argument structure k'u etc.

[Koontz John E]

On Fri, 1 Apr 2005, ROOD DAVID S wrote:
> 	Thanks for the question.  To my way of thinking, the decisive
> behavioral property is the indexing on the verb.

"Indexing" is the word I was looking for, I think, where I cobbled
together an expedient "representing."

I think it's been suggested that languages which select the "higher
ranked" object (recipient over patient) be called "primary object
languages," the primary object being the recipient if specified or
specifiable and the patient otherwise.  However, I'm not sure this is much
more than a more an expanded terminology for "case marking/agreement"
patterns.  It labels the pattern, but doesn't explain anything.

> Contrast the causatives, which have the morphological tools for
> expressing three arguments: I made you buy the horses can be
> s^uNkawakhaN ki ophewichathun-chiye.

Does it seem fair to say that causatives can do this because they have two
stems to attach indexing morphology to?  Also, is the case that the kind
of indexing morphology allowed with the embedded stem (here ophe..thuN) is
restricted to third person marking?  If so this may be a way of
demonstrating that third person (plural) marking is different in character
from non-third person marking - less inflectional, more derivational.

Dakotan is a bit of a special case because of wic^ha-, but my inclination
in Dhegiha is to see parallel uses of wa- as a "third person plural
object" as a sort of drafting of the derived indefinite object form as an
indexed form.

> 	I can't think of any purely syntactic tests for deciding if
> something is an "argument" or not, ....

Me either, though I'm guiltily aware that this may be because I haven't
looked.  It's something of a relief that your more sophisticated and
extended investigations of Lakota also leave you without any ready
instances of such tests!  Does anyone else have any suggestions?

kk