Re: Referent Tracking

>--
>M.H. Klaiman, in his Grammatical Voice book, discussed some Mayan
>languages which had both Topic marking and Focus marking.  If I both
>remember correctly, and understood correctly in the first place, the
>verbs in these languages were marked for Voice, in order to specify the
>semantic roles played by the Topic- and Focus- marked participants.
>
>
I would be very interested to see more details about these systems. Next
time I'm at university I'll look see if they own the book you mentioned.

>I suppose that may not have been all there was to that system; just as
>that isn't all there is to the Focus system in Tagalog, and the Trigger
>system in Tagalog is, according to Chris, not only about Focus; and, as
>Klaiman says, in the Philippine languages with Information-Salience
>Voice, "the closest thing to a 'Subject' is a focussed agent".
>
>
The paper I cited agrees that that is the only type of argument that has
all the features associated with subject in other languages. I have
though used the terms "subject" and "object" in relation to Tagalog
because, as I mentioned, the syntactic pivot in Tagalog is S (argument
of intransitive clause) + A (actor of transitive clause)... ie
syntactically Tagalog does exhibit in some ways the notion of subject,
but the subject role is seriously impoverished in other ways compared to
other languages except when the subject is also the trigger.

>--
>As for pivots -- is it not the case that, in some ergative natlangs,
>the pivot is indeed the patient (the absolutive argument)?
>---
>
>
Yes, although the fact remains that in such languages, as in typical
languages with nominative alignment, the syntactic pivot is mainly role
defined (ie S + A or S + P). In topic languages, by far the most
important thing for resolving things like zero anaphora is not the
identity of the subject or absolutive of the previous clause but the
identity of the discourse topic.