Sound Change Quickie Thread

[dɮ the phoneme]

Hmm, so you want e.g. /panaɾ/ → /panɾ/? That doesn’t sound at all likely to me… it would end up with your language gaining a lot of strange consonant clusters.

Strange clusters are basically the intention. But Richard W reminded me that this did basically happen in French (but the weird clusters where patched up by the addition of a paragogic /e/). Funnily enough the conlang that this is for basically uses the Latin stress rule, meaning that I might just end up ripping off exactly what French did but keeping the clusters around. I do wish I had another type specimen of this kind of change to compare with the French case though, since my language is really not phonologically like French at all.

edit: Actually, bradrn's particular example reminds me: what happened in on the way from proto-Germanic to Proto/Old Norse? IIRC the nominative singular ending is -r from PG -az/-iz/-uz, even following a consonant.

[Richard W]

Hmm, so you want e.g. /panaɾ/ → /panɾ/? That doesn’t sound at all likely to me… it would end up with your language gaining a lot of strange consonant clusters.

So what do you think of /vɑ̃dʁ/ in that conlang called French?

[Richard W]

French isn't like Latin any more!

edit: Actually, bradrn's particular example reminds me: what happened in on the way from proto-Germanic to Proto/Old Norse? IIRC the nominative singular ending is -r from PG -az/-iz/-uz, even following a consonant.

The Icelanders couldn't keep it up - they stick in <u> or lose consonants.

[bradrn]

I do wish I had another type specimen of this kind of change to compare with the French case though, since my language is really not phonologically like French at all.

Several Papuan languages seem to have undergone a similar process: huge vowel loss followed by huge vowel epenthesis. This study has a detailed study (mostly of Kalam but also discussing other languages).

Hmm, so you want e.g. /panaɾ/ → /panɾ/? That doesn’t sound at all likely to me… it would end up with your language gaining a lot of strange consonant clusters.

So what do you think of /vɑ̃dʁ/ in that conlang called French?

I think that French has a pretty strange phonology, which is nothing new. And reading a bit more about this, it seems that these clusters tend to undergo either resyllabification, epenthesis or deletion, all of which indicates that such weird clusters aren’t terribly stable in the scheme of things.

[Nortaneous]

I vaguely feel like I've asked this before? But I can't find it anywhere so who knows. Is there an attested case of a language (possibly over the course of multiple changes) deleting all vowels in word-final syllables?

If the Index Diachronica is to be trusted, this is very common, e.g. in Abenaki, Munsee Delaware, Nixumwak-Nêlêmwa, Marshallese, (Middle) Welsh… There’s also some Nilo-Saharan languages where final vowels tend to be devoiced (Ik is the one I remember, also possibly Kambaata–Alaaba–Kʼabeena) — perhaps that’s an intermediate stage in the process?

I don't just mean vowels in absolute final position, but all vowels in word-final syllables (including before a coda)

Most proto-languages of East Asia probably had Austronesian-like root structure + strong final stress and simplified through deletion of unstressed pretonic vowels, as the first step in a general areal procession followed by mergers of 'preinitials', loss or transphonologization of preinitials, and development of trochaic word structure through extensive unstressed derivational suffixation, which can then be lost in turn.

Contrived example:
- Fake Proto-Sino-Tibetan *rɔpɔŋ
- Fake Nancowry ruˈpaŋ (pretonic vowel mergers - I don't know of any ST language that preserves this stage)
- Fake Rgyalrong rpaɣ (pretonic vowel deletion)
- Fake Written Tibetan dpuŋ-pa (morphological extension-driven trochaicism without preinitial loss, some initial simplification)
- Fake Amdo Tibetan hpuŋ-pa (trochaicism with preinitial simplification)
- Fake Lhasa Tibetan puŋ-pa (morphological extension-driven trochaicism with preinitial loss)
- Fake Sikkimese puːm (post-trochaicism secondary monosyllabization)
- Fake Cantonese poŋ (preinitial loss without compensation)
- Fake Mandarin puŋtsɿ (preinitial loss without compensation, trochaicism)
- Fake Vietnamese voŋ (preinitial loss with conditioned lenition of initial voiceless plosives)
- Fake Tangut pã˞ (preinitial [and final] loss with transphonologization)

Could do something similar. Best precedent is probably Germanic, but there it hasn't progressed to the point of general monosyllabization. You'd need to think about what would happen to the resulting unstable clusters. Probably loss (with or without transphonologization) in many cases. Restricted inventory of coda consonants would make it easier - then you could have -p-n -t-n -k-n > ˀm ˀn ˀŋ etc. Metathesis is also an option.

[dɮ the phoneme]

French isn't like Latin any more!

Right, I just meant the way vowel loss would interact with the metrical structure would be very French-esque, and maybe result in a similar looking outcome.

This seems like mostly vowel loss in initial syllables. Are you saying to do the same thing in reverse (start out with trochaic stress and lose post-tonic vowels, then (potentially) switch to an iambic system via unstressed prefixation)?

[Nortaneous]

This seems like mostly vowel loss in initial syllables. Are you saying to do the same thing in reverse (start out with trochaic stress and lose post-tonic vowels, then (potentially) switch to an iambic system via unstressed prefixation)?

Right. There's probably some asymmetry involved (wouldn't have most plosives merging to h- as happens in some Tibetic varieties, etc.) but that's manageable.

There's also the question of where lenition happens. Vietnamese had lenition of plosives to voiced fricatives / approximants conditioned by the existence of a preinitial; more common in ST is reduction of preinitial plosives to fricatives. So e.g. *tapak > vak or spak; reversing this, *tapak > tavk or tapx. Both are plausible IMO.

Contrived example:
- Proto-Shit Germanic *tapak, *tamak, *tapaŋ, *tapar, *tamar, *tap
- Shit German tapx, tamk, takŋ, tapx, tamx, tap
- Shit Danish tapp, tãkk, tãpp, tarp, tarm, taːp
- Shit Icelandic tahp, tahk, tahk, taps, tams, tap
- Shit Frisian tawk, tamʔ, tawŋ, tawr, taːm, taːp
- Shit English taːp, taːm, taːp, taːp, taːm, tap

If figuring out how to deal with unstable clusters is too annoying, you can just have words with difficult clusters be subject to increased churn and be preferentially replaced - a similar process is claimed for Russian, where words with difficult and unsonorous clusters fell out of use more readily.

[dɮ the phoneme]

Contrived example:
- Proto-Shit Germanic *tapak, *tamak, *tapaŋ, *tapar, *tamar, *tap
- Shit German tapx, tamk, takŋ, tapx, tamx, tap
- Shit Danish tapp, tãkk, tãpp, tarp, tarm, taːp
- Shit Icelandic tahp, tahk, tahk, taps, tams, tap
- Shit Frisian tawk, tamʔ, tawŋ, tawr, taːm, taːp
- Shit English taːp, taːm, taːp, taːp, taːm, tap

This looks close to what I want. This is all part of redoing the sound changes for an old conlang of mine, so I already have a number of other shifts in place that I'm happy with, including various types of lenition.

The main asymmetry that concerns me here is the [claimed] Iambic/Trochaic Law, which among other things implies that this sort of widespread syncope should be less common in trochaic systems than iambic ones. There's no trochaic version of Eastern Ojibwe, as far as I know. But at least some trochaic systems have selective post-tonic syncope based on vowel hight (Latin > French, Asia Minor Greek apparently), so maybe I can accomplish this through successive raising follow by syncope .

[StrangerCoug]

What are some plausible conditions I can impose on /t͡ɬ/ → /ɬ/? The original /t͡ɬ/ contrasts with /t͡ɬʰ/ and /t͡ɬʼ/, all of which occur only word-initially and intervocalically as far as I've thought this out. I want aspiration and ejection to block lenition, but I also want to keep some tenuis /t͡ɬ/. (It's fine if /t͡s t͡ʃ/ → /s ʃ/ would expected in the same environment; I'm asking because, unlike /s ʃ/, I don't have /ɬ/ yet and I want to split it off from /t͡ɬ/.)

Edit: Typo fix—I don't want to merge /t͡s t͡ʃ/→ /ʃ/ as I implied before the edit.

[Man in Space]

Leave it affricated following nasals (or resonants more generally).

[Qwynegold]

It couldn't be preceded by a consonant.

Just lenite it when intervocalic and unstressed, or between two unstressed syllables?

[StrangerCoug]

Just lenite it when intervocalic and unstressed, or between two unstressed syllables?

I like

[linguistcat]

Japanese had a lot of /k/s disappear at one point of its development, especially before /i u/ (tho /k/s were reinserted before /u/s in most dialects, or some verbs and adjectives had forms with a /k/ and forms with it dropped). Question: could there be a similar sound change for /t/ in some language and what vowels would it be likely to drop in front of? Or could I just have whatever conditional loss of /t/s in front of whatever vowels I wanted?

[Qwynegold]

Just lenite it when intervocalic and unstressed, or between two unstressed syllables?

I like

Now I came up with another suggestion: in the environment V_V[+closed syllable].

[Qwynegold]

Linguistcat: Huh? What exactly were the sound changes for /k/?

[Man in Space]

Linguistcat: Huh? What exactly were the sound changes for /k/?

I am wondering if he meant Ryukyuan. Some of those languages had cool stuff happen with *k.

[linguistcat]

Linguistcat: Huh? What exactly were the sound changes for /k/?

I am wondering if he meant Ryukyuan. Some of those languages had cool stuff happen with *k.

I've had trouble explaining so I'm just going to use examples. This is Standard Japanese and well attested as occurring in the Muromachi period, which is part of why I had difficulty. Another reason is that it seems to do two things that sound changes are said not to do, namely, act on specific parts of speech (verbs and verbal adjectives), and that one of the sound changes reverses itself with the exception of a few words.

/kakite/ from "kaku" (to write) > /kaite/ possibly going through [kagite] or [kaɣite]. /ojogite/ "oyogu" (to swim) > /ojoide/*

/akaki, akaku/ (red, adnominal and infinitive forms, respectively) > /akai, akau/ BUT /akau/ and many other adjectives regained the -ku form for the infinitive. However, some set phrases that used adjectives like modern <arigatou> kept the -u form and went through further sound changes to their current forms.

*notice the compensatory "voicing" although at this point "voicing" was actually prenasalization and a lot of "unvoiced" consonants had voiced but unnasalized allophones intervocalically.

[bradrn]

This seems like mostly vowel loss in initial syllables. Are you saying to do the same thing in reverse (start out with trochaic stress and lose post-tonic vowels, then (potentially) switch to an iambic system via unstressed prefixation)?

Right. There's probably some asymmetry involved (wouldn't have most plosives merging to h- as happens in some Tibetic varieties, etc.) but that's manageable.

There's also the question of where lenition happens. Vietnamese had lenition of plosives to voiced fricatives / approximants conditioned by the existence of a preinitial; more common in ST is reduction of preinitial plosives to fricatives. So e.g. *tapak > vak or spak; reversing this, *tapak > tavk or tapx. Both are plausible IMO.

Contrived example:
- Proto-Shit Germanic *tapak, *tamak, *tapaŋ, *tapar, *tamar, *tap
- Shit German tapx, tamk, takŋ, tapx, tamx, tap
- Shit Danish tapp, tãkk, tãpp, tarp, tarm, taːp
- Shit Icelandic tahp, tahk, tahk, taps, tams, tap
- Shit Frisian tawk, tamʔ, tawŋ, tawr, taːm, taːp
- Shit English taːp, taːm, taːp, taːp, taːm, tap

If figuring out how to deal with unstable clusters is too annoying, you can just have words with difficult clusters be subject to increased churn and be preferentially replaced - a similar process is claimed for Russian, where words with difficult and unsonorous clusters fell out of use more readily.

I was perusing this thread when I realised this process is in fact very similar to what I’m trying to do with my current conlang. However, I’ve been struggling to figure out the exact details of how this will work, so here’s some questions:

1. From what I can see, Proto-Germanic had consistent initial stress (though I’m very unfamiliar with the language, so please correct me if that is incorrect). This makes the syncope rule consistent — in disyllabic words, it is always the second vowel which is deleted. By contrast, my language has weight-sensitive stress, which usually ends up on the first syllable, but not always. Is posttonic vowel syncope still plausible with this stress rule, or does that make it implausible?
2. Along related lines, what happens in multisyllabic words? e.g. would *tapakar with initial stress turn into tapkar, or tapakx, or what? What about long words with secondarily-stressed syllables?
3. You list a bunch of ways in which final consonant clusters get reduced. This is of particular interest of me, as one of the biggest problems I’ve been having has been getting rid of final clusters in a nice way, without messing up the paradigms too much. (Or, more accurately, my whole goal with these sound changes to mess up some paradigms horribly, but keep the rest sane.) Do you happen to have any source going into more detail about exactly how the cluster reduction proceeded in the various Germanic languages? I feel this would help me get a better sense of what options I have. (Currently I’m deleting word-final obstruents then doing C [+cont] / [+cont] ː / _, which gives lovely results, but I’m not all that happy with that last change.)

[Pabappa]

i can certainly believe in the /k/ > [ɣ] > /Ø/ path especially given that the language at the time had very little vowel hiatus. /p/ > [ɸ] > [w] > /Ø/ was in the process of happening too but had not reached the Ø stage yet.

Another reason is that it seems to do two things that sound changes are said not to do, namely, act on specific parts of speech (verbs and verbal adjectives), and that one of the sound changes reverses itself with the exception of a few words.

i think this can be a sign of a sound change that got mixed with a grammatical substitution ... e.g. maybe /k/ only lenited in a few environments, but got generalized to others, and at the same time the lack of lenition also got generalized to environments in which it had originally occurred. there was also a -si adjective formation, right?

[linguistcat]

...
i think this can be a sign of a sound change that got mixed with a grammatical substitution ... e.g. maybe /k/ only lenited in a few environments, but got generalized to others, and at the same time the lack of lenition also got generalized to environments in which it had originally occurred. there was also a -si adjective formation, right?

Correct. -si was the original sentence final or conclusive form. The change from the -si form to only the -ki/-i form was actually a grammatical change. Like many verbs, the conclusive got replaced by the adnominal. This change itself was from the adnominal being used for emphasis and then becoming the default form. Not really related to sound changes but an interesting change nonetheless.

It does seem to me that coronal consonants in Japanese were more resistant to lenition than peripheral consonants, and I do have reasons for the reverse in this language, despite it being in close contact with Japanese and other Japonic languages at the time this sound change would be happening. But as for specific vowel that would condition this change, I'm not seeing much evidence.