Dhegiha Plurals and Proximates

[rankin]

> In OP, my sense is that =i implies that you are
talking about the
> real world, while =bi means you are talking about the
concept, or
> a hypothesis about the real world.  I can't think of
a really
> good word for this in Latin.

In Osage lots of posttonic voiceless stops and
affricates simply vanish intervocalically in more rapid
speech.  Carolyn would have to be the one to say
whether it ever happens with -api/-ape or not, but it
certainly happens word-internally.  In Kaw I never had
a single case of -i (i.e., loss of the /b/) anywhere in
my data and there is no trace of it in Dorsey's Kaw
data either.  Same for Dorsey's Quapaw where the
particle is -awi/-awe regularly.  Even the
phonologically "weaker" /w/ stays.

So I wonder if we simply have a more rapid delivery
when we "know what we're talking about", but when we
voice hypotheses and the like, we simply slow down and
the phonology doesn't routinely suffer so much
fast-speech reduction?

As for the initial /a/ of -api, -azhi, etc., of course
it could historically be a separate morpheme.  Actually
I think John has had a theory that this was so since
the mid '80's that he can elaborate on.  But there is,
of course, always a temptation to keep reducing things
until, basically, every phoneme is a morpheme.  It's a
question of evidence.  It is the case throughout
Dhegiha that the positionals take the "a" form
of -tte/-tta, but whether the explanation is analogical
or morphological remains unclear, at least to me.
Until I find more contexts that would help put a label
on it, I personally take the explanation to be
analogical.

Bob