Siouan and other highlighting morphology.

[R. Rankin]

I agree with Ardis.  Moreover, this seems very often to be a feature of
"pronominal argument" languages like Siouan and Muskogean.  Since arguments are
clarified by prefixes on the verb, the actual nominals in the sentence (which
many linguists take *not* to be "arguments" of the verb) strongly tend to have
their *own* grammar.  And that grammar is most often discourse-based, i.e., its
morphology is *not* marking grammatical relations such as "subject" or "object".
Rather it is signaling the sort of relationships Ardis is suggesting for what we
call "proximate/obviative" in Dhegiha.  It may be topicalizing or focusing or
simply highlighting, among other things.

In Muskogean languages, for example, there are discourse markers (-t and -n)
that signal "center-stage" and "off-stage" -- -t highlights what is central.
But within Muskogean, Choctaw has additional particles that mark 'topic' (-o$).
For years these -t/-n particles were erroniously described as marking 'subject'
and 'object/oblique', but then, when identical particles occurred on entire
clauses they were described as marking 'switch-reference'.  But
'switch-reference' is another misnomer; it's all one system, and it marks
central vs. "obviative" in the discourse.  The system is free to do this because
the essential argument structure is covered by the pronominal prefixes.  And I
suspect such systems are much more prevalent than linguists have believed (the
noun and identical clause markers in Walapai are another case).

I really hate to move away from sentence-based grammar and into discourse
myself, but if we're going to understand these sub-systems, like article and
(-abi) verb suffix use in Dhegiha, it is the direction we have to move.  Lead
on, Ardis.

Bob

> 4. THe articles should really not be called 'focus' markers.  They
> don't mark the linguistic concept of focus regularly.  Often, they are
> marking given material (topic - not focus, which is new material).
> Consistently, however, they mark characters (or things, as per 3)
> which are of central concern, centerstage in narratives, topic of
> conversation.