Clarification *wy- and 'think' (Re: 'ice' and 'pot')

[Koontz John E]

On Tue, 24 Jun 2003, Koontz John E wrote:
> ... We don't know a lot about PMV *py (*wy) < PS *wVy-, but there is
> this.  The inflection of *DEM=...ye 'to think' (Da epc^a 'I think', a
> "defective" stem in traditional temrinology - with only this first
> person; OP ebdhe(gaN)) shows that Dakota py for *wy and Dhegiha has
> *bdh.  A few noun sets with *py and *ky show that Dakota tends to lose
> initial p in pc^, so PrePDa *wyegha might well come out pc^hegha ~
> c^hegha.

Between a typo concerning Dakotan py (actually pc^ ~ c^h) for *wy (*py,
*by, all non-contrasting, at most conceptual variants) and lack of detail,
I don't think this was very clear.  What I was trying to say is that,
while Siouan has abundant evidence of *wr (*pr, *br), we have only limited
evidence for *wy.  However, there is a verb, something like *DEM=...ye,
e.g., *e=...ye, 'to think' (of a quotation), that shows that there was
such a cluster.

In Dhegiha this behaves like an *r-stem, with the first person ebdhe- for
the stem edhe-, for example, in OP.  Actually, in OP this verb has to be
followed by egaN 'like that', so the attested forms are ebdh=e'gaN,
edh=e'gaN.  I take egaN here to be acting like an adverbial modifier "sort
of" so that historically the construction is "I sorta think that ..." and
so on.

In Dakotan this stem is defective, i.e., existing only in some of the
potential forms, in particular, only in the first person, which is epc^a.
Since c^h is normally a refex of *y, this is probably from *e-p-yA, i.e.,
it is a *y-stem instead of an *r-stem.

So, although Dhegiha does usually distinguish *y and *r, it doesn't do so
in this case.  In fact, since the third person is not expected *e-z^e <
*e-ye then we'd have to call the third person irregular and analogical
(first person ebdhe < *epye => third person edhe), except that we could
probably also argue that this is a case of intervocalic *-y- treated as
epenthetic and so developing as *-r-.  But as only some intervocalic *-y-
develop in that way, we can see that a certain amount of analogy or
abstraction of some sort has to be involved here, one way or another.
About the best we can do here is to point out that *y in *e=ye is always
intervocalic, because of the proclitic demonstrative, while in other cases
it is either always initial, or at least occasionally initial.

In any event, alerted by the case of 'think', we can discover a few other
cases of *wy (*py, *by), one of which is apparently 'mosquito', which
was something like *pyaphuNka, leading to Dakotan c^haphuN'ka, but OP
na'haNga.  Here Dakotan has c^h rather than pc^.

We are aided in understanding *py by the behavior of *ky, e.g., *kye'praN
'ten' as in Dakotan (wi)kc^e'mna vs. OP gdhe'baN (earlier gdhebdhaN), or
*kye'taN 'hawk' as in Dakotan c^hetaN' vs. OP gdhe'daN.  Here *ky > kc^ ~
c^h in Dakotan, but gdh in OP.

JEK