Stem Truncation in Dhegiha (was Re: More bears)

[Koontz John E]

On Wed, 28 Feb 2001, Erik D. Gooding wrote:
> What about lions and tigers? (that's me playing the straight man for
> someone else)

Well, Omaha-Ponca iNgdhaNsiNsnede 'mountain lion' is interesting in
providing two examples in one word of stem truncation, in which a longer
stem is truncated in a compound.  I've already mentioned this in passing,
including truncation in kin terms, where it appears to be diminutive in
origin.

This word is iNgdhaN(ge) 'cat' + siN(de) 'tail' + snede 'long'.

Comparable examples are s^aNttaNga 'wolf' with s^aN(ge) 'dog' ('horse in
historical usage) and waz^iNttu 'bluebird' with waz^iN(ga) '(small) bird'.

One way to analyze this is as a Dhegiha implementation of the C-final stem
pattern that some stems (C-final ones) show in Dakotan, e.g., s^aN <
*s^aNk < s^aNge, rather than direct truncation.  Dhegiha languages tend to
avoid C-final forms in independent words and have a greatly reduced
cluster inventory, so neither *siNt or s^aNk nor *siNtsnede or
*s^aNkttaNga would be expected.

Note that Omaha, however, does have a tendency to reduce word final CV
sequences to Ch[V-voiceless] which comes across as Ch#.  This is
especially frequent with the articles akha and khe.  But it also occurs
with the article ama and the homonophonous quotative.  Medially you also
get niks^iNga for nikkas^iNga 'person', ttapska for ttappuska 'student;
school', and so on.

I'm not positive what the conditioning is, but it looks like something
along the lines of C__# (in enclitics) and C__C[voiceless] (medially).

Incidentally, as long as we're looking at obscure borrowings, that
'student, school' word is also attested in Pawnee.  The origin is unknown.