"WOUND"

[Koontz John E]

On Wed, 3 May 2006, Rankin, Robert L wrote:
> Our question about this verb had to do with the conservative pronominal
> prefixes that commonly occur with stems beginning with */?/.  ...
> Languages that have reflexes of regular *wa+?uN, *ra+?uN are not
> germaine to the discussion.

Well, not germane to that aspect of the discussion, but I think Jimm's
primary concern is actually just what the inflection of the verb o(o)
'wound' was in IO, in the absence of attestation, and in that case the
patterns in IO for uN(uN) are relevant, as are, perhaps, the ones in
Winnebago. I think Jimm knows that this verb is one of the awkward ones,
though, and he's looking for anything to shed light on the problem.

I think it's likely that the verb was A1 *ha?oo, A2 *ra?oo, A3 oo, maybe
A12 *hiN?oo, and that at some point before that (maybe most recently in
proto-Chiwere-Winnebago) it must have been something like A1 *ha?oo, A2
s^?oo or *s^oo, A3 *oo, A12 *hiN?oo, to judge from the attested Winnebago
forms.  For that matter, I could imagine more complex possibilities like a
second person *ras^?oo.  The actual forms must have varied across time,
and, given the number of different groups speaking IO c. 1800, they may
well have been different in space, too.  But for Jimm I think the question
is less one of the reconstructive possibilities at given points, than a
philological problem of what he is justified in putting in his dictionary
and and an applied one of what he could suggest that would be speakers
use.

Philologically he can't put anything, I think, except A1 unknown, etc.
But in an applied sense he might want to recommend something by analogy
with uN or the regular paradigm, and I think the two are close enough to
suggest a working answer.  However the applied use reader would want to
know that in this case the form is a guess, and, of course, a historical
linguist would very much want to know what was attested and what was
guesswork.

> As John says, *?o: 'wound' is the only known case of a ?-stem with an
> oral vowel.  So, what was the conservative (syncopating) pronoun
> allomorph used before analogy yielded regularized forms?  The bottom
> line is that apparently we still don't know.

See my comment on Randy's remarks.

> An educated GUESS might be that the pronominals should be oral versions
> of */m-/ and */y- ~ r/ in the 1st and 2nd persons, but what would those
> forms be?  ....

> Would Dakotan have been {A1} *[bo], {A2} *[do], {A1}[bo], {A2} [no],
> {A1} [mo], {A2} [no] or what?

I'd guess PS A1 *woo, A2 *yoo, A3 *[?]oo, but more like PMV A1 *boo (maybe
*woo), A2 *z^oo, A3 *oo.

I'm pretty sure that the n in the second person of Dakotan glottal stems,
e.g., nuN 'you do' is contamination from the y-stems (*r-stems), and so I
plump for the Dhegiha second persons in z^-, e.g., OP z^aN as the model,
over Winnebago s^?uNuN or Dakota nuN.  The Dakota pattern is even seen in
some verbs in Dhegiha, cf. OP 'the (sitting)':  A1 miNkhe, A2 (s^)niNkhe ~
(s^)niNkhe=s^e, A3 dhiNkhe;  or OP 'to interrogate':  A1 imaNghe, A2
i(s^)naNghe, A3 iwaNghe ~ idhaNghe.  (I think these are all OP forms, but
other Dhegiha adds further alternates, including first persons on the
order of *ibdhaNghe.)   I suspect these verbs are just irregular
glottal stop stems, unusual in that they have second and third persons on
the *r-stem model, but showing their true colors in the first person.

In the second case the -w- and -dh- after the locative i- are potentially
epenthetic.  It's not so clear what's happening in the first case.
However, extra or epenthetic -dh- < PMV *r in these forms tends to explain
how *r-stems manage to contaminate the *?-stems:  the epenthetic *r in the
third person is simply taken as organic and new second and/or first
persons are produced, modelled on the *r-stems.  If you consider that
stems like *?uN 'to do' seem to have had derivatives like *i...uN 'to do
with;  to use', it begins to look like most *?-stems might have had a
potential for *r-stem contamination.

In addition, even if we consider *...(i)iNk(e)=...he and *i...uNx(e) not
to be glottal stop stems, even if we decide they are a special kind of
stem traceable to PMV, they do provide a basis for the analogical change
of *?-stem second persons from *z^- or *s^- to *n-.  We do have to assume
that this n- is from earlier *s^n-, but it's pretty clear that l- in the
second person of Dakotan y-stems (*r-stems) is from earlier *s^l-, too, so
this shouldn't be a problem.