Re: Stacked sound change?

On the 16th, John wrote:

<<
If a language had a sound change X->Y going on that in several
environments

(IF A), (IF B), is there any chance that the results would be cumulative

(X->Z) in case of (A AND B)??



Example: historical sound changes changed English /&/ to /O/ before /
l/ and

after /w/ ("war", "fall"). "Stacking" would here mean approximately
"wall"

becoming /wUl/ rather than /wOl/.

Note that the following analysis

1) /&/ -> /O/ if before l or after w

2) /O/ -> /U/ if between w and l

would *not* be possible, since a stacked sound change would ONLY
apply to

older /w&l/, not hypothetical older /wOl/.



Does this sound the least bit plausible?

 >>

That sounds very interesting.  Why does /&/ go to [U]
specifically, though?  How is this stacked sound change
constrained?  Just whatever seems right based on the
phones involved?

John continues:
<<
If not, how about cumulative results of two different sound changes?
Example

#2 follows.

Assuming that /d/ and /tS/ already exist:

/t/ -> /tS/ before /i/

/t/ -> /d/ between /n/ and a vowel

so /t/ -> /dZ/ between /n/ and /i/

while /ndi/ and /ntSi/ remain intact?

 >>

This isn't actually as out-of-the-ordinary as the first you
described (if that, indeed, is out of the ordinary).  This kind of
thing happens all the time.  With your specific example, though,
I'd be at least a bit surprised to find a language where /t/ > [tS] /
_[i],
but /d/ > [d] / _[i].  Nevertheless, the point it illustrates is sound.

Here's an interesting one.  Imagine:

(1) /t/ > [tS] / _[i]
(2) /ts/ > [tS] / _[i]
(3) /s/ > [ts] / C[+nasal]_

Now, let's say (1) and (2) happened much earlier on in the language.
You could have the following situation:

(a) *mati > matSi
(b) *matsi > matSi
(c) *masi > masi
(d) *manti > mantSi
(e) *mantsi > mantSi
(f) *mansi > mantsi

Add intervocalic voicing for fricatives only, but also obstruent
voicing after nasals...

(a) *mati > matSi
(b) *matsi > matSi
(c) *masi > mazi
(d) *manti > mandZi
(e) *mantsi > mandZi
(f) *mansi > mandzi

Well, maybe.  It would depend on the status of the new sound
[tS] (the phonemic status), and also the status of the intervocalic
[t]...  Well, no, I guess it wouldn't, since [s] is an obstruent.
Anyway, yeah, that works.

(P.S.: I'm back from a very restful vacation.)

-David
*******************************************************************
"A male love inevivi i'ala'i oku i ue pokulu'ume o heki a."
"No eternal reward will forgive us now for wasting the dawn."

-Jim Morrison

http://dedalvs.free.fr/