Locative Postpositions

[Koontz John E]

On Wed, 27 Oct 1999, regina pustet wrote:
> e' can also combine with cha to serve as a topic marker.

This is one of the interesting things that comes out in Ingham's article,
namely that often forms like ec^ha, which have a fairly obvious
morphological analysis, have a function which goes beyond what the
morphological analysis suggests.  The comparable form in Omaha-Ponca is
egaN which has a number of grammatical functions (but not topic marker)
that arise historical out of 'be such a one; like that', but clearly have
a life of their own.  Etu may be another such example in Dakotan.

> What these three function have in common is the semantic
> denominator of "pointing at", in some sense, and from that, I'd say, it's
> just a short way semantically to a demonstrative function. In other words,
> John's "neutral demonstrative" e' and David's "verbal" e' might be one and
> the same, at least if we go back far enough in language history.

Tada!

> I'd also
> like to make a connection between e' (whatever type) and the locative
> prefix e'- 'at' towards', which occurs, for instance, in e'-gnaka 'to put
> (somewhere)' (vs. gna'ka 'to put'). e'- is not listed in the Boas/Delora
> grammar as a locative prefix that occupies the same slot as the more
> common locative prefixes a-, o-, and i-, if I remember that correctly,
> although it certainly does. Though e'- is not particularly productive in
> Lakota any more.

There are similar things in Omaha-Ponca, though what comes to mind is
eppaze 'to spend the night at a place', which may not be strictly
analogous.  There are some other examples, but they're not coming to mind
at the moment.

Are these forms really analogous to locatives in Dakotan in the sense of
preceding first and second person, but following inclusive?  I think
they're more like outer instrumentals or other incorporations in
Omaha-Ponca, i.e., precede everything.

> > The -l/-n
> > marker is, in my opinion, the reduced form of e'tu, not the -ta
> > postposition,
>
> I'd also vote for -tu rather than -ta. ...

> The pitfall is that stem-final -ta is also regularly reduced to -l in
> Lakota.

I'd been arguing that tu = l for a while when I ran into Ingham's article.
The data there left me with the distinct impression that there were also
reduced el forms that matched -ka ~ c^a (the c^a alternative is normal
with pronominals, as they normally end in e).  I can't remember now if
there were -ta cases, too, but presumably not with e-, since that
introduces a transitional -k- to form ekta, an interesting irregularity
that I can't account for at all.  I belive there are some other e +
postposition cases where -k- is introduced, but they're not coming to me
at the moment.

> > 	I am not aware that modern day speakers use the "thil" form for
> > the locative of 'house';
>
> Never heard that either. thil might be just a rapid speech contraction of
> thi + el.

tti=adi (and tti=atta) and tti=the=di (with the 'the upright inanimate')
are normal in Omaha-Ponca.  There is also an anlogy to maNhe with
(cognate) maNthe, though I think tti=maNthe may be restricted to
earthlodges.  maNthe is normally 'under(neath)'.