New topic -- stative pronouns with reflexives?

[R. Rankin]

It wouldn't be typologically unusual if Lakota were an ergative language where
transitivity had a lot to do with pronominal selection, but it seems a little
peculiar to me to find it in Siouan.  Here we'd expect the patient set with
stative and (sometimes) experiencer subjects, not just intransitive ones.

Biloxi doesn't distinguish active/stative pronominal sets, and it looks as
though all the Mississippi Valley Siouan languages except Dakotan keep the
active pronominals in reflexive constructions.  What about Crow and Hidatsa?
The 1st plural pronominal should reliably indicate which set is used (1st and
2nd singular might not, since the reflexive prefix begins with /i/).  Or Mandan?

One of my students *wrote a dissertation on the grammatical evolution of
reflexives in European languages (Germanic, Romance and Slavic).  It isn't as
heavy on theory as I might have liked, but his implicational (evolutionary?)
hierarchy was:

1.  reflexive
2.  reciprocal
3.  inchoative (inan. subj.)
4.  inchoative (anim. subj.)
5.  impersonal (transitive)
6.  impersonal (intransitive)
7.  agent-licensing adverbial (inan. subj.)
8.  agent-licensing adverbial (anim. subj.)
9.  resultative (inan. subj.)
10. resultative (anim. subj.)
11. agentive passive (inan. subj.)
12. agentive passive (anim. subj.)

It would be interesting if Dakotan reflexives split along these lines somehow,
but I take it that all reflexive constructions with /ic?i-/ require the same
subj. pronominal set.

Bob

*Sohn, Joong-Sun. 1998.  _The Functional Evolution of the Reflexive Pronoun in
Romance, Slavic and Germanic_.  Univ. of Kansas Ph.D. dissertation.  University
Microfilms International.

----- Original Message -----
> This is not typologically all that unusual -- Geraldine Legendre has
> argued that the reflexive in French is a de-transitivizing morpheme, not a
> direct object (cf. the conjugation with "etre", e.g.)  Cross
> linguistically both patterns seem to occur.  Of course, I agree that it's
> odd to see fairly closely related languages split on this.
> David