| |
Abstract:
Intuitively, our understanding of short passive sentences like
"The ship was sunk" and intransitive sentences like "The ship
sank" differ in that only the first entails an agent responsible
for sinking the ship. Recently Mauner, Tannenhaus, and Carlson
(1995) demonstrated that readers' representations of short
passive but not intransitive sentences include an unexpressed
agent. Although they hypothesized that implicit agents are part
of a verb's representation, agent encoding was not evaluated at
the passive verb but rather at a later word position. We
demonstrate that implicit agents are encoded immediately as soon
as a passive verb is encountered.
Cross-modal naming, which has been used successfully to
demonstrate that readers immediately integrate information into
sentence representations (e.g., Marlsen-Wilson, 1973; Tannenhaus,
Trueswell and Kello, 1993), was used to determine whether readers
encode implicit agents as soon as they recognize a passive verb.
To engender an expectancy for an agent, sentence-initial
rationale clauses, which are infinitives that require an agent in
an adjoining clause, were paired with short passive sentences
whose verbs are hypothesized to introduce an implicit agent (1a)
and intransitive sentences which have no agent in their lexical
representations (1b). Crucially, short passive and intransitive
probes were identical. We predicted that verbs which satisfy the
agent requirement would be named faster than verbs which do
not.
(1a) To reduce the noise from next door, the heavy door was
... SHUT.
(1b) To reduce the noise from next door, the heavy door ...
SHUT.
Sentences were auditorily presented up to but excluding the
final verb and were followed by visually presented verb probes
for a naming response. As predicted, naming times for passive
verbs with hypothesized implicit agents were significantly faster
than for intransitive verbs whose representations included no
agent. This demonstrates that implicit agents are encoded
immediately upon encountering passive verbs.
This finding replicates Mauner et al.'s results using a
different experimental methodology and extends them by
demonstrating that implicit agents that are hypothesized to be
part of the representations of passive verbs are encoded as soon
as these verbs are recognized. Additionally, in contrast to
Boland's (1997) arguments that cross-modal naming is sensitive to
only syntactic integration effects, these results clearly show
that cross-modal naming is also sensitive to semantic information
(i.e., verb argument structure) when it is not conflated with
semantic plausibility computed from conceptual knowledge as it
was in Boland's studies.
Boland, J.E. (1997). "The relationship between syntactic and
semantic processes in sentence comprehension."
Language and Cognitive Processes,
12, 423-484.
Marslen-Wilson, W.D. (1973). "Linguistic structure and speech
shadowing at very short latencies."
Nature,
244, 522-523.
Mauner G., Tanenhaus, M.K., and Carlson, G.N. (1995). "Implicit
arguments in sentence processing."
Journal of Memory and Language,
34, 357-382.
Tannenhaus, M.K., Trueswell, J.C., and Kello, C. (1993). "Verb
specific constraints in sentence processing: Separating effects
of lexical preference from garden-paths".
Journal of Experimental Psychology: Learning, Memory, and
Cognition,
1 9, 528-553.
Trueswell, J.C., Tanenhaus, M.K., and Garnsey, S.M. (1994)
"Semantic influences on parsing: Use of thematic role information
in syntactic ambiguity resolution."
Journal of Memory and Language,
33, 285-318.
|
