Postulated wo- 'food' in Dakotan (Re: argument structure k'u etc.)

[Koontz John E]

On Thu, 21 Apr 2005, REGINA PUSTET wrote:
> I'm now quoting Buechel in detail on ok'u (p. 393): 1. 'to lend anything
> to one. 2. (of k'u). 'to give to, e.g. food; to give a portion to'.

> There is not much to be done about the hypothesis that wo- means 'food'.
> Either the prefix exists, or it doesn't, at least you can always posit
> such an element. However, I recommend eliciting semantic combinations
> containing wo- 'food' in conjunction with verbs which don't have initial
> o-. My guess is that this is not grammatical.

My guess is that the problem with this approach is that any wo- in this
sense of 'food' probably wouldn't be productive enough to elicit arbitrary
examples of it.  So it may come down to a philological exercize using
whatever examples Boas & Deloria cite to support themselves, plus any
others in sources like Riggs and Buechel.  This is something that one of
us could probably do fairly quickly.

> (Another hint: unlike a true wa-o contraction, as a hypothetical
> classificatory prefix, wo- 'food' shouldn't carry stress).

Assuming that wo- 'food' would be from attested wotA, a attested
peculiarity of yutA, and not from woyutA per se, we'd probably be dealing
with *wol-k?u > wo'k?u, right?  Would the stress on a compound of this
form be wolk?u'?  I'm more than a little vague on the finer points of
Dakota accentuation, but I seem to recall that compounding is one of the
contexts in Dakotan where accent is not restricted to the "first root" in
a word, e.g., c^aNlwa's^te, if I recall the example correctly.

> But there is nothing to be done about the hypothesis that wo- = wa-o
> either, because it reflects a highly regular contraction process in the
> language.

And in Mississippi Valley Siouan generally, to the point that it might be
treated less as a process in Dakotan than as a still more or transparent
process of Proto-Mississippi Valley Siouan.  But when is something like
this a process and when is it an artifact?

> Given the fact that some of us have stated before, namely that a
> reduction of woyute 'food' to wo- 'food' is unlikely because of the
> phonetic complexity involved, my vote is clearly and emphatically in
> favor of the wa-o hypothesis.  I realize that this is just a minor
> issue, but I feel it deserves clarification.

I quite agree that this deserves clasificaiton.  It is the sort of
superficially minor point on which important morphological issues can
turn, and it is also a very longstanding conundrum introduced by Boas &
Deloria themselves.  Do we accept or reject their account of wo- 'food'?
This is a good question for all you Dakotanists and I'm interested to see
how it plays out.

It seems to me that the general issue is, does Dakotan display an
occasional pattern of more or less arbitrary or drastic reductions of
incorporated elements to which we can appeal in resolving the structure of
forms like wok?u 'give, e.g., 'giving portions' rather than as a gloss on
(w)o-?

Without having reviewed other proposed wo- 'food' forms, I would be
doubtful that there is a regular formation of this nature, though it's not
inconceivable.  But, Regina (and Jan and Bruce et al.) I wonder how you
feel about the possiblity of an irregular reduction of hypothetical
*wolk?u to wok?u?  Would accentuation resolve this one way or another?
Would you consider *wolk?u itself plausible?