Winnebago Nominal Ablaut (Re: Winnebago -ga in Kinterms)

[Koontz John E]

On Sun, 3 Mar 2002, Koontz John E wrote:
> The last of these, 'big turtle' (= snapping turtle?) seems to be another
> case of -e- appearing before -ga, since this is keecaNk' in Miner.

So, as far as I can see there are now two examples, perhaps fossilized, of
0 / elsewhere ~ e / __ ga with Winnebago nouns in -k, kec^aNk and
was^c^iNk.  It might work better syncronically to treat this as ga /
elsewhere ~ ega / k __.  It's a sort of left-handed nominal ablaut, in the
sense that Dakotan nominal ablaut essentially involves a ~ e ~ 0, though
with well-developed e ~ 0, a ~ 0, a ~ e, and a ~ e ~ 0 subcases.  Also,
while Dakotan has the 0-alternant in compounds and reduplication,
Winnebago has it as the unmarked (uncompounded) form.

The occurrence of -e- in these contexts is not at all surprising as
Winnebago (by comparison with Ioway-Otoe) seems to lose final -e after C
(but not CC), as in naNaNc^ge' 'heart'.  In essence the =ga forms an
environment in which this loss of e is prevented.  For example, compare
Omaha-Ponca mas^tiNge ~ mas^c^iNge (in many of the JOD texts) 'rabbit' and
kke( )ttaNga 'big turtle'.  The latter has -ga < *-ka), but it appears
that in Winnebago (and Ioway-Otoe) *a > e after any velar obstruent.
This is one reason I was interested in Winnebago deega' 'uncle'.  The ga
looked irregular.  So, I was happy to be (probably) able to reduce it to
the usual exception for =ga.

JEK