Historical Explanation for *pi as Plural and Proximate and Nominalizer

[R. Rankin]

> For what it's worth, I'm inclined to suspect that the current
> "object/obviative" articles were the original set, and that the subject or
> proximate animate pair akha/ama were somehow grafted onto this later,
> whereas, particularly with this new approach I've suggested, I'm forced to
> assume that =pi as a proximate marking scheme is older.

Don't forget that Quapaw lacks the -akha and -apa set (they have the/a
second -(a)pa marker) entirely.  There is little doubt that the positionals are
the original set, with -akha, -apa as add-ons.  Logically, of course, Quapaw
could have been the one to lose the two rather than the other 4 gaining
morphemes, but when this sort of loss occurs, it virtually always leaves traces,
and I've found none.

Bob