e and he.

[Koontz John E]

Actually, this is all on e, and how certain predicative forms in e in
Dakotan aren't the *he attested in Dhegiha and Winnebago and Chiwere.

On Thu, 28 Oct 1999, Robert L. Rankin wrote:
> Is there evidence for the *h anywhere in Dakotan?  The place to look would
> be frozen compounds.  How many morphemes is echa historically?  If it's
> related to ekta, then could -ha be ablauted *he.

This is another aspect of the question whether e in e, etu and ec^ha
used predicatively in Dakotan is the e demonstrative or a special
predicative e.  So far I've observed:

1) That I'm looking at the issue etymologically or historically, and that
I don't have any quarrel with the predicative distinction as a synchronic
fact.

2) That the e demonstrative and its derivative e=di (cf. e=tu, e=l) is
also used predicatively in Omaha-Ponca.

What I haven't done is shown that the predicative e participates in
paradigms of demonstratives like it was the e demonstrative.

The data here is from Boas & Delora 1941:116 ff:

Row 14:  e'l, le'l, he'l, ka'l, tukte'l
Row 17:  e'tu, le'tu, he'tu, ka'tu, tukte'tu
Row 29:  ec^ha', le'c^ha, he'c^ha, ka'kha, to'kha

The absence of a vowel e in the ka forms and to forms (where they exist)
is indicative of the e being separate from the following morpheme l, tu,
kha.  Note that kha remains kha unless preceded by e to condition
affrication to c^ha.  A form like *ketu or *kec^ha would be a dead give
away that the e was organic, or at least 'in addition to' the
demonstrative.

It's clear from the stress anomalies of ec^ha' and from discussion in the
notes that this form is far advanced in grammaticalization outside its
nominal place in this table.  It's also clear from the way el and etu are
not placed adjacent in the table that they are not associated by Boas &
Deloria.  This is consistent with their evidence functional divergence and
with the possibility that the =l forms actually have several sources.  \

There are several forms built on tu, e.g., -haNtu, -haNtuya, c^haNtu (only
with e, so perhaps khaNtu), -khetu, -khetuya.  There are also -haNl,
-c^haNl (also with e only, so perhaps -khaNl), -khel, and -khelya.  The
-haN and khe forms all occur separately, too.  I'm going to guess that in
each case -l represents both -tu and -ka ~ -c^a at least. Forms with -haN
(probably from PMV *thaN) have to do with time, as do the c^haN forms.
The -khe forms seem to refer to manner.  There's also a -tkiya that might
be -t(u)-kiya, but B&D say it's -t(a)-kiya (see below).

I've also pointed out that some forms seem to insert a -k-.  These
include:

Row 27:  -k-ta:  e'kta, he'kta
Row 28, -k-takiya:  e'ktakiya, he'ktakiya
        -k-tataNhaN:  e'ktataNhaN

These compare with the postpositions -ta and -takiya.

The Omah-Ponca correspondences -tta and -ttathaN always have the -k-, or,
to specific, always have the -tt- that presumably reflects -k-t-.

It's possible that the -k- also appears in suffixes like -kha or -khe or
might represent the article -ki(N).  In that case, presumably the final
-ha/-he here might represent the he 'to be' form Bob discusses, but at the
moment I don't have enough understanding of these forms to be certain of
this, and for the moment I'm not sure where to look for *he in Dakotan.

JEK