Wa-/Wicha history.

[R. Rankin]

I'm afraid a comparative perspective on the wa- vs. wicha- problem doesn't
provide as much useful information as one might wish.  The problem is that
wicha- exists only in the Dakotan subgroup, but it seems to be found throughout
Dakotan.  This means that it's history is difficult to trace within Dakotan and
impossible to trace outside it.

It is probably realistic to reconstruct *wa- within Mississippi Valley Siouan at
the very least.  But its precise semantics/function would need to be determined
by a careful grammatical comparison of Dhegiha (which John has pretty much
covered already), Chiwere, Hocank/Winnebago, etc.  The implication seems to be
that, wherever you find wicha- used today in the Dakotan verb complex, you may
have had wa- at an earlier time.  This may not be 100% so, but as a working
hypothesis, it's a start.  To the extent that it is true, wicha- can be said to
have replaced wa- first in human and then animate contexts, i.e., wicha-
(always??) spreads at the expense of wa-.

We would then expect to be able to seriate the derivational vocabulary to a
certain extent (e.g., Konstantin's nice listing).  Older derivations would be
expected to show wa- where, today, we might expect wicha-.  Broadly, derivations
with pan-MVS cognates should have (or have had) wa-.  Those without cognates
outside of Dakotan should have the wa/wicha split.  I believe John said
something about like this in a recent posting.  More recent derivations should
yield wicha-, not wa-, in animate contexts.  But, of course, older constructs in
wa- may have been reanalyzed with wicha- analogically.  We would not generally
expect replacement of wicha- with wa-, however.  Replacement should be one-way.

Since grammar change, as opposed to sound change, is lexically gradual
(analogical), we might expect (1) that different Dakotan dialects will show
different distributions of wicha- vs. wa- (this is where Linda's Assiniboine may
come in handy, also Stoney, as well as the better-attested dialects), (2) that
different generations of Dakotan speakers may show different distributions of
wicha/wa, so that, (3), using these distributional differences, we may be able
to trace the categorial spread of wicha- within Dakotan.

This is all very common-sensical, so I apologize for its obviousness.  Anyway,
that's the perspective of a comparativist.  Now all that is needed is somebody
to do all the related work!!  A great doctoral dissertation topic.

Bob