| |
Abstract:
Subject-verb agreement processes have received much recent
attention in the production literature. Results from these studies
have suggested that agreement errors are more likely when a noun
intervenes between the subject and its verb, specifically when the
local noun and the head noun mismatch in number (i.e., "the key to
the cabinets were"). Moreover, a number of studies have found
little or no effect of semantic or phonological factors on
agreement in English, suggesting that it is a primarily grammatical
phenomenon (although studies in other languages have found robust
semantic effects). For example, Bock and Miller (1991) manipulated
the animacy and concreteness of the intervening noun, reasoning
that animate/concrete nouns may be more likely to disrupt agreement
due to their high correlation with subjecthood. They observed
little effect of either variable.
The present study provides a more stringent test of semantic
factors by manipulating not only the animacy of the local noun, but
the plausibility of its relation to the verb, which has been shown
to have significant effect on sentence comprehension (e.g.,
Garnsey, Pearlmutter, Myers, & Lotocky, 1977). In our
production task, a modified version of the task used by Bock and
Miller (1991), participants saw a verb on a computer screen,
followed by the auditory presentation of a sentence preamble. They
were asked to create a complete passive sentence beginning with the
preamble and using the verb, and errors on the auxiliary (always
"was" or "were") were measured. Both the match between the number
of the nouns (match versus mismatch) and plausibility of the local
noun as a subject of the passive (plausible versus implausible)
were manipulated. Plausibility ratings norms were conducted to
ensure that the manipulation was robust. Example stimuli are
presented in (1).
The results revealed a significant effect of plausibility, such
that agreement errors were more likely in the implausible condition
than in the plausible condition. Thus, semantic factors do seem to
affect subject-verb agreement in English; our results suggest that
agreement may be easier when the head noun is the only plausible
subject for the verb. The results of a related comprehension
experiment will also be discussed, highlighting the links between
the production and the comprehension of agreement.
|
