| |
Abstract:
The means by which the structure-based constraints of Binding
Theory interact with other information sources in the
interpretation of referentially dependent expressions has been
characterized in two conflicting ways. Nicol and Swinney (1989)
have argued from selective associate priming results that the
Binding Principles function as an initial filter on all
content-based processing. On this account, a syntactically
inaccessible phrase should be invisible to mechanisms that
identify, evaluate and select candidate antecedents. In contrast,
Badecker and Straub (1998) have argued, based on results from
word-by-word self-paced reading studies, that very early evaluation
processes can have access to candidates that are excluded by
Binding Theory principles. They argue that associate priming
provides an incomplete picture of the early evaluation process, and
that if a more direct means of probing candidate activation were
employed, then one should observe evidence for activation of both
accessible and inaccessible candidates.
This paper presents data from three cross-modal
probe-recognition experiments which support this claim. In one
experiment, probe names are presented either immediately at the
off-set of a pronoun, or 500 ms following the pronoun. At the early
probe position, both accessible and inaccessible candidate names
were facilitated in comparison to a no-pronoun control; while at
the later probe position, response times to the two probes
diverged: the probe for the inaccessible candidate showed
inhibition while the probe corresponding to the accessible
candidate continued to show facilitation. A similar priming pattern
was observed for contexts with reflexive anaphors in the second and
third experiments presented here. Four probe points were examined.
When the onset of probe presentation was at the identification
point of the reflexive (i.e., at the second vowel in the
reflexive), no effect was observed in comparison to the no-anaphor
control. However, when probe onset occurred 500 ms after the
off-set of the anaphor, then facilitation for accessible candidate
probes and inhibition for inaccessible candidate probes was
observed. Evidence of facilitation for both the accessible and
inaccessible candidate probes was observed at two intermediate
probe points. These data support the view that at the onset of
coreference calculation, evaluation processes may have access to
antecedent candidates which will subsequently be excluded as
candidates on the basis of the structure-based principles of
Binding Theory.
References
Badecker, W., & Straub, K. (1998). The processing role of
structural constraints on the interpretation of pronouns and
anaphors. Manuscript, Johns Hopkins University.
Nicol, J., & Swinney, D. (1989). The role of structure in
coreference assignment during sentence comprehension.
Journal of Psycholinguistic Research,
18, 5-19.
|
