Re: Referent Tracking

> Are (wannabe) trigger systems also Topic systems? (Not
> intending to tread loose [lostreten?] a debate about trigger
> systems)
 Leaving aside for now various arguments about how exactly to
characterise a trigger system, I would say no, for this reason: although
the topic of a clause is what the clause is about, there is no
restriction in languages which mark topic separately from grammatical
roles (ie, what most people think of as a topic system) for the topic to
occur in a certain subset of roles, whereas with a trigger system, while
the choice of the trigger does seem to be motivated by factors related
to topicality, the referent chosen to be trigger must occur in a certain
subset of possible roles. Indeed, the stereotypical "double subject"
(that elephant, trunk is long) construction that such topic languages
possess is, I think, not possible in (as far as I know all) trigger
languages such as Tagalog.
 This is, of course, assuming that the trigger can be identified with
the topic of the clause. While they do share many features in common
(for instance, in Tagalog the trigger must be chosen to be definite
whenever possible, and similar requirements hold for the topic in topic
languages) there also do seem to be differences between the two. One big
unexpected use of the trigger if it is in fact the topic is the sheer
number of times the trigger is the object (and indeed the fact that
object focus seems to be more basic from a marking point of view than
subject focus)... since subjects are generally much more topical than
objects, you'd expect the opposite if the trigger were simply the topic
in such languages.
 We also have to consider the question of syntactic pivots. In a
language like Japanese or Chinese with a well developed topic system,
typically the topic acts much more like a syntactic pivot than any other
role. This simply isn't true in Tagalog and probably other Phillipine
languages with trigger systems... in Tagalog, according to papers I've
read on the issue, the syntactic pivot remains the subject role even
when the trigger is not the subject. Thus, while the trigger system
(whether you consider it voice or not) alters the pragmatic marking of
the clause (equivalent to say, stressing a pronoun or topic fronting in
English), it doesn't alter which argument controls zero anaphora,
whereas changing the topic in a topic language does.
 This is of course only relevant to the trigger systems as found in
Tagalog and related languages. Ayeri, which I think you've said before
is a trigger language, needn't function in this way, and its triggers
may indeed be much closer to topics in topic languages. The main
questions are, I think:

1) are your triggers the syntactic pivot of the clause?

2) what factors are involved in the choice of trigger?

[If you want to read more about trigger vs topic vs subject in Tagalog,
I suggest you read a paper in the book:

Subject and Topic
Edited by: Charles Li

If you can get hold of it, that is]

Replies