=be Plural/Proximate in Kaw; =pe Plural/Proximate in Osage

[Koontz John E]

On Wed, 31 May 2006, Rankin, Robert L wrote:
> I've always suspected that the male Kaw form was probably -abi, just as
> in Omaha, and, as you say, that -abe is feminine -abi+(y)e.  I think
> Justin ran across -abi in some of Dorsey's Kansa notes at the NAA just a
> few days ago.  He can clarify that -- I can't remember what speaker was
> being quoted, but it was undoubtedly a man.

> Truth is, however, that we don't have very many instances of anything
> but the -abe form even from males in the 19th century.  Kansa -e and
> Omaha -i are not a regular vowel correspondence in any event.

Agreed.  I think the Omaha-Ponca =bi and =i represent the *=pi lineage,
matching the contextually resgtricted cases of =bi in Ks, =pi in Os, and
=wi in Quapaw.  The forms =be, =pe, and =we in these latter involve
something additional added, and I was wondering how good a handle we have
on it.

Secondarily, of course, it would be interesting to know the environments
in which the two alternates occur.  I think one of your Kaw or Quapaw
Lexicons, I forget which at the moment, says that the -i variant occurs in
subordinate clauses.

In OP [pace Larson] the major variants are =i and =bi, conditioned by
direct vs. indirect reportage, among other things.  In effect, it appears
that there are a series of micro-environments in which reflexes of *=pi
survive, different in each language perhaps, while =i on the one hand and
*=pe on the other, replace it as the more productive variants.  I don't
mean to imply that =i and =pe have exactly the same distribution, of
course.

For the non-Dhegihanists, note that (most) Omaha (but not Ponca? and not
all Omaha?) seems to lose =i entirely in most modern usages.  It is
retained when certain other enclitics follow it, e.g., =the EVIDENTLY or
=ga MALE-IMPERATIVE.

All Dhegiha languages have reflexes of *=p (i.e., *=pi with -i elided) in
the context of the plural/proximate negative, e.g., adha=b=az^i 'he didn't
go; they didn't go'.

At the moment, though I don't have much information on *=pe vs. *=pi in
Ks, Os, and Qu, it looks like this might be another example of an isogloss
separating these from Omaha-Ponca.   Other such isoglosses would be the
treatment of the regular dative and the dative of *k-stems.

The Regular Dative Situation

        Southern Dhegiha   Omaha-Ponca

Dat A1    aa-                ee-
Dat A2 *dhaa-              dhee-
Dat A3  *kii-               gii-

The Southern Dhegiha pattern merges the dative with the *ka- STRIKING
instrumental in the A1 and A2 forms.  The third persons still contrast
with *ka- vs. *kii-.  I think changing that *ka to *kii is part of what
happened to shuffle the prefix system so thoroughly in Winnebago and
Ioway-Otoe.  The OP pattern is a bit more like Dakota (though the direct
analogy is with the Dakota suus forms).

The *k-stem Situation (with 'to make')

       Southern Dhegiha   Omaha-Ponca
A3     *kaa'ghe           gaa'ghe
Dat A3 *ks^ii'ghe         gia'ghe

(I think we don't actually have the dative forms for 'make' in Quapaw.)

There are a few other patterns like this.

One is *h (?) in *(h)aNp- 'day' and *(h)a INTERROGATIVE-DEMONSTRATIVE.

I think the Southern Dhegiha group all have h-, cf. Osage haNaNpa 'day',
hanoN 'how many', while OP lacks it, cf. OP aNba' day', and anaN 'how
many'.  Again, OP is like Dakotan, which has aNp(etu).

There are also some differences in the set of positional articles between
Southern Dhegiha and OP.  SD has *dhiNke (not same as *dhiNkhe), while OP
has dhaN.

For *W and *R OP has m and n (ma 'snow', nu 'potato') while SD has
something more like *p and *t, e.g., Os pa and to.  This may not be that
significant.

I suppose there are other things that pattern like this, and some of
these, e.g., the articles or the treatment of *W/*R, may not prove very
reliable or significant.