Re: Synthesis of many topics

On Sat, 29 Apr 2000, Nik Taylor wrote:

> dirk elzinga wrote:
> > Chemehuevi, a Numic language of the Lower Colorado River Valley,
> > underwent the change from Proto-Numic *m to w intervocalically:
>
> What evidence exists for that change?  Presumably cognates with /m/, but
> how is it known that /w/ wasn't the original form?
Here's a partial Uto-Aztecan cognate set for *tama 'tooth':

Shoshoni     tama
Tubatulabal  taman-
Luiseño      tama-
Hopi         tama
Guarijio     tame
Cora         tame
Nahuatl      tlan-

It's obvious that Proto-Uto-Aztecan had medial *m. It's also
obvious that Chemehuevi cognates have medial /w/. The Shoshoni
word shows lenition of intervocalic /m/ to [w~]. Southern
Paiute, closely related to Chemehuevi, shows the same lenition.
If Chemehuevi also had [w~], it is a simple matter to denasalize
it to yield [w].

Dirk

--
Dirk Elzinga
dirk.elzinga@m.cc.utah.edu