Date: Tue, 05 Nov 1996 16:19:50 -0500 (EST)
From: ENG_SEELY@EMUVAX.EMICH.EDU
Subject: DEFINING ANAPHORS, AN INVITATION
To: LINCONF@tamvm1.tamu.edu
Message-Id: <01IBHRE0G2EQ9BXKIK@EMUVAX.EMICH.EDU>
COMMENT AND INVITATION
A. COMMENT/QUESTION
I know it's late, and this is a rambling thing which I don't have the
time to tighten up, but I'll put it out just the same.
A general question, to all authors (and conference participants):
How are anaphors defined?
With the related question:
What is binding theory about?
[These are prompted by a number of observations, just two of which
I'll trace here:
1. If an anaphor is defined as a referentially dependent element with
a theta-role distinct from its antecedent (the "standard" definition),
then switch reference markers (see Broadwell) are not anaphors.
[The same holds of other extensions of BT; for instance, to
negative polarity elements (in Progovac's recent work).
NP-trace is similar too (an NP-trace and its antecedent have the same
theta-role and a trace is created during the course of a derivation--on
many views) but at least for Chomsky (1981) it has the same feature
makeup as (semantic) anaphors.]
But, in the spirit of Burzio (1991) "The morphological basis
of anaphora," J. Linguistics 27, it is important to have
a definition of anaphor independent of the binding principles.
(And in sorting out what progress we've made on our conference
theme; it seems relevant.)
Are SR markers anaphors because they must be locally bound?
Well, that yields a tautology:
X is an anaphor if X is locally bound.
An anaphor is locally bound (BT-A).
Thus: X is locally bound if X is locally bound.
Are SR markers anaphors morphologically?
Apparently not (see Broadwell's conference discussion).
SR markers are "treated" as anaphors since doing so allows a more
parsimonious analysis of a certain range of data.
But without an independent definition of anaphor, we face
Burzio's problem.
2. It is still not entirely clear to me the extent to which an anaphor
determines the semantics (i.e. interpretation) of a predicate (in the
relevant sense) or the predicate determines the choice of the anaphor.
(In the Lidz-Reinhart discussion and this is relevant in the Reeves
paper as well.)
Lidz himself (usually) seems to say that differences in interpretation
lie in the predicate, not in the reflexive elements themselves.
Reinhart's response is that "SELF anaphors in reflexive predicates have
properties different from intrinsic reflexivization."
and Reinhart says: "...the SELF anaphor has the property that it can be
interpreted as in (12b)."
But Lidz isn't always entirely clear:
"What the Madame Tussaud examples illustrate is that SELF anaphors
allow an interpretation in which the anaphor is referentially
dependent on its antecedent but not necessarily identical with it."
What is clear from Lidz, Reeves, and others is that the standard
notion of coindexation entailing coreference (of the usual
sort given QPs, etc) will not do. But again, how are the
various types of anaphors defined?]
B. AN INVITATION
I would also like to invite everyone to perhaps give
a final statement on where they stand on the central
issues of this conference. It could be a good way
for us to assess what progress might have been made.
--Daniel Seely