bikkude

[Koontz John E]

On Tue, 31 May 2005, Rankin, Robert L wrote:
> BUT the actual Dakotan morpheme /-khota/, with its allomorph /khola/
> 'friend' has an organic final /-a/, and so would have /-a/ in all the
> Dhegiha dialects.

I'm not positive the final a in -khota is wholly organic.  Taking into
account only forms like lakhota and lakhol- (a combining form) we'd have
the a straightforward case of "non-organic" -a, unless we're using the
term organic (or the concept organic -a) differently.  The oddity comes in
with khola which seems obviously connected and clearly doesn't follow the
pattern for non-organic -a.

But the whole 'friend' set is very odd:  a collection of resemblant, but
non-corresponding forms with look-alikes in Algonquian and Muskogean.  I
think it's what you'd call a Wanderwort, though students of the other
families claim the look-alikes as native forms, I think.  Either they're
right or they're a bit innocent.  (It's easier to be right if you're
innocent.)  But I'm pretty sure the various resemblant forms are loans in
Siouan, or in at least two out of three cases in Siouan.

The 'friend' forms that I recall off hand are:

- khola (and the lakhota ~ lakhol- pair), etc., in Dakotan

- that Omaha-Ponca form iNdakkudha (which I take to be essentially 'my
friend'), almost certainly a loan from Dakota.  Dhegiha otherwise has
*kHake ~ *i-kHake=...rE  for 'friend'.

- Winnebago hic^akoro (with fossilized *i-hta- alienable possessive
prefixed)

The last is not a loan from Dakota, but doesn't represent a plausible
cognate with it, either.

I think there is a Chiwere form, and maybe some others, too.  These and
the resemblants in other families have been discussed on the list in the
past, so I won't try to track them down now.

Anyway, my suspicion is that -khota in Dakota is from khola, etc., by a
process of back-formation.  The original is khola, which permits a no
longer extant derivative *lakhola, which leads to lakhol- in compounds,
and the last leads by analogy to lakhota.  I'm assuming that initial
da/la/na- are an archaic form of the 'by heat' instrumental, something
I've argued before.  I admit, though, that the step from lakhola to
lakhol- is one that seems irregular.  (See the list archives under
http://www.linguistlist.org.)

> Thus the Omaha word cannot be an actual cognate with 'friend'.  It has
> to come from somewhere else. And it would be in homonymic clash with
> khota (the allomorph of 'friend') in Dakotan.

Omaha does have (iN)dakkudha 'friend' in some songs, but I think this is a
Dakota loan word.