| |
Abstract:
Psycholinguists have long known that the parsing of syntactic
structure takes place incrementally, frequently on the basis of
incomplete information. Recent evidence using eye movements has
also demonstrated the incrementality of semantic interpretation
with respect to a visually-presented model (Eberhard et al., 1995).
The use of an eye monitoring technique allows for the investigation
of spoken language processing in a fairly naturalistic environment
that involves no special skills (such as reading) or specific tasks
(such as phoneme monitoring), making this a potentially useful
technique for studying processing in children. Trueswell, Sekerina
and Hill (1998) reports data from eye movement patterns with
children investigating the use of visual context in syntactic
ambiguity resolution.
The present study investigates more fine-grained temporal
measures in stimuli that involve temporary referential
indeterminacies, but no syntactic ambiguity. Fifteen 5-7 year-old
children heard spoken instructions such as "Pick up the red cat",
and were required to manipulate cards with colored images of
animals. Trials containing objects that resulted in referential
ambiguity at the earliest content word in the critical noun phrase
(e.g., two red animals) were compared with trials for which it was
possible to identify a unique referent at this point. Latencies of
eye movements to the target object as measured from the onset of
the head noun were significantly different across the two
conditions (335 ms in the No Ambiguity condition vs. 571 ms in the
Ambiguity condition, p< 0.01), indicating that children were
attempting to incrementally identify the referent for the noun
phrase on the basis of partial evidence. In addition, eye movements
to the "competing" object in that Ambiguity condition (e.g., the
second red animal) were compared to an object of a different color
in the same location in the No Ambiguity condition. A notable
difference between these results and previous results from similar
displays with adults (Eberhard et al., 1995) was that children were
more likely to look at objects in the display that were not related
to the speech stream; however results still show a significant
difference in the percentage of trials that contained a look to the
"competing" object in the Ambiguity condition (72%) as opposed to
the unrelated object in the "No Ambiguity" condition. We are
currently using the same experimental materials to test referential
comprehension in 3-4 year-olds.
References
Eberhard, K., Spivey-Knowlton, M., Sedivy, J., & Tanenhaus,
M. (1995). Eye movements as a window into real-time processing in
natural contexts.
Journal of Psycholinguistic Research,
24, 409-436.
Trueswell, J., Sekerina, I., & Hill, N. (1998). On-line
sentence processing in children: Evidence from eye movements during
listening. Paper presented at the 11th Annual CUNY Conference on
Human Sentence Processing, Rutgers University, New Brunswick,
NJ.
|
