Osage 'eight'

[Rankin, Robert L]

Re: Wichita /kiyatawha/ 'eight'
 
David wrote:

> The Siouan words cited in this exchange look like they're built on a
combination of this "prefix" and the Siouan word for 'four'. If that
element is borrowed in either direction, it's meaning has shifted between
"five plus" and "two times".  My instincts are that that's a lot of
semantic change for an element in the basic counting system, but the
phonetics is certainly intriguing.  I'm pretty sure Bob R. worked his way through all of this a while
back in a study of the numbers.
 
Well, that doesn't mean I understand it.  :-)   I suspect what happened is that, when the word was borrowed, the /tawha/ was (re)interpreted as Siouan *to:pa 'four' (Lord knows how speakers of OS and KS would have adapted [wh] into their speech).  Then the remaining [kkidha] HAD to be interpreted as having some meaning that would turn 4 into 8.  So semantic change in the sense of some sort of steady progression probably wasn't involved.  Just a gestalt replacement.  I think OS and KS have some words that sound like [kiya] that facilitated this interpretation.  Dorsey interprets the parts as meaning "again four", and he lists KS /kkiya/ as meaning 'separate, apart'.  You can see how, especially if you're using finger counting in sign language, "four separate" or "four apart" might be construed as totaling eight (i.e., perhaps four on each hand).  But the truth is hard to know.  
 
Carolyn wrote:
 
> EIGHT. the only forms I have for 'eight' are hkietoopa (accent on -too) and its less frequent variant. hkidhetoopa  (accent on dhe).  (No hkidha...  forms for 'eight'.)

I think I'm detecting a problem, since kidha 'each' etc. (accent on either syllable) seems to occur with a plain k, rather than hk. Unless these forms are unrelated.

 

For OS 'eight' I recorded /kkidha-/ as the first part in about 1980.  Also Laflesche has "kitha" (with K-dot) from the turn of the century, and it can't be one of his "Poncaisms" because the word doesn't occur in OM or PO.  I suspect the forms in /e/ are recent, although anything is possible.  In that vein, there is a Kaw variant given to me by Walter Kekahbah, /kkialoba/, with a mysterious /l/.  Go figure.
 
Bob