Dative Marking and the Primary Argument Language Type

[Koontz John E]

On Sat, 2 Apr 2005, ROOD DAVID S wrote:
> Bob's comment about the Dryer "primary/secondary" argument typology is
> relevant, except that Lak. has a very robust "dative case" that it can
> and does utilize quite often to show "secondary" argument (i.e.
> obliquely marked) status for recipients and beneficiaries.  I have
> somehow internalized the idea that Dryer's theory is a typology of
> systems (some languages do it one way, some the other), not a variable
> for individual verbs.

This may be one of those places where I know just enough to be dangerous,
so please correct me if this is wrong!

I had the impression that the dative construction in Lakota (and
Omaha-Ponca and so on) amounted to a device for permitting a
recipient-object to occur with a verb that, in underived, non-dative form,
agreed with its patient-object.  In effect, the dative marker is a way of
deriving a ditransitive verb from a mono-transitive one.  And ditransitive
verbs in primary object languages have to agree with the recipient-object.
You can also think of it as a way of marking the case of the primary
object as dative.  You can a;sp think of it as a requirement to raise the
expressed possesor of an object to object.  There are other ways it could
be done, but "dative marking" in the verb is the mechanism in most Siouan
languages.

I tend to think that some mechanism like this is more or less essential in
a primary object language in which the primary object is indexed in the
verb.

Anyway, from this perspective I'm not sure I would want to describe the
dative marker as way to "show secondary argument (i.e. obliquely marked)
status for recipients and beneficiaries."  Actually, I think it
substitutes them for the patient-object as primary arguments."  The
recipient does kidnap the patient indexing property, right?  (This may be
where I am confused.)  So, it seems to me that the old primary argument
that is now secondary, which is as it should be relative to a recipient in
a primary argument language.  I guess we can't call the patient object
oblique, since oblique (or indirect) is a term for direct object systems.

Along these lines, I don't believe there is any way in Lakota or
Omaha-Ponca to include a recipient-object in a mono-transitive verb
clause.  So, unlike English, you can't say 'give x to y' but only 'give y
x'.  Nor is there any way in Omaha-Ponca that I can see to form a
mono-transitive from a ditransitive, e.g., to convert ?i 'give' into a
verb that agrees with the patient-object and doesn't mention the
recipient-object or demotes it to an adpositional form like 'to y' in
'give x to y'.  It may be possible to do something along those lines with
wa- in Lakota but I wasn't clear on the details.  It looked to me like
this was an area where Omaha-Ponca and Lakota went different ways.