Dhegiha Negatives (RE: ognayehci- vs uNgnayehci)

[Koontz John E]

On Thu, 8 Sep 2005, Jan F. Ullrich wrote:
> The two words you gave are derivates of: ogna' 'in, in the way of' and
> uNgna' 'perhaps, by chance, suddenly'. So the question is whether the two
> are different words or not? In my experience based both on work with
> speakers and texts they are.

(Thanks to everyone for pointing out the porential for #o- ~ #uN-
confusion in Lakota.  I'd overlooked that.)

I think that historically the -gna element is what amounts to a vertitive
(in k-) of the 'sitting'/'round' positional.

The comparable OP form would be gdhaN, though I think OP ugdhaN (<
*o-k-raN(kE)) is 'to put (a round thing) in, to insert (a round thing)'.
That would be a linguist's analytical gloss, not a colloquial one.  I'd
have to check that the form occurs, but I seem to remember it.

The uN- element in Lakota uNgna I don't recognize.  I'm assuming that the
underlying stem is again -gna, which it might not be.  Potentially the
form could be something like (?) *uNk-la I suppose.

The only parallel for uN- or uNk- that occurs to me is OP aNkkaz^i 'no'.
This is actually a sort of particle aNkkV to which has been appended the
somewhat inflected OP negative.

A1  aNkka=m=az^i 'me not'
A2  aNkka=z^i 'you not'
A3/Plural  aNkka=b=az^i 'she/he/it/they/us/y'all not'

The negative morpheme is =(a)z^i, which conditions the a-grade, as does
the plural/proximate =b(i).  The only actual inflected form is maN 'I do'
in the first person.  The rest is just negative base aNkkV,
plural-proximate =b(i) and =(a)z^i negative in various combinations.

I don't control the idiomatic use of the personal forms here, but I think
you could gloss them 'I say no(t)', etc.  'Yes' could be hau (ho)
'expression of agreement, approval' if male is speaking to male, but I
think verification and agreement are more usually expressed as egaN 'like
that', comparable to Latin sic, the source of Spanish si, Portuguese sim,
etc., in Romance.

The form egaN is itself inflectable, e.g.,

A1 egimaN
A2 egiz^aN
A3 egaN

I am not sure of plural(s) at the moment!  I don't think the personal
forms are used in agreement (i.e., saying 'yes'), but I don't know.

Anyway, returning to aNkkV, its internal structure is obscure to me, but I
suppose it might be *uNk-ki-, parallel with e=g(i)-aN < *e-ki-?uN.  In
that case the aN < *?uN reflected in A1 aNkka=m=az^i might be a parallel
with the more complete paradigm of aN < *?uN in egimaN < A1 *e-ki-m-?uN,
etc.

If that's so then aNk- in aNkkaz^i 'no' might well be historically from
*uNk- 'doubtful', and be used in parallel with *e- 'thus, as stated' in
egaN 'yes, thus'.