Dhegiha -akhe.

[Koontz John E]

On Thu, 20 Sep 2001, Rankin, Robert L wrote:
> Rory: Another argument against it is that akha', as a
> positional, ought to come *after* the (-i | -bi)
> particle if it occurs at the end of the sentence, as
> "the" does in the fairly frequent form [Sentence]
> bi=the'=ama. I think the examples you cite above
> indicate that akhe' functions as a non-ablauting -e
> stem verb, however it was derived.
>
> Bob: 1. "as 'the' does in ... frequent form"  Actually, this
> is the "the" that is not historically a positional I
> think. It's the one that some linguists have translated
> 'narrative' and which has the cognate in Hidatsa "rahe"
> 'rumored'.  John's 2000 Siouan Conf. paper was on
> synchronic aspects of this particle and my ICHL paper
> last month was on diachronic aspects of it.

I was thinking of the general positioning of the articles as imperfect
auxiliaries, but, on further reflection, I don't believe they ever have a
plural before them, either.  The future works just like the "imperfective"
in this respect, and there, for example, it's ....=tta=akha (typical third
person) or ...=tta=miNkhe (typical first person).

I've just realized that I'm not positive that the auxiliaries always
condition ablaut, or, to phrase it more carefully, the a-grade.  In fact,
I think they don't, except with the future.  (So, maybe the future and the
imperfective are *not* exactly the same in formation.) Yet another thing I
have to check.  I'm also fairly sure that the last stem in a relative
clause before a definite article doesn't ablaut.

Although the evidential or narrative the (narrative is Dorsey's gloss)
probably does have a separate origin from the article the, I've also
demonstrated that in OP at least it clearly now alternates with khe, dhaN
and even ge in its evidential capacity.  On the other hand, this is a
rather special use of the and may not parallel uses of the animate
articles as imperfective (progressive?) auxiliaries.

> 2. "non-ablauting -e stem verb"  There are no
> non-ablauting -e stems in Dhegiha as far as I know.

Actually, I've just demonstrated that neither e (demonstrative) nor tte
'buffalo' ablaut in Omaha-Ponca.  Akhe doesn't, either, though I'd argue
that this was because it incorporates e.  If it doesn't incorporate e,
it's a third example. However, exceptions are very limited, and fairly
special in nature.  I posted something on this just recently, but I've so
voluminous I'm not surprised it got lost in the storm.

> Oh yeah, there was one other thing.  The fact that e ~
> ai ~ abi all occur in the texts (or even the same text)
> does not mean that they are semantically or
> morphologically distinct forms. They may be individual
> or simply fast-speech variants of one another much like
> some of the plural allomorphs Connie listed for Dakotan
> -- "contractions" if you will. This is why additional
> field investigation is so important. Dorsey may have
> tried to normalize his notation in publications, but he
> didn't always understand everything.

Agreed about Dorsey and understanding and normalization.  However, I
believe there is nothing chance about the alternation of a=i and a=bi and
I don't think e enters into the pattern at all.  There are stems in which
ai seems to lead to e by contraction (ppe < *ppahi 'sharp' is the only one
I can think of). Also, pez^i occurs as an alternate for ppiaz^i.
Finally, the a+i => e contraction also occurs with wa-i, wa-gi, dha-i, and
a-i in the prefixal system of verbs.  These are special cases.

On the other hand, There are tons of places where it ai > e could occur
with =i, and it never, ever does, unless one insists on including just the
akhe and e existential cases for some reason, while ignoring the e=i and
akh(=)e=i cases that show that =i actually follows these.  In fact the
normal development of final ...a=i## is not e, but a, which occurs with
most Omaha speakers - including all those I have ever heard personally.