Status of PS Glottal Stop (was Re: Locative Postpositions)

[Koontz John E]

On Sun, 31 Oct 1999, Robert L. Rankin wrote:
> These recurring, non-predictable glottal stops are found in a handful of
> words including ?u:N 'do, be', ?iN 'wear about the shoulders', ?e:
> 'general demonstrative', ya?iN 'think', ?o: 'shoot at and hit, wound',
> ?iN- 'stone, rock' and a few others.  Some are nouns, some verbs.  Most
> seem to be word-initial, which is suspicious of course.  There are a few
> others I think.  Very messy, but there they are.

The main environments in which *? might be audible, are, of course,
initial position (not really audible in Omaha-Ponca) or intervocalically
(also not really audible, at least in eaN 'how', which I have heard) , if
k or other morpheme final stops merge with ? to produce ?.  The ? from *k?
is very audible on Omaha-Ponca, as in a?i 'I gave' < *ak?u.

Unfortunately, I haven't heard Omaha-Ponca aNaN 'we do', so I don't know
if it's aNaN (with, say, rearticulation) or aN?aN.  Any of the other
Dhegiha languages would, perhaps, clarify this.  I seem to recall that
none have aNk?aN, but can be positive.

In fact, I think that only Dakotan and maybe Winnebago have C? with any
*?-stop stem forms.  Dakotan has k? with the inclusive and Winnebago has
s^? with the second person, if I remember correctly.  These two forms are
prima facie evidence that ?-stop was perceived as part of the stem at some
point in the history of PS.  On the other hand, all these C? sequences are
well preserved in Mississippi Valley Siouan, but are rare in the
inflectional pattern even there.  Hence, I conclude that there is
something odd about the status of ? in these stems.  When you realize that
only Dakotan and Winnebago make much of a big deal about initial ? at all,
wonder if the C?-forms don't say something about these particular
languages, rather than Siouan at large.

Winnebago is actually far more relevant here than Dakotan, because it
alone substitutes hV for all those *V initials that regularly become ?V in
Dakotan.  Thus, Winnebago alone seems to have an unpredictable ? in these
stems (even if it seems to be secondary) without using ? with all
V-initial stems.  But why, s^? if ? is not original?  One possibility that
occurs to me that doesn't involve *? is that there was an additional
prefix in the third person, e.g., *i, leading to *i?uN > *?uNuN or
something like that in the third person.  Or maybe *uNuN was just long.
Long vowel initials don't get epenthetic h.  In this case, s^?V in the
second person is by evidence with the recent third person only.

Back when I was looking at *?-stems, I also noticed that a number of the
languages seemed to have substituted *r-stem personal inflections for some
or all *?-stems stems, especially in the second person.  Dakotan has nV,
for example, which doesn't correspond regularly with either Dhegiha *z^V
or Winnebago s^?V.  I believe this is because the forms in question
sometimes have an epenthetic *r in the third person, due to prefixes,
making them appear to be *r-stems.

Since I could account for the *nV second persons and the Winnebago form
looks secondary, too, I have to conclude that Dhegiha *z^V is the original
form.  Dhegiha does have *nV with some stems, incidentally, though curwe
usually call these *w-stems, because of the epenthetic w that develops in
the third persons, which have either a prefix *o or an initial *u, etc.
Since all the inflectional forms with *C? are restricted to one language
or another, and the various other inflectional forms, *mV first person,
*z^V second person (perhaps *nV second persons, if they don't come from
the *r-stems), don't suggest *? at all, I suggested that these stems might
actually be *V-initial.  The only real counter-evidence is the Winnebago
second person in s^?V, secondary though it presumably is.

This is a summary, without the paradigmatic evidence tables, of my
position on *?-stems.  Notice that it's somewhat wobbly.  I think it looks
like the *?-stems were actually *V-stems that frequently acquired regular
epenthetic ?-initials in the third person, which sometimes resulted in
secondary ?-introduction in other persons (second or inclusive), but I
wouldn't want it to be forgotten that *? is also a possibility, and that
Winnebago provides the best evidence of this, albeit not incontrovertable
evidence.

JEK