Lakota chaNnuNpa

[Rankin, Robert L]

> MVS      *RaNri'                *RaNri'-uNpa

> Or perhaps the dubious phoneme sets are due to 'tobacco' being borrowed after Dakotan and Dhegihan diverged.

Probably *iraNri with initial i- accounting for the Dak. ch-, as in several kinterms.  But you're right, this word was widely borrowed and may or may not be reconstructible even at the subgroup level.  There are Iroquoian look-alikes and I doubt there is a good Proto-Siouan reconstruction.

Bob