Historical Explanation for *pi as Plural and Proximate and Nominalizer

[Koontz John E]

On Mon, 9 Feb 2004, REGINA PUSTET wrote:
> It would be interesting to check on the etymology of opha 'to join' as
> well. Maybe all three forms, i.e. -pi, ob, and opha, are, ultimately,
> related. The semantic fit of ob 'with' and opha 'to join' is just to
> close to be neglected, at least to me.

Now that she points it out, I agree with Regina that ob seems like it
might derive from o'phA, which in Buechel is glossed 'to go with, follow;
be present at, take part in'.  This seems to derive from a more complex
gloss in Riggs 'to go with, to follow; to pursue, as opa aya [i.e., ophA +
ayA 'to go' form 'to pursue'];  to go to , attend, as a school or meeting,
to be present at; to be a member of, as an association or church; to go
in, as in a canoe ...'.  We've just verified that Dakota verbs in final
ChV (aspirated stop) can reduce to -C (final unaspirated stop) in
subordinated forms.

In the CSD the Crow-Hidatsa forms and ob are only mentioned in the over
all notes for this slip.  I don't know anything about the history of this
slip.  I think the grammatical slips rather languished after the first
year.  It looks like this one could use some modifications in light of a
better current appreciation of why Dhegiha plurals sometimes end in -e or
-a instead of -i (gender-coding declaratives e or a?), and the
morphophonemics of the Omaha-Ponca plural are eliminated.  It seems to
antedate the "proximate" terminology for Dhegiha third person singulars
with "plural" marking.

Actually, I am pretty amazed to discover that there are Crow-Hidatsa forms
that might provide an explanation of the origin of +(a)pi that is in line
with the 'pre-eminent among others' sort of focus that I was appealing to,
and I appreciate Bob making the connection and pointing it out.

What the slip says is that Hidatsa has aapi 'with' and Crow has a'appaa in
which the latter part -paa derives from the common adverbializer -haa,
which looks in turn like a cognate of the =ha postposition in Omaha-Ponca
that is glossed 'in places, in directions'.  For example, du'(u)ba=ha 'in
four places', dhabndhiN=haN 'in three parties', gu'=di=ha 'further off
(yonder direction)', etc.  This is particularly common with numerals and
in songs.

I suppose this analysis requires that the Proto-Crow-Hidatsa form is
*a'api, and Crow has =haa added to reduced *a'ap?

I'm vague on what happens to aspiration in Crow-Hidatsa.  I think it more
or less disappears, i.e., that *hp and p, for example, behave the same.
So, PS *hpa(re) 'bitter' with Hi (ara)pari 'bile', OP ppa 'bitter', and
also PS *paN(he) 'call', with Hi paa (imp. sg. pah) 'shout', OP baN.
'call'.

Now, if the PCH forms is *a'api and not *a'aphaa, it seems unlikely that
the root here is the root in Da o'phA, which looks more like the 'go,
travel' root *phE that appears as -hE in Omaha-Ponca, e.g., (udh)uhE 'to
follow (a trail)', Osage (odh)o[ps^]e, IO (ir)owe(=are).  (This is the
comparison made in Dorsey 1895, if I recall, and I think it is correct.)
In fact, the two forms look like cognates, though the CSD seems to cross
up the OP form with some other Dhegiha -hV forms probably of a different
origin and compares Da okhihe 'follow' instead.

JEK