| |
Abstract:
Two experiments examined the mechanism underlying syntactic
feature-tracking in comprehension. If features are handled by
passing them along a hierarchically-organized structural tree, then
structural distance should determine how much interference an
intervening number-marked element creates. Thus in (1),
painting(s)
is structurally closer than
house(s)
to
lamp
and to
was,
so
paintings
should interfere more with subject-verb agreement computations
than should
houses.
If instead a linear-string-based memory system handles syntactic
features, then linear distance should determine degree of
interference, and because
houses
is closer to the verb in (1), it should create more interference
than
paintings.
(1) The lamp near the painting(s) of the house(s) was damaged in
the flood.
For Experiment 1, twenty stimuli like (1) were constructed, each
with 4 versions created by varying the number of the second and
third nouns (N2 and N3). The head noun was always singular, and the
verb (always
was)
immediately followed N3. Sixty participants read the stimuli
(word-by-word self-paced reading with comprehension questions)
mixed with 90 fillers. At the verb and the following word combined
(was damaged),
an interaction was present, because the condition in which all
nouns were singular was faster than the other three, indicating
that an intervening plural either near the head noun or near the
verb created interference (of comparable size).
Eberhard's (1997; Bock & Eberhard, 1993) proposal that the
singular form is unmarked provides an explanation for this result:
Because the head noun and verb were singular, even a relatively
distant (but overtly marked) plural may have been able to
interfere. Experiment 2 addressed this possibility by examining
plural-head versions of (1). Twenty-four stimuli were created, like
those in Experiment 1, except that the head noun was always plural,
and the verb was always
were.
Sixty participants read the stimuli mixed with 72 fillers,
including 12 items which had a singular head noun and verb (always
was) but were otherwise identical in structure to the experimental
stimuli. At the verb and the following word, plural N2 conditions
were reliably faster than singular N2 conditions. There was no
effect of N3 number and no interaction. The pattern across the
experiments indicates that hierarchical distance between agreeing
and interfering elements, rather than linear distance, determines
the amount of interference; but that even relatively distant
elements can interfere when the original number-specifying element
and the agreement target are not strongly marked. This extends but
is generally compatible with other findings from sentence
production and comprehension, using the same or related
constructions (e.g., Bock & Cutting, 1992; Nicol et al., 1997;
Vigliocco & Nicol, 1997).
References
Bock, K., & Cutting, J. C. (1992). Regulating mental energy:
Performance units in language production.
Journal of Memory and Language,
31, 99-127.
Bock, K., & Eberhard, K. M. (1993). Meaning, sound and syntax
in English number agreement.
Language and Cognitive Processes,
8, 57-99.
Eberhard, K. M. (1997). The marked effect of number on subject-verb
agreement.
Journal of Memory and Language,
36, 147-164.
Nicol, J. L., Forster, K. I., & Veres, C. (1997). Subject-verb
agreement processes in comprehension.
Journal of Memory and Language,
36, 569-587.
Vigliocco, G., & Nicol, J. (1997). The role of syntactic tree
structure in the construction of subject verb agreement.
Manuscript.
|
