Stars.

[Koontz John E]

This doesn't add much to Bob's account, but may be easier for some folks
to follow.

On Mon, 19 Feb 2001, RLR wrote:
> > In Lakota the word for 'star' is wichah^pi .
> > Normally the prefix wicha- refers to animate plurals.  Does anyone
> > know why it turns up in the word for star?
>
> Hard to say. I'd guess either by accident or, perhaps, by
> folk-etymology. The form in several related languages is nasalized:

Mississippi Valley 'star':

  Dakotan      wic^haxpi   < *Wi  hkax pi
> Omaha-Ponca  mikka'?e    < *miN hkax ?e
> Kansa        mikka'k?e   < *miN hkax ?e
> Osage        mihka'k?e   < *miN hkax ?e
> Quapaw       mikka'x?e   < *miN hkax ?e
  IO           bikha'x?e   < *Wi  hkax ?e
  Wi           (wiira guNs^ge)

The affrications of k to c^ after i is regular in Dakotan.  The k ~ c^ set
is otherwise typical of the proto-Mississippi Valley "preaspirate"
correspondence vs. the "(post)aspirate" and "unaspirated"
correspondences.

The spaces here are just to line things up.  They don't necessarily
correspond to morphological boundaries, but might.  In particular, note
that the first syllable in each case actually is the term in each language
for 'luminary', that is 'sun; moon'.  These forms have the same irregular
corespondence, with *W in Dakota, IO, and Winnebago, and m + nasalized
vowel in Dhegiha.

In the Dhegiha languages, terms for 'moon' add aNba (or uNba?) and in OP
the compound is changed then to niNaNba.

The Winnebago term is based on wiira 'the luminary' = wii 'sun; moon' + ra
'the'.  GuNuNs^ge is 'skunk', though guNuNs is 'teach, create'.  The -ke
(-ge in some Wi orthogrpahies)  is the Winnebago and Ioway reflex of the
nominalizer/"qualifier" -ka in Dakotan and Dhegiha.  I'm not sure if
'star' should be rendered 'skunk luminary' or 'creator/teacher luminary'.
The latter seems more likely.

That sense might make some sense of *hkax in the other languages, since in
all Mississippi Valley languages but IO and WI developments of *kaghe (Da
kagha, OP gaghe, etc.) are 'to make'. The preaspiration in hkax is,
however, something that only occurs when some other morpheme is prefixed
to *kaghe e.g., *ki in some of the languages. cf. Da kic^haghe 'to make
for' or OP gikkaghe 'to make one's own'.  Since nothing is clearly
prefixed here, we're left suspecting that we're going down a false trail.

> The suffix is different in the two subgroups also. It appears as though
> -?e is a suffix in Dhegiha and the same -pi suffix we find so well
> represented in Dakotan nouns turns up there. So the glottalized
> fricative found originally in Dhegiha is historically most likely a
> composite with a morpheme boundary in the middle. Then the result
> underwent the usual Dhegiha change: *x? > k? > ? with Kansa/Osage
> retaining the middle form and Omaha-Ponca showing the last stage. There
> are plenty of other examples of this change.

For example, 'woman' OP wa?u, Os wak?o, Qu wax?o

The ...x-pi sequence here is reminiscent of the one in the 'cloud' forms
we have been discussing.  We could possibly take it as a plural (as
opposed to the "singular" in -e in 'sky'), but since the singular 'sky'
form lacks the x?e reflexes, there are problems with that.

About the best we can do is line the evidence up, notice the problems, and
wait for some further enlightenment.

Incidentally the *W set is Dakotan w (sometimes b in Santee or p in
Teton), OP m, Ks b, Os p, Qu p, IO p, Wi w.

JEK