Hawk.

[Koontz John E]

I just dashed this off over the weekend without thinking much about it, to
provide the forms supporting Bob's comment.   The material is from the
CSD draft, of course, but with one change and some omissions of
Southeastern forms that might not fit as well.

On Sun, 28 Nov 2004, Koontz John E wrote:
> PS *kyetaNhe ~ *kyetuNhe 'hawk'
>
> Ma titaNhe 'hawk'  (Ma ti- : PS *kye- not otherwise attested)
...  (the uN vowels in IO and Wi are unexpected)
> Bi *kyetoNhi 'duck hawk' (maybe oN is unexpected)

I added the *-h(e) on my own authority, based on the Mandan and Biloxi
forms.  I suspect it's just an h-final stem.  That is, some CVC-stems are
CVh-stems.  The editors have been coming to grips with these in the course
of the editing.  This element of Siouan morphology was a discovery of CSD
editors I believe.  I'm not sure who recognized what, except in one
detail.  The *-h(e) is usually lost pretty much everywhere in such stems,
but traces can remain

in Crow-Hidatsa, e.g., Crow final diphthongs -ia and -ua where
intervocalic h is lost (no CH forms attested for this stem);

in Mandan, e.g., -h resurfacing with some noun stems before what I might
call the absolute marker -e;

in Mississippi Valley if -ka follows, yielding, e.g., Da -kha, OP -kka, IO
-khe, Wi -ke, from *-h-ka.

The behavior of *-h-ka was, I think, first noted by Bob Rankin.  In this
case it appears that Biloxi follows the Mandan pattern.  I'm not sure if
this has been noticed in a Biloxi form before, and so, perhaps, I have
misanalyzed the behavior here.  I love the *-h-ka forms because they help
confirm that *-a > -e / [velar]__## occurs in Winnebago as well as
Ioway-Otoe, even though most final -e are later lost there.

Biloxi does have final -i in various contexts that behaves like the -e
increments to stems elsewhere.  Particularly widespread is -di < *-r-e
where presumably the *r is epenthetic.  This is something noted by Dick
Carter early on.  I don't recall whether he felt *r was epenthetic or
organic.  I think everyone who has pondered the matter has argued either
way at various times.  At the moment I favor epenthetic myself.  Carter
compared Biloxi -di to the underlying -r that resurfaces before -e in some
Mandan nouns.  In Mandan some superficially V-final nouns add -re to form
the absolute, just as some add -he (per the above) or -?e or -r?e.  A
similar -r appears with some verb stems when the declarative -o?s^ is
added, etc., and -di appears with both nouns and verbs in Biloxi as far as
I can recall at the moment without looking.

The following was the other 'hawk' form that Bob and I remembered, he
apparently in more detail than I.

> PS *kyaNs^ka' 'hawk'
>
> Da c^haNs^ka' 'large hawk'
>
> OP gdhaNs^ka'
> Qu xnaNs^ka'  (kn, not xn, expected)
>
> Bi *kiyaNska' 'marsh hawk'  (s^ka, not ska, expected)

This is unusual in that the initial *kyV- has a kiyV reflex in Biloxi.
This obviously recalls the verb 'to fly'.  This set has a number of
irregularities in it.  The final stress seems a bit odd to me, too.
Final -ka in this stem doesn't behave at all like the -ka formant rather
common in terms for animals, etc., in various Siouan languages.

The following is just an example of *ky in another context.

> PMV *rukyaN' 'think'
>
> Da yukc^aN 'to comprehend'
>
> OP i'dhigdhaN 'to decide'
>
> IO iirugraN 'to think'
> WI rukaNraN' 'to manage'