Companion Terms for 7 and 8 (Re: 'eight' some more)

[Rankin, Robert L]

[RLR]:  Actually, this is the best idea I've seen lately.  There's your
Siouan Conference paper for you.  I've added my own Proto-Dhegiha
reconstructions (as I understand PDh to be), but the upshot is exactly
the same, of course.  The source and meaning of PDh *hpee- has been
bothering me for years.  This makes good sense.

[JEK]:
> I was wondering about that, but bothered by *hpe vs. *-pe.  Now it
occurs to me that one possible explanation of *hpe is *kpe, perhaps from
*s^aakpe, cf. Mandan ku'pa 'seven' ?< PreMA *saakupa, cf. PS *s^aakowiN.
In this case one would have to assume that the pattern in Dhegiha was
modified along arithmetical lines:

    PMV              PDh			PDh II (rlr)

 1  *wiN'            *wiN			*wiN'-xti
 2  *nuNuN'pa        *ruNuN'pa		*nooN'pa
 3  *raa'priN        *raa'priN		*raa'briN
 4  *too'pa          *too'pa			*too'pa
 5  *saa'ptaN        *saa'ptaN		*saa'htaN

 6  *s^aa'k=pe       *s^aa'kpe(wiN)		*$aa'hpe-(wiN)
 7  *s^aa'k=owiN          *kpe'ruNuNpa	*    hpee'noNpa
 8  *s^aa'k=???           *kpe'raapriN	*    hpee'raabriN
 9  ???              *s^aNaN'kka		*$aaN'hka
10  *kyee'rpaN       *kyee'praN		*kree'braN


[RLR]:  With Proto-Chiwere having something closer to the PMV forms, the
above paradigms would also explain why Chiwere has a /k/ in the initial
syllable of /kree-raabriN/ 'eight', although it doesn't explain the /r/.
I would have expected maybe **/kwe-/, instead of /kree-/.  Vowel length
in Dh *hpee- and Ch *kree- is also a loose end.  This is a really
interesting working hypothesis though.

Bob