Funny W

[Koontz John E]

On Thu, 26 Oct 2006, Rory M Larson wrote:
> So *R is the phoneme behind the Dakhota/Nakhota/Lakhota divide?  I.e.,
> *Rakhota ?  What about Assiniboine and Stoney?  Where do they come out?

*R > n in Assinboine and Stoney.  In fact, since bl (< *p-R < **p-r) is md
in Santee and mn in Assiniboine and Stoney, I've wondered if *R was */n/
in Proto-Dakotan.  Richard Carter once mentioned to me that he'd had
similar thoughts (which would have been earlier than mine!).

> And *R => d in IOM and Winnebago, n in Crow (-Hidatsa ?),

I'm not sure about Crow-Hidatsa, since in both of them n seems (today) to
be a conditioned allophone of /r/.  Not the same conditioning in the two
cases.  In Crow /r/ is [d ~ l ~ n], all written separately in the popular
orthography.  I think n occurs finally and in geminate sequences.

> n in Biloxi and Tutelo, l in Ofo, t in Quapaw and Osage, d in Kansa, and
> of course n in OP.

In some of these cases the reflex is identical with the reflex of *r or
*t.  In fact, on reflection, only Dakotan and Winnebago have fairly
unique developments:

PS   Da   Wi
*r   y    r
*R   l    d
*t   t    j^

> Is Mandan unknown?

I think it's /r/ (like *r), but, of course, /r/ is [n] before nasal
vowels.

> And *W => w in Dakotan and Winnebago, I believe, which isn't hugely helpful
> to the model I proposed earlier today.

It's m in Omaha, e.g., in outer instrumentals like mu= 'by shooting' and
ma= 'by cutting'.  And *R is n, as in na= 'by heat, or spontaneously', vs.
the inner instrumental naN- 'by foot'.

I'll let Bob supply *W = *w-w data, but I think some of the outer
instrumentals have #VwV- in some languages.