ/ > /θ/ and /tʃ/ > /ʃ/. Now you have /ʃ/ > /x/, as in Spanish, and then /x/ > /k/. Result: /t/ is lenited to /θ/ and /k/ comes full circle to /k/. Accordingly for the voiced counterparts.

I posted this over on the CBB, but I was wondering if Sardinian could be used as a natlang precedent here, where /tj dj/ underwent palatalisation but /kj gj/ did not, unlike the rest of the Romance languages.

Equally, some varieties of English? For me tube and chew begin with [tʃ], June and dune with [dʒ], but cute retains [kj] (can't think of a good word with /gj/...)

I think I'll go the Sardinian path.
Also, how many years would these sound changes take?

230 years, four months and about a week and a half, excluding public holidays.

[احمکي ارش-ھجن]

Thats funny, but be serious....

[KathTheDragon]

Assume an arbitrary number of generations for each sound change to run its course.

[Dewrad]

Thats funny, but be serious....

There's no way of telling. The rate of soundchange is not something which we've been able to find a scientific metric for: a soundchange could hit all the given lexemes in a language in less than a generation, or the shift could be slow, incremental and take several centuries: there are just too many factors at play. Really, the answer to your question is "how long is a piece of string?"

[Salmoneus]

Worth re-iterating the second part of that. The problem is not that there isn't an answer, but that there isn't a question. "Sound change" isn't, in reality, broken up into discrete, easily-measured-or-defined 'sound changes'. Each 'sound change' is, in most cases, a matter of a shift along a continuum, and that shift must in turn propagate both throughout the whole vocabulary and throughout all the dialects and registers. And then again, phonemic sound changes involve both the allophonic change itself and then the other change that makes the first one phonemic. Which can, in particularly, lead to very 'lumpy' progression of sound changes - you can have masses of allophonic changes building up, as it were, until suddenly a small surface change like 'loss of final schwa' makes all the others suddenly phonemic. And it's also important to remember that... well, I'm not sure this is absolutely true, but it's a good starting point to consider that sound changes are blind to their significance, they don't know whether they're going to end up phonemic or not.

All of which means that you can sometimes have a feel for this being too many changes or that being too few changes for a given period, but you can't really put figures on that.

In this case, and having expressed how stupid I am to give an answer: a simple change like /it/ > [its] > [itT] > /iT/... I think would take at least two or three generations, given a reasonable population size and connectivity? I would probably put a lower bound of 50 years, probably more likely 150 years? There isn't really an upper bound - the last part ([itT] > /iT/) is likely to happen quickly, I would imagine, but [it] and [its] could both stick around a long time.

Of course, in certain circumstances it could all happen virtually overnight, who knows?

[Daedolon]

Is this plausible ?
ɐ̃ ĩ õ ũ ẽ > ɐn in on un en
or
ɐ̃ ĩ õ ũ ẽ > ɐ: i: o: u: e:
or
ɐ̃ ĩ õ ũ ẽ > ɐ i o u e

And what could be done to with [w̃] and [j̃] ?