Sound Change Quickie Thread

[sangi39]

Ah, yeah, the changes I presented over on the CBB weren't meant to indicate free variation, but variations of each of the original Proto-Hasjakam vowels found in closed and open syllables respectively. Since I didn't present any changes that could have caused those variations to become phonemic, I continued to present them in those pairs.

Oh... thanks. How do non-syllabic vowels develop for diphthongs when there were originally none. Do they just... Spontaneously come from an approximant?

I don't see a reason why not, and I'm sure such a development has been suggested here before. I could be wrong though.

[vokzhen]

Offglides can spontaneously appear just like vowels can spontaneously change position (e.g. a: > o: or u > y). English has i: u: a: o: > ai au ei əu. German has i: u: y: > ai au ɔy. Icelandic's got a: æ: e: o: > au ai je ou. Some of those might be due to a large number of vowels pushing on each other, so offgliding gives them more "distance" from each other, but I'm not sure it's necessary. It could also be triggered by something in particular, an i or u in the following syllable triggering breaking or something along those lines.

[احمکي ارش-ھجن]

How would a voiceless alveolar palatal fricative become a voiceless palatal lateral fricative?

[Porphyrogenitos]

How would a voiceless alveolar palatal fricative become a voiceless palatal lateral fricative?

That sounds like something that could happen unconditionally, without any steps in between.

[احمکي ارش-ھجن]

How would a voiceless alveolar palatal fricative become a voiceless palatal lateral fricative?

That sounds like something that could happen unconditionally, without any steps in between.

I mean /ɬ/ -> /ʎ̝̊/, I seem to have said "alveolar-palatal" rather than "alveolar lateral"

[Pogostick Man]

How would a voiceless alveolar palatal fricative become a voiceless palatal lateral fricative?

That sounds like something that could happen unconditionally, without any steps in between.

I mean /ɬ/ -> /ʎ̝̊/, I seem to have said "alveolar-palatal" rather than "alveolar lateral"

Seems like you could just have it in a palatalizing environment—around /i/ or a front vowel or /j/, or maybe something involving consonant harmony.

[Zaarin]

How would a voiceless alveolar palatal fricative become a voiceless palatal lateral fricative?

That sounds like something that could happen unconditionally, without any steps in between.

I mean /ɬ/ -> /ʎ̝̊/, I seem to have said "alveolar-palatal" rather than "alveolar lateral"

In the environment of front vowels and other palatal consonants like /j/ or /c/ seems like an obvious choice; change the vowel, insert an epenthetic vowel, or change the palatal consonant and you've got a phoneme.

EDIT: Ninja'd.

[Nortaneous]

could happen unconditionally, bet that's where a lot of those weird velar laterals come from too. lateral fricatives/affricates can vary in POA on their own.

[vokzhen]

Yea I'd say unconditionally works. Hadza's lateral affricates are palatal in release and alveolar~palatal in onset, Welsh <ll> is apparently moving towards [ç], Archi's got pre-velars in place of every other Caucasian language's alveolars so it was presumably backed via a palatal.

[Zaarin]

How does this change look?

/VstV/ /VθθV/ /VːʃV/

[Pogostick Man]

How does this change look?

/VstV/ /VθθV/ /VːʃV/

I'd think it'd be more like this:
VstV → VθtV or VsθV → VθːV → VːθV → either straight to VːʃV, or VːsV and palatalization before some front vowel.

[Zaarin]

How does this change look?

/VstV/ /VθθV/ /VːʃV/

I'd think it'd be more like this:
VstV → VθtV or VsθV → VθːV → VːθV → either straight to VːʃV, or VːsV and palatalization before some front vowel.

Excellent, that works as well.

[Max1461]

How plausible is vowel metastasis across consonants? Things like uCa -> Cua, or uCa -> aCu? What kinds of environment could trigger it?

[Bristel]

Can apical consonants become an alveopalatal laminal consonant before a front (labial, coronal and palatal) consonant?

ex. [l] becomes [ɬ̱ʲ] before a front consonant?

Also, after alveopalatal obstruents or nasals can apicals become retroflex? ex. [l] becomes [ɬ̢]?

These are some things gleb gave for a random allophony, and while I like these ideas, I don't know if they are realistic, and there are also more allophony processes that I don't want to use from the same seed, I'm using a simplified version of the phonology, maybe adding a few other things to it that I do like.

Here's the seed number if you want to check it out in full: 1620359910

[vokzhen]

Mmmmaybe. I wouldn't expect apicals to become palatalized before other apicals, though you could have a collapse of an apical-laminal distinction before certain sounds. I'm not aware of labials patterning along with the (non-retroflex) coronals; Australian languages pattern labials with velars as peripheral consonants, and afaik labials and velars are more similar in how they affect vowels than either are to coronals (except perhaps retracted/retroflexed). I doubt the second, a cluster like [tɕʈ] or [ʎʂ] I really couldn't see spontaneously happening for /tɕt/ or /ʎs/, the tongue movement is too awkward, unless you used the retroflex syllables as shorthand for something like an apico-alveolar where the normal apicals are articulated against the root of the teeth. Actually, maybe this could explain my problem with the first change: the only coronal clusters that are allowed in the first place is palatalized-apical, and any apical-apical were replaced (especially if palatalized-apical was much more common than apical-apical or apical-palatalized). Otherwise perhaps you could have dissimilation where the second C in CVC shifts to retroflex when the first is palatalized.

[Bristel]

I'm just confused because the gleb phonology doesn't say what is apical or laminal, or when one consonant is apical or laminal, and only distinguishes coronal and palatal alveolar, so I might as well just ignore the whole part of the allophony besides some of the simple aspiration and ejective rules that also were there.

[Daedolon]

How about this?

Stressed short vowels in open syllables become long.
ɛː ɔː > eː oː
ɛ ɔ > je we
ã ẽ ĩ õ ũ > an en in on un

ʦ ʣ ʧ ʤ > s z ʃ ʒ
z ʒ > s ʃ
ʃ > x
x > h
ʎ > j
ɲ > j̃ > j

[Porphyrogenitos]

How about this?

Stressed short vowels in open syllables become long.
ɛː ɔː > eː oː
ɛ ɔ > je we
ã ẽ ĩ õ ũ > an en in on un

ʦ ʣ ʧ ʤ > s z ʃ ʒ
z ʒ > s ʃ
ʃ > x
x > h
ʎ > j
ɲ > j̃ > j

All of those seem perfectly plausible.

Though, the consonants, taken together - why is it that z ʒ (and no others) are being devoiced, when all the other fricatives are moving back? It seems more likely to have s z > ʃ ʒ. Not that z ʒ > s ʃ isn't possible, even in the context you describe. But it just doesn't seem to fit quite as well.

[8Deer]

All of those seem perfectly plausible.

Though, the consonants, taken together - why is it that z ʒ (and no others) are being devoiced, when all the other fricatives are moving back? It seems more likely to have s z > ʃ ʒ. Not that z ʒ > s ʃ isn't possible, even in the context you describe. But it just doesn't seem to fit quite as well.

If /z ʒ/ are the only voiced fricatives then I don't see a problem. In fact, didn't a very similar shift happen in the history of Spanish? Although with the way these rules are ordered, original /dʒ/ will end up as /h/, which may not have been intended...

[احمکي ارش-ھجن]

I've got two dialects: Uzerian Vrkhazhian and Mukhebic Vrkhazhian.

What ways for the Mukhebic dialect of a language to gain prenasalized voiced stops and fricatives, as well as for it to eliminate pre-aspirated consonants and uvularized consonants? Assume that the consonants can occur word-initially, -medially, and -finally.

The inventories in question (if necessary);
Uzerian:
/a e̞ i o̞ ə u/
/ae̯ ao̯ ɛo̯/
/m̥ m n̥ n ɴ/
/p ʰp pʶ b t ʰt tʶ d k ʰk g q ʰq ʔ/
/t͡s c͡ç ʰc͡ç ɟ͡ʝ/
/ɸ β s z ç ʝ x ɣ χ h/
/ɹ̥ ɹ j w/
/r ʀ/
/l̥ l ʎ̝̊/
/m̩ n̩ ɹ̩/

Mukhebic:
/a e̞ i o̞ ə u/
/ae̯ ao̯/
/m̥ m n̥ n ɴ/
/p b ᵐb t d ⁿd k g ᵑg q ʔ/
/t͡s c͡ç ɟ͡ʝ/
/f v s z ⁿz ç ʝ x ɣ ᵑɣ χ h/
/ɹ̥ ɹ j w/
/r/
/l̥ l ʎ̝̊/
/m̩ n̩ ɹ̩/

[vokzhen]

Chain shift of hp > p > b > mb, gaining additional prenasals by merging NC clusters into unitary phonemes. Uvularization could be lost to vowel quality which later levels, e.g. ʶi ʶe ʶu ʶo ʶə ʶa > ɨ ɛ ʊ ɔ a ɑ > ə e o a a a. Depending on what else you're doing to mess up the vowels between the two, other things might happen too; if secondary /a/ is really common - like if all of /ə ʶo ʶa ʶə/ push towards it - maybe that pushes original /a/ to /e/, with /ə/ being rare as it only now comes from /ʶi/.

Given that you have uvularization patterning as a change in voicing, I'd presume under such a scheme they would end up voiced, just as the other voiceless non-pre-aspirated consonants do. I'd also expect that the uvulars would probably end up doing whatever the uvularized ones do too (i.e. tʶi > tə > də, then qi > kə > gə and χi > xə as well), though I don't think it's too much of a problem if they act like a POA rather than voicing.

[احمکي ارش-ھجن]

Chain shift of hp > p > b > mb, gaining additional prenasals by merging NC clusters into unitary phonemes. Uvularization could be lost to vowel quality which later levels, e.g. ʶi ʶe ʶu ʶo ʶə ʶa > ɨ ɛ ʊ ɔ a ɑ > ə e o a a a. Depending on what else you're doing to mess up the vowels between the two, other things might happen too; if secondary /a/ is really common - like if all of /ə ʶo ʶa ʶə/ push towards it - maybe that pushes original /a/ to /e/, with /ə/ being rare as it only now comes from /ʶi/.

Given that you have uvularization patterning as a change in voicing, I'd presume under such a scheme they would end up voiced, just as the other voiceless non-pre-aspirated consonants do. I'd also expect that the uvulars would probably end up doing whatever the uvularized ones do too (i.e. tʶi > tə > də, then qi > kə > gə and χi > xə as well), though I don't think it's too much of a problem if they act like a POA rather than voicing.

I was also pondering having two velarized nasals /mˠ nˠ/ that would only occur before high/close vowels such as /i u/. Those two are strange phenomenon, as they are not really phonemic as they only occur before those vowels, nor are they allophonic, as there can be normal /mi mu ni nu/. It might just be phonotactics.
By the way, the uvular trill is considered an "uvularized consonant"...
I plan on unconditional raising of mid vowels, with /e o/ arising from somewhere else like the diphthongs? They already mutate to high vowels as part of morphology: q-w-zh "be" quzh as opposed to the standard stem CuCaC.
Lastly, the prenasals would be analyzed as pure nasal + plosive clusters in codas and word-finally. There is a phonetic difference between the two.

According to sound changes;
Uzerian / Mukhebic:
Horn: [tʶam] / [dam]
Person: [metʶaɟ͡ʝ] / [midaɟ͡ʝ] or [medaɟ͡ʝ]
Food: [zaqəd] /[zagad] (I don't think this is diachronically likely, as /q/ has existed in the language family, it is not, diachronically, a uvularized /k/, but it may be analyzed by its speakers as such)
Roof: [Xak] / [ɣak]
Bell: [ʀod] / [rad]
Chest (body): [qwaɸ] / [qwaf]
Bird: [qɹoç] / [qɹiç] or [qɹoç]
Bear: [maʰp] / [mamb]
Spirit: [ʰtaʔ] / [ⁿdaʔ]
Throne: [muXəl] / [muɣal]
Poison: [pʶas] / [bas]
Arm: [meɴqa] / [miŋga] or [meŋga]
Pendulum: [maʀum] / [marom]
Something like that?

A problem though: My conlang has a triconsonsonantal root system. How will that affect the triconsonsonantal roots. They'd have to be different. For example, z-q-d "to eat" would become z-g-d " to eat" which might merge with another root of the same consonant. I don't think this happens in Semitic languages, or rather, I can't think of examples where this occurs, either between two different Semitic languages, or among a language's dialects

[vokzhen]

Looks like you got confused on the consonant chain shift (to be fair, that notation is a bit confusing). The preaspirates become plain, the plain voiced, and the voiced prenasal. So:
Person: [metʶaɟ͡ʝ] / [midaⁿɟ͡ʝ] or [medaⁿɟ͡ʝ]
Food: [zaqəⁿd] /[zagaⁿd] (or [zaqaⁿd])
Roof: [Xak] / [ɣag]
Bell: [ʀod] / [raⁿd]
Bear: [maʰp] / [map]
Spirit: [ʰtaʔ] / [taʔ]
And so on. I'd really expect the palatals to participate in the chain shift, unless they've been reanalyzed as being more like fricatives than stops. /q/ is a bit more iffy, just that they start out as different than the uvularized doesn't necessarily mean much, especially when you're treating /X ʀ/ as uvularized and not just uvular, and especially if they pattern with the uvularized ones (there's no /kʶ/, and if /q/ appears in places where you'd diachronically and morphophonologically have /kʶ/ that would strengthen it). Also, you could go X > ɣ, but as it's not a stop you could just ignore the voicing change as have X > x.

Lastly, the prenasals would be analyzed as pure nasal + plosive clusters in codas and word-finally. There is a phonetic difference between the two.

Can you reword this? Are you saying that /nd/ contrasts with /ⁿd in these positions, or not? I was assuming that a word like /kanda/ would end up as /gaⁿda/, and either there wouldn't be a /ganda/ anymore or it would have come from something like /kanuda/ with a medial unstressed vowel being dropped. That's the only way my proposed changes get the nasalized fricatives (from e.g. /kanza/ > /gaⁿza/ and /minz/ > /miⁿz/), and it would be likely to give you ᵐv and ⁿʝ as well, unless they undergo further changes.

[احمکي ارش-ھجن]

Looks like you got confused on the consonant chain shift (to be fair, that notation is a bit confusing). The preaspirates become plain, the plain voiced, and the voiced prenasal. So:
Person: [metʶaɟ͡ʝ] / [midaⁿɟ͡ʝ] or [medaⁿɟ͡ʝ]
Food: [zaqəⁿd] /[zagaⁿd] (or [zaqaⁿd])
Roof: [Xak] / [ɣag]
Bell: [ʀod] / [raⁿd]
Bear: [maʰp] / [map]
Spirit: [ʰtaʔ] / [taʔ]
And so on. I'd really expect the palatals to participate in the chain shift, unless they've been reanalyzed as being more like fricatives than stops. /q/ is a bit more iffy, just that they start out as different than the uvularized doesn't necessarily mean much, especially when you're treating /X ʀ/ as uvularized and not just uvular, and especially if they pattern with the uvularized ones (there's no /kʶ/, and if /q/ appears in places where you'd diachronically and morphophonologically have /kʶ/ that would strengthen it). Also, you could go X > ɣ, but as it's not a stop you could just ignore the voicing change as have X > x.

Nah, we and Sangi discussed this.
I'm not including the voiced palatal affricate into the sound change. I'm not going to pattern it like a stop. I'm going to analyzed /X/ as pure uvular.

Lastly, the prenasals would be analyzed as pure nasal + plosive clusters in codas and word-finally. There is a phonetic difference between the two.

Can you reword this? Are you saying that /nd/ contrasts with /ⁿd in these positions, or not? I was assuming that a word like /kanda/ would end up as /gaⁿda/, and either there wouldn't be a /ganda/ anymore or it would have come from something like /kanuda/ with a medial unstressed vowel being dropped. That's the only way my proposed changes get the nasalized fricatives (from e.g. /kanza/ > /gaⁿza/ and /minz/ > /miⁿz/), and it would be likely to give you ᵐv and ⁿʝ as well, unless they undergo further changes.

I'm saying that /ⁿd/ and /nd/ do not contrast in coda position and word-finally. The palatal fricative will be spare from the prenasalization, for reasons.
Now, I'm not entirely sure how I feel about prenasal stops anymore, they are going to mess up my triconsonantal root system's roots more than the de-uvularization. I wouldn't mind, but this dialect is supposed to be mutually intelligible, and the creation of prenasals from NC clusters that were originally parts of roots will screw up a lot of things.

[sangi39]

I don't think it's going to affect your triconsonantal system too much (just look at palatalisation in the Semitic languages of Ethiopia or the loss of /n/ in certain verb forms in Hebrew. It causes a change in some parts of the system, but it works more or less the same way it did before. A chain shift would hardly touch triconsonantalism as a whole).

Mutual intelligibility, though, would likely become an issue here very quickly. Loss of uvularisation (but maintaining voiclessness of Vrkhazhian uvularised consonants, and keeping uvulars as uvulars) after allophonic vowel lowering might not be too much of an issue, especially if the same allophony occurs in Uzerian without loss of uvularisation.

A chain shift, though, might mess that up. De-uvularisation would cause one merger, but I don't think it would be that bad overall, since the vowel lowering in some forms would match up with allophonic lowering in Uzerian (so, say you've got a form that generally speaking is CiCuCa, but appeaars as CeCuCa if the first consonant is uvularised, CiCoCa if it's the second and so on. The Uzerian listener might fine the consonants of the Mukhebic speaker weird, but the vowels might compensate for that somewhat). However, once you, say, shift voiceless plosives to voiced plosives and pre-aspirated plosives to voiceless plosives, that level of predictability and intelligibility might decrease, especially if it's an unconditional shift.