double inflection

[Koontz John E]

In spite of my sally about Siouan languages not including multiple
plurals, I've thought of a counter example.  Those pseudo-inflected
negatives in Dhegiha (or at least Omaha-Ponca) involve a plural form that
can be pluralized.  The forms are A1 =m=az^i perhaps from =maN=az^i, A2
singular and A3 singular obviative =az^i, plural and A3 singular proximate
=b=az^i < =bi=az^i.  The regular plural can follow this:

dhahu'ni=b=az^i=*bi*=ama 'it did not draw him into its mouth, they say'

hna'tha=b=az^i=*i*=a 'why do you (all) not eat'

wiaN'bahaN=b=az^i=s^te=aN=*i* 'we do not know at all' (we know something
not soever)

aNdaN'ba=b=az^i=xti=aN=*i* 'we have not seen him at all' (we have very not
seen him)

It seems that the cases where =b=az^i occurs with nothing following might
be (third singular) proximates with no following enclitics or cases
followed by additional independent verbs that preempt the plural/proximate
marking.  I'm not sure that accounts for all the exceptions to double
marking.  It may provide a sort of test for enclisis, since declaratives -
which Dorsey always writes as a separate word - don't seem to condition
multiple plurals (but imperatives and interrogatives do).