| |
Abstract:
Frequency effects of linguistic elements have long been
recognized in language comprehension, with more frequent elements
typically being processed more quickly or easily than less frequent
ones (e.g., Just & Carpenter, 1980). Constraint-based sentence
comprehension theories make extensive use of argument structure
frequency information in explaining ambiguity resolution (e.g.,
MacDonald et al., 1994). For example, Garnsey et al. (1997) studied
the sentential complement/direct object (SC/DO) ambiguity and
showed that readers had no difficulty at disambiguation in
sentences with SC-biased verbs, but read slower at disambiguation
in sentences with DO-biased verbs (compared to unambiguous
controls).
Argaman et al. (1998) applied work in lexical semantic theories
(e.g., Pinker, 1989) to suggest that such theories could also
account for frequency biases. Looking at argument structure
frequency measures for SC-taking verbs and corresponding nouns
(e.g.,
propose-proposal),
derived from both corpus and completion data sources, they found:
(1) overall reliable correlations across verb-noun pairs; (2)
reliable differences in frequency bias measures between semantic
categories identified by Levin (1993); (3) correlations across
verb-noun pairs within Levin's semantic categories; and (4) no
reliable influence of derivational morphology or amount of
noun-verb orthographic overlap on argument structure properties.
These results implicate semantics as a source for argument
structure biases, and the purpose of the current investigation was
to examine how well semantic category and finer-grained semantic
properties predict variation in comprehension difficulty in an
on-line ambiguity resolution study.
Participants read 36 sentences containing an SC/DO ambiguity or an
unambiguous control in a word-by-word self-paced reading paradigm.
Twelve verbs in each of three semantic categories, chosen to cover
different ranges of SC bias, were used to trigger the ambiguity:
manner of speaking (low SC bias; e.g.,
barked, whispered),
transfer of message (medium bias; e.g.,
demonstrated, showed),
and say verbs (high SC bias; e.g.,
claimed, declared).
The stimuli were always resolved as SCs but were mixed with
DO-resolved fillers and other structural types. A reliable
ambiguity effect was present for each verb type at disambiguation,
but the magnitude of the effects did not differ across verb type,
providing preliminary evidence that semantic category itself does
not have a substantial influence on ambiguity resolution. The
influence of finer-grained semantic properties is being
investigated by examining correlations between individual verb SC
biases and on-line performance. These results will be discussed in
terms of the relationship between broad lexical-semantic categories
and finer-grained semantics, and with respect to current theories
of sentence comprehension.
References
Argaman, V., Pearlmutter, N. J., Garnsey, S. M., Mendelsohn, A.
A., Randall, J. H., & Myers, E. (1998). "Lexical semantics as a
basis for argument structure frequency biases." Poster presented at
the 1998 meeting of the Psychonomic Society, Dallas, TX.
Garnsey, S. M., Pearlmutter, N. J., Myers, E., & Lotocky, M.
A. (1997). The contributions of verb bias and plausibility to the
comprehension of temporarily ambiguous sentences.
Journal of Memory and Language,
37, 58-93.
Just, M. A., & Carpenter, P. A. (1980). A theory of reading:
From eye fixations to comprehension.
Psychological Review,
87, 329-354.
Levin, B. (1993).
English Verb Classes and Alternations: A Preliminary
Investigation.
Chicago, IL: University of Chicago Press.
MacDonald, M. C., Pearlmutter, N. J., & Seidenberg, M. S.
(1994). The lexical nature of syntactic ambiguity resolution.
Psychological Review,
101, 676-703.
Pinker, S. (1989).
Learnability and Cognition
Cambridge, MA: MIT Press.
|
