Zompist Bboard Again

Posted: **Fri May 24, 2019 9:50 am**

I'm sorry, I forgot to reply to this question! Sercquiais has [ʙ̥r], apparently from /fr/ or something?

Posted: **Fri May 24, 2019 11:23 am**

Frislander wrote: ↑Fri May 24, 2019 7:02 am The other thing I would like to note is that it's not entirely clear to me why a trilling action made by the tongue should "jump" to one made by the lower lip like this.

I had a teacher in high school who had br > ʙ

Posted: **Fri May 24, 2019 12:42 pm**

Nortaneous wrote: ↑Fri May 24, 2019 11:23 am
Frislander wrote: ↑Fri May 24, 2019 7:02 am The other thing I would like to note is that it's not entirely clear to me why a trilling action made by the tongue should "jump" to one made by the lower lip like this.
I had a teacher in high school who had br > ʙ

This keeps making me whisper [ðə ˈʙɪɾɪʃ ˈɛmpʰajɹ̩] or [ˈʙɪɾɪʃ ˈʙɪɾɪʃ ˈʙɪɾɪʃ] to myself.

Posted: **Fri May 24, 2019 1:50 pm**

Vijay wrote: ↑Fri May 24, 2019 12:42 pm
Nortaneous wrote: ↑Fri May 24, 2019 11:23 am
Frislander wrote: ↑Fri May 24, 2019 7:02 am The other thing I would like to note is that it's not entirely clear to me why a trilling action made by the tongue should "jump" to one made by the lower lip like this.
I had a teacher in high school who had br > ʙ
This keeps making me whisper [ðə ˈʙɪɾɪʃ ˈɛmpʰajɹ̩] or [ˈʙɪɾɪʃ ˈʙɪɾɪʃ ˈʙɪɾɪʃ] to myself.

Try "break" - this was woodshop. Which for whatever reason was a requirement in the sci/tech program.

Posted: **Sun May 26, 2019 6:42 am**

Vijay wrote: ↑Fri May 24, 2019 9:50 am I'm sorry, I forgot to reply to this question! Sercquiais has [ʙ̥r], apparently from /fr/ or something?

Is there a non-Wikipedia source for this?

Posted: **Tue Jun 04, 2019 6:40 pm**

What can I do with the English rhotic in terms of diachronics to make it go away in all clusters and get replaced with something else otherwise? I'm slowly coming to the conclusion that my current /ɹ/ → /ˤ/ → vowel quality and /ɹ/ → /ʕ/ → /ɣ/ → /g/ → /ŋ/ shift sequences are not really something I like. And since I'm no longer getting the velar nasal out of the rhotic, how else can debelop it so that it's only found in onsets and not in codas?

Posted: **Tue Jun 04, 2019 7:11 pm**

Knit Tie wrote: ↑Tue Jun 04, 2019 6:40 pm What can I do with the English rhotic in terms of diachronics to make it go away in all clusters and get replaced with something else otherwise? I'm slowly coming to the conclusion that my current /ɹ/ → /ˤ/ → vowel quality and /ɹ/ → /ʕ/ → /ɣ/ → /g/ → /ŋ/ shift sequences are not really something I like. And since I'm no longer getting the velar nasal out of the rhotic, how else can debelop it so that it's only found in onsets and not in codas?

Dorsey's Law in Ioway-Oto: *CRV₁ > CV₁RV₁.

Posted: **Tue Jun 04, 2019 7:15 pm**

dhok wrote: ↑Tue Jun 04, 2019 7:11 pm Dorsey's Law in Ioway-Oto: *CRV₁ > CV₁RV₁.

Friggin genius! Thanks!

What do you say I should change the approximant itself into?

Posted: **Tue Jun 04, 2019 7:16 pm**

Knit Tie wrote: ↑Tue Jun 04, 2019 7:15 pm
dhok wrote: ↑Tue Jun 04, 2019 7:11 pm Dorsey's Law in Ioway-Oto: *CRV₁ > CV₁RV₁.
Friggin genius! Thanks!

What do you say I should change the approximant itself into?

*l > r, *r > j, *j > r and *w > j are all attested.

Posted: **Tue Jun 04, 2019 8:07 pm**

dhok wrote: ↑Tue Jun 04, 2019 7:16 pm
*l > r, *r > j, *j > r and *w > j are all attested.

So ɹ → j unconditionally is plausible?

Posted: **Wed Jun 05, 2019 2:54 am**

I'm starting with a two-tone system, where tones are assigned to morae rather than syllables (i.e. basically what Navajo does), and I want to turn coda [s] into tone. I'm thinking something like Vᵀs > Vːᵀᴸ (where T is either tone, and L is of course low tone). Basically coda [s] turns into a floating mora baring low tone. I think this is plausible (please correct me if I'm wrong), but in any case I'm not sure what to do with [s] after a vowel which is already long. Maybe add another mora to create overlong vowels? I don't like the aesthetic of that at all —would it be realistic for [s] to just delete in this context.

On another note, how does lf ls lx ll > ɬ(ː) look?

Posted: **Wed Jun 05, 2019 4:18 am**

s → h → ᴸ certainly works. If there's a problem with compensatorily-lengthening a short vowel but not a long one, then I don't know what it is.

When you say the result is a floating tone, though---do you mean it doesn't end up linked to the final vowel? I'd have thought Vːᴴᴸ would give you a falling tone; maybe Vːᴸᴸ would violate an obligatory contour principle, but I don't know how floating the new tone would solve that issue.

(Disclaimer: not at all an expert.)

Posted: **Wed Jun 05, 2019 9:31 am**

Knit Tie wrote: ↑Tue Jun 04, 2019 8:07 pm
dhok wrote: ↑Tue Jun 04, 2019 7:16 pm
*l > r, *r > j, *j > r and *w > j are all attested.
So ɹ → j unconditionally is plausible?

Yes

Posted: **Wed Jun 05, 2019 9:48 am**

I'd rather expect coda /s/ to give a high tone.

Posted: **Wed Jun 05, 2019 10:02 am**

WeepingElf wrote: ↑Wed Jun 05, 2019 9:48 am I'd rather expect coda /s/ to give a high tone.

In Chinese it's supposed to have resulted in a 去聲 (falling tone).

Posted: **Wed Jun 05, 2019 10:12 am**

Max1461 wrote: ↑Wed Jun 05, 2019 2:54 amOn another note, how does lf ls lx ll > ɬ(ː) look?

Perfectly plausible.

Posted: **Wed Jun 05, 2019 11:01 am**

WeepingElf wrote: ↑Wed Jun 05, 2019 9:48 am I'd rather expect coda /s/ to give a high tone.

Tonogenesis can do whatever, since a minor shift in phonetic detail can produce a completely different tone - in this case, s > h > H, s > h > ɦ > L.

Posted: **Wed Jun 05, 2019 10:28 pm**

Posted: **Thu Jun 06, 2019 4:39 pm**

akam chinjir wrote: ↑Wed Jun 05, 2019 4:18 am s → h → ᴸ certainly works. If there's a problem with compensatorily-lengthening a short vowel but not a long one, then I don't know what it is.

My concern is, what how is tone affected when [s] deletes from a syllable which is already bimoraic without it? I'm trying to achieve something like the following (please excuse the shit diagrams):

Code: Select all

		    σ		σ	...	
		   / \		|
		 /     \	|
		T	L	|
		|	|	|			
		μ	μ	|
		|	|	|
		V	s	C	...
		
		→
		
		    σ		σ	...	
		   / \		|
		 /     \	|
		T	L	|
		|	|	|			
		μ	μ	|
		|	/	|
		|    /		|
		V		C	...

But I'm not then sure what happens when we have:

Code: Select all

		    	σ		σ	...	
		   /	|    \		|
		 /  	|      \	|
		T	T	L	|
		|	|	|	|			
		μ	μ	μ	|
		|    /		|	|
		 V		s	C	...

Since a vowel can't be trimoraic, and tones are constrained to one per mora. I don't really want to drop that constraint; is it a plausible alternative for the final tone to just drop entirely?

Posted: **Thu Jun 06, 2019 8:54 pm**

When consecutive morae have the same pitch, it's likely that you've got a single tone linked to both morae. So, for your initial situation, you've likely got something like one of these:

Code: Select all

      L       H   L
     /|\      |\  |
    μ μ μ     μ μ μ
    |/  |     |/  |
    V   s     V   s

If you lose the s, the following results are reasonable:

Code: Select all

      L       H   L
     /|       |  /
    μ μ       μ μ
    |/        |/
    V         V

(Aside: I was wrong to think you might end up with a new violation of an obligatory contour principle.)

(Further aside: I've omitted the σ, since I don't see a way to include them without a third dimension---it's not the tones that group into syllables.)

Zompist Bboard Again

Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread

Re: Sound Change Quickie Thread