Omaha fricative set

[Koontz John E]

On Thu, 28 Sep 2006, Rory M Larson wrote:
> Is the Kaw reflex of common Dhegiha preaspirates also hp/ht/hk, as in
> Osage?

I think preaspiration as opposed to length and tenseness is usually
associated only with Osage in terms of modern attested phonetics.
However, the forms that can't be tied to stop-stop clusters in Mississippi
Valley are generally reconstructed as *hC preaspirates in
Proto-Mississippi Valley and thus in Proto-Dhegiha.  These are the forms
in the Dakotan Ch :  Dhegiha CC : IO C(h) : Winnebago C / __V sets, like
Da tha : OP tta : IO t(h)a : Wi taa vs. (a cluster set) Da pte : OP tte :
IO c^(h)e : Wi c^ee

> John mentioned before that in stop clusters, the second element usually
> won, except when the first element was t, which usually took over in any
> case.  Is this a more sophisticated explanation of that observation?

Essentially yes.  My explanation along those lines was more in the line of
classifying results.  Bob's approach attempts to explain how tk and kt
both become tt in phonetic terms.  The inspiration and justification for
this analysis is the treatment of *tk in Ioway-Otoe vs. Winnebago, I
believe.  In 'bow', for example, I seem to recall that IO has maN(aN)hdu
where Wi has maNaNc^gu, presumably both from something like *maNaNtku.
Something similar happens - independently - to *tk in Stoney.

> I.e., MVS *tk regularly -> Early Dh. *kt -> Later Dh. *ht ?

That's the idea.

> Does MVS *tp do the same thing?  Can you offer any example words?

Not off the top of my head, but I think are some.

> >  So these are all borrowings and diffused words.  Unfortunately this
> > reduces the usefullness of those three sets of apparent z/s
> > correspondences I posted.  'Five' is still a mystery though.

I tend to suspect that 'five' might not also involve borrowings and
diffusion, even within Siouan.

I remember helping someone a while back with terms for some aquatic and/or
tuberous plants and discovering that they were rife with this same kind of
irregularity.   I think the forms mostly involved (p)Se/i(N).

> What about the 'squirrel' term?  Would that still be good?

I thought 'squirrel' terms were squirrely by definition.

> But if borrowing/diffusion can explain the 'chicken' term, I would think
> that a standard number term would be even easier, as a tool for trade
> between groups.  The *s/z/aptaN term isn't even common to Siouan outside of
> MVS, is it?

I tend to agree.  The relative stability of numerals in Indo-European led
to an early impression that they were basic, but I think consulting
additional data tends to suggest that they are "cultural."

> Maybe these problematic s/z terms ...

Hu Matthews must be chuckling sympathetically right about now.  I think
the reference is Matthews 1970 in IJAL.

> all date to a period shortly after the spread of MVS, after significant
> differences in pronunciation had developed between dialects, but while
> they were still pretty well mutually comprehensible, and while the
> speakers still recognized a common ethnicity.

It's probably worth pointing out that only Mississippi Valley
distinguishes voicing or sharp/muted oppositions in fricatives at all,
which is a large part of what Matthews was wrestling with.

> Do we have any others?  (Trade terms?  Hunting small animals?  Spread of
> bow and arrow?)

The 'bow' terms are perhaps trade terms, since they seem to be of
Algonquian origin.   Terms for horticultural items are also sometimes
rather problematic, including particularly 'tobacco'.  We got into 'cat'
briefly a while back (for the nth time).

One of the things you gradually recognize in dealing with Proto-Siouan is
that a number of common sets contain small irregularities that we have
gotten used to ignoring over the years.  I mean irregularities that can't
be accounted for in terms of morphological context, obvious analogies,
etc.