Assiniboine verbs with verbal complements requiring -pi

[Koontz John E]

On Tue, 30 Jul 2002, Linda Cumberland wrote:
> This is a follow-up to the list I distributed at the conference last
> month, of verbs that require either -pi or -kta on verbal
> complements, e.g.:
>
>     wachipi wachi~ka     'I want to dance'
>     wachikta washka~     'I'm trying to dance'
>     wachipi i~mnushki~   'I enjoy dancing'

I looked at some of these in OP without turning anything new up.

I did notice that OP verbs under imperatives don't repeat the subject,
something I hadn't realized before (or had forgotten):

iNwiN'kkaN    was^kaN'=ga                  JOD 1890:642.3
(you)-me-help try      IMP

a'kkihide           was^kaN' gi'=i= ga     JOD 1890:695.4-5
(you) attend to it try      return IMP

I also noticed that was^kaN often has an instrumental and suffix (not the
causative, as it isn't inflected itself) attached, though I couldn't
discern why:

aNwaN'dhittaN  aNwaNgas^kaNdha=i           JOD 1890:680.12
we work        we try it
< wa-dhittaN   < wa-ga-s^kaN-dhe

I also noticed an interesting case in which an instrumental was repeated,
apparently to clarify or make more transparent irregular morphology:

u's^kaN  i'kkigdhagas^kaNdha=bi= ama       JOD 1890:230.19
deed     he tried it for himself QUOTE

Here the underlying stem is ga-s^kaN-dhe as in the preceding example.
Adding the reflexive kki (here with benefactive sense) to ga- should
produce kki-g-dha-, with transformation of ga- to g-dha-, but kki-g-dha-ga
is what appears.  (There is also a locative i- in the actual verb.)  This
is analogous to the change of ka- to g-la- in Dakotan, in the suus.  In
Omaha-Ponca, ga- changes to gi-g-dha- there.  The additional gi- then
distinguishes this from the case of dha-, which becomes just g-dha-,
comparable to Dakotan ya- becoming g-la-.