ska, s^ka, xja, <th>ka, etc.

[Koontz John E]

On Fri, 16 Feb 2001, Lance Foster wrote:
> > I think hka is a variant of (older) ska ~ (newer) <th>ka.
>
> I think there was a change from shka to ska to a kind of lisping ska/thka, then
> finally thka. Whitman describes this. Don't know why.. that's why I'm here.. to
> learn :-)

The comparativist's appreciation of the situation is that most of the
Mississippi Valley Siouan languages have s/s^/x contrasts (and at least
some voiced z/z^/gh contrasts) in the fricatives.  These occur in older
examples of Ioway-Otoe more or less matched with the other languages in
cognate sets, like *ska 'white'.  However, for a long time now Ioway-Otoe
has been shifting s to <th> (i.e., theta) and s^ to s.  My recollection is
that x stays put.  Z/z^/gh troop/do not troop along in unison with
s/s^/x.  In addition there is some tendency for sk to become hk, which, of
course, conflicts with the tendency of s to become <th>, leading to <th>k
~ hk doublets.  And, in older sources you find sk still. The shift of s^
(or sh) to s is less absolute, if I recall, so that s^ variants tend to
still occur.

Naturally, this is all somewhat complicated by the existence in Ioway-Otoe
of the same pattern of sound symbolism with fricatives that occurs in
other Siouan languages, so that one might expect for ska sound symbolic
grades of s^ka and *xka.  I think xk tends to be rare to non-existent
because of constraints on clusters.

Some Stoney dialects also shift s to <th>, and Ofo in the Southeast shifts
it to f.  In Mandan, to everyone's amazement, s and s^ are interchanged.
Hu Matthews has a paper in IJAL c. 1970 addressing the Mandan situation
and some others.  Siouanists pondering these problems tend to spend a lot
of time squinting askew, frowning, and counting on their fingers.

JEK