zompist bboard

I don't think it's odd.

Thanks.

Suppose I have a conculture where a significant minority speaks a conlang (let's call it Language A for the purposes of this question) with a (C)V(C) syllable structure whose consonant inventory is pretty much Skolt Sami plus /ʔ θ ç h/, but has only /a ə ɨ/ for its phonemic vowels. It's in contact with speakers of the conculture's majority language (call it language B), which has /a e i ə ɨ o u/ for its vowel inventory and otherwise isn't structurally far off from SAE. What allophonic processes are plausible for Language A that might cause speakers of Language B to perceive backness distinctions that don't really exist in Language A? (In other words, what might cause Language B's speakers to hear Language A's /ə/ as closer to a Language B /e/ or /o/, and similarly for Language A's /ɨ/?)

StrangerCoug wrote:Suppose I have a conculture where a significant minority speaks a conlang (let's call it Language A for the purposes of this question) with a (C)V(C) syllable structure whose consonant inventory is pretty much Skolt Sami plus /ʔ θ ç h/, but has only /a ə ɨ/ for its phonemic vowels. It's in contact with speakers of the conculture's majority language (call it language B), which has /a e i ə ɨ o u/ for its vowel inventory and otherwise isn't structurally far off from SAE. What allophonic processes are plausible for Language A that might cause speakers of Language B to perceive backness distinctions that don't really exist in Language A? (In other words, what might cause Language B's speakers to hear Language A's /ə/ as closer to a Language B /e/ or /o/, and similarly for Language A's /ɨ/?)

Fronting when adjacent to palatals, rounding when adjacent to labials, backing adjacent to velars (less likely; I think this is attested but I'm not certain, it would certainly make sense).

Sounds that velar to glottal and in between tend to cause backing, while sounds that have some palatal quality tend to front. Labials cause rounding.

These are just tendencies though: Velars are less likely to cause backing than uvulars and other sounds further back, and think I saw at least one natlang grammar that said uvulars caused rounded front vowels!

[h] already exists as an allophone of /x/.

Here's an example I made up:

1. After /p b m f v/ ⟶ [ɒ̈ ɵ̞ ʉ]
   Before /t d n t͡s d͡z t͡ʃ d͡ʒ s z θ ð r l/ ⟶ [ä ə ɨ]
   Before /ʃ ʒ j/ ⟶ [a e̞ i]
   Before /c ɟ ɲ ʃ ʒ ç ʝ ʎ/ ⟶ [ɛ e i²]
   Before /k ɡ ʔ ŋ x ɣ h w/ ⟶ [ɑ ɤ̞ ɯ]
   Before /w/ ⟶ [ɒ o̞ u]
   So now we have /ɨ/ [i i² ɨ ʉ ɯ u], /ə/ [e e̞ ə ɵ̞ ɤ̞ o̞] and /ä/ [ɛ a ä ɒ̈ ɑ ɒ]
2. After nothing ⟶ [ɪ i ɪ̈ ʊ̈ ɯ̽ u e e̞ ə ɵ̞ ɤ̞ o̞ ɛ æ ä ɒ̈ ɑ ɒ]
   After /p b t d k ɡ ʔ/ ⟶ [i i ɨ ʉ ɯ u e e ɘ ɵ ɤ o ɛ æ ɐ ɞ̞ ʌ̞ ɔ̞]
   After /c ɟ/ ⟶ [i i i y ɯ̈ ü e e e ø ɤ̈ ö ɛ æ æ ɶ̽ ɑ̽ ɒ̽]
   After /m n ŋ r l/ ⟶ [i̞ i ɨ ʉ ɯ u e e̞ ɜ ɞ ʌ ɔ ɛ æ ɐ ɒ̈ ɑ ɒ]
   After /ɲ/⟶ [i i y ɯ̈ ü e e ë ø̈ ɤ̞̈ ö̞ ɛ æ æ̈ ɶ ɑ̈ ɒ̈²
   After /t͡s d͡z t͡ʃ d͡ʒ s z ʃ ʒ f v θ ð x ɣ h/ ⟶ [i̞ i ɪ̈ ʊ̈ ɯ̽ ʊ e̞ e̞ ɜ ɞ ʌ ɔ æ a ä ɒ̈ ɑ ɒ]
   After /ʃ ʒ ç ʝ j ʎ/ ⟶ [i ɪ ʏ ɯ̽ ʊ e e ë ø̈ ɤ̞̈ ö̞ ɛ a a ɶ ɑ̈ ɒ̈²
   After /w/ ⟶ [i̞ i ɪ̈ u u u e̞ e̞ ɜ ɔ ɔ ɔ æ a ä ɒ ɒ ɒ]
3. Assume [i ɨ y ɪ ʏ ɪ̈ e ø ë ø̈ ɘ e̞ ə ɛ ɜ æ æ̈ a ä] as Group 1 and the other values as Group 2
   All vowels in Group 2 become Group 1 vowels if the first vowel of the word is a Group 1 vowel and vice versa
4. Language B borrows the word
   [i y ɨ ʉ ɯ̈ ü ɯ u ɪ ʏ ɪ̈ ʊ̈ ɯ̽ ʊ e ø ë ø̈ ɘ ɤ̈ ö ɤ o e̞ ə ɵ̞ ɤ̞̈ ö̞ ɤ̞ o̞ ɛ ɜ æ æ̈ ɶ̽ ɐ ɞ̞ ɑ̽ ɒ̽ ʌ̞ ɔ̞ a ɶ ä ɒ̈ ɑ̈ ɒ̈² ɑ ɒ] ⟶ /i i~u ɨ u ɨ u i u ɨ u ɨ u ɨ u e ə e~ə ə ə ə o ə o e ə ə ə o e ə ä ä ə ä ä ä ä ä ä ä ä ä ä ä ä/

mèþru wrote:[h] already exists as an allophone of /x/.

I had intended /h/ and /x/ to be separate phonemes, but hey, maybe they merge at some point

Edited to add: Or even better, Language B speakers can't hear the difference between /x/ and /h/ and they get merged when it makes it way over to Language B. I'm not willing to say no...

Edited post

When it comes to bilabial trills, I know they can originate from /b/ before rounded vowels and from /mb/ clusters, but does the alternative origin of bilabial stop + flap/trill sound reasonable as well?

Sure

How would I be able to make /ɹ̩/ and /az/ cognates?

احمکي ارش-ھجن wrote:How would I be able to make /ɹ̩/ and /az/ cognates?

Language A: *əz > əɹ > ɹ̩
Language B: *əz > az

Most lenitions of /s/ seem to take /z/ or /h/ as initial pathways; i.e.

/s/ > /h/ > /∅/
/s/ > /z/ > /ʐ/ or /r/ or /ɹ/

Can anyone think of any other outcomes for lenited /s/, or any other initial pathways? Especially in a voiced environment?

How would you approach the following ones?

/k/ → /tsk/ // V[+short]_{a o u}, V[+short]_$
/k/ → /ts/ // _{a o u}
/θ/ → /tx/
/ks/ → /x/

Pole, the wrote:How would you approach the following ones?
/θ/ → /tx/

θ ---> ṭ (dental) ---> t (known from some AmEng dialects) ---> tʰ ----> tx
Any preeixisting /t/ would have to be sheltered from the change by a Duke-of-York sound shift, since /tx/ without bare /t/ would be very peculiar.

/ks/ → /x/

Similar. ks----> kx ("strongly aspirated" ) -----> x


Others not likely without similar shifts affecting /k/ in other environments. How big is the pohonology?

Soap wrote:Others not likely without similar shifts affecting /k/ in other environments. How big is the pohonology?

It looks roughly like that:
/m n/
/p b t d (tʃ) k ɡ/
/f v (θ) s z (ʃ) x/
/l r j w/

Later on, /tʃ ʃ/ → /tʃˠ ʃˠ/ → /tsx sx/.

Edit:
Would it be plausible to have /θ/ → (fortition) /tθ/ → /tx/ instead?

And also /k ɡ/ → /kː ɡː/ // [V+short]_ → (dissimilation) /tk dɡ/ → /tsk dɡ/?

Pole, the wrote:

Soap wrote:Others not likely without similar shifts affecting /k/ in other environments. How big is the pohonology?

It looks roughly like that:
/m n/
/p b t d (tʃ) k ɡ/
/f v (θ) s z (ʃ) x/
/l r j w/

Later on, /tʃ ʃ/ → /tʃˠ ʃˠ/ → /tsx sx/.

Edit:
Would it be plausible to have /θ/ → (fortition) /tθ/ → /tx/ instead?

And also /k ɡ/ → /kː ɡː/ // [V+short]_ → (dissimilation) /tk dɡ/ → /tsk dɡ/?

How about /kː ɡː/ → /t͡ʃk d͡ʒg/ // [V +front]_, followed by /t͡ʃ d͡ʒ/ → /t͡s d͡z/? You don't need intermediate steps for /tʃ ʃ/ → /tsx sx/, [t͡s] and [t͡ʃ] are in free variation in some Dravid language I can't remember.





Given that vowels have inherent slight tone differences - going from highest to lowet [ɛ] [a] [ɔ] - how likely is that tone difference of a second unstressed vowel turning into pitch accent, while the vowels themselve reducing to [ɪ] [ɐ] [ʊ]? A quite common word type is CV(J)CV, stress is on first syllable.

احمکي ارش-ھجن wrote:How would I be able to make /ɹ̩/ and /az/ cognates?

Boring but straightforward version: /ɹ̩/ < *ɹ̩z > /az/.

Porphyrogenitos wrote:Can anyone think of any other outcomes for lenited /s/, or any other initial pathways? Especially in a voiced environment?

Nganasan has /j/ as the voiced alternant of /s/ (the other pairs are /h/ : /b/, /t/ : /ð/, /k/ : /g/). This is from earlier palatal *ś, though.

Pole, the wrote:/k ɡ/ → /kː ɡː/ // [V+short]_ → (dissimilation) /tk dɡ/ → /tsk dɡ/?

Geminate dissimilation is not really a thing that happens.

For a bit more of a roundabout, this should be workable: kː gː > ʰkː ɠ > hk ʔg > hk ɦg > xk ɣg > çk ʝg > tɕk dʑg > tsk dg.

Tropylium wrote:

Pole, the wrote:/k ɡ/ → /kː ɡː/ // [V+short]_ → (dissimilation) /tk dɡ/ → /tsk dɡ/?

Geminate dissimilation is not really a thing that happens.

It might be anecdotal, but Polish miękki, lekki → dialectal/colloquial miętki, letki.

What about
kː ɡː → k.k ɡ.ɡ → t.k d.ɡ

Although I think that either both voiced and voiceless plosives should affricate or neither do.

I think I could see that happening.

Pole, the wrote: Would it be plausible to have /θ/ → (fortition) /tθ/ → /tx/ instead?
?

yes, I should've thought of that first,sorry.

I'm on phone only for time being,sorry for messy typing.

Has there been recorded an unconditional change of [ɔ/o] to [œ/ø]?

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote:Has there been recorded an unconditional change of [ɔ/o] to [œ/ø]?

More or less, yes, at least if it's part of a larger vowel system where there's another shorter/laxer o-like consonant. E.g. in some English varieties, such as the Baltimore dialect, /oʊ/ is fronted almost all the way to /øʊ/ or even /eʊ/. Meanwhile the shorter, laxer /ɔ/ stays behind. This is because long vowels tend to front or be raised.

So, is there a possibility it'd become fronted without any other o-like vowel? I'm thinking of something similar to the change in French from to [y].

What are ways to develop word-initial geminates of obstruents (although they would be phonetically realized as tense)? Looking mostly for /p t d k s/.