Dakotan ''wichasha'' 'man'.

[Koontz John E]

On Thu, 28 Apr 2005, Koontz John E wrote:
> I guess there's no reason why wic^ha-s^a 'man-red' wouldn't work
> analytically, though I'd feel better if I could explain the variants
> wic^hasta, etc., in the same breath.  ...

Wic^has^ta.  (Though I think s^ > s in Stoney?)

> One possibility that occurs to me is that wic^has^a is a punning
> substitution for wic^hasta.  ...

Or, rather, for wic^has^ta.

Or even just a lexical substitution of 'red man' for 'man + ???'.  Note
that -s^ta might be preceived as resembling -ska 'white'.

> If you look at Dhegiha forms, which as far as 'person' proper aren't
> cognate, you'd expect second elements in 'person' compounds to be
> something meaning 'little', cf., nikkas^iNga, s^iNgaz^iNga, etc.
>
> As far as -sta, could this be a fricative-grading variant of -xta in the
> sense 'real, true'?

The actual Dakotan enclitic here is =xc^A, though it's probably reasonable
to see it as deriving from earlier *=xta.

I should add that Bob Rankin, intrigued by this possibility, looked
through the SA Buechel files for any trace of putative *=s^ta with no luck
at all.  I looked in various grammatical references for something like
=sta or =s^ta in any sense with no luck either.  So this remains a rather
far-fetched suggestion.

Constantine Chmielnicky's observation that there are parallels for
reduction of Sa -s^tV to Te -s^V (in accentually weak locations, it
seems), is probably better supported and more plausible.