Epenthetic Initials and Problematic Initial h-

[Koontz John E]

On Fri, 30 Jan 2004, Koontz John E wrote:
> Yes.  Winnebago has an epenthetic h on vowel-initial words, with certain
> exceptions, ...

One interesting postscript to epenthetic h is that the h with the first
persons A1 ha, P1 hiN, A12 hiN, which behaves in Winnebago like one of the
epenthetic h's, is also found in IO, which doesn't have epenthetic h
otherwise.  So these pronouns have h- in IO, but the locatives, etc.,
don't.  In Dhegiha there is no h on these pronouns either, cf. OP A1 a, P1
aN (P1+dative = iN), A12 aN (A12+dative = iN).  It's hard to tell if the h
here is a reflex of *w in IO (and Wi?) or some sort of incipient
epenthesis, though the former seems more plausible to me at the moment.

The other more or less problematic h- in Siouan is the one that appears in
'day' in Wi, IO, and Dhegiha-less-OP, but is missing in OP and Dakotan.

If I recall, this set is Wi haNaNp, IO haNaNwe (least sure of this one!),
Os haN'pa, OP aN'ba, Da aNpA.  The Da A ablauts, and is e before, e.g.,
=tu, as in aNpe'=tu.

There's a similar h- in the indefinite/interrogative pronoun base *(h)a-
where it occurs, e.g., OP anaN 'how many' vs. Os hanaN.  I think there are
traces of this ha- in Wi as well.

The 'day' set is Chafe's example of PMS *rh (or was it *hr?), from the
naN- prefix on 'day' in Southeastern and no doubt Iroquoian data I have
forgotten.  I'm not sure if that explains the h ~ nil alternation or not.

JEK