| |
Abstract:
n verb agreement, the critical information about a subject
noun's number can come from two possible sources: (1) the abstract
grammatical number reflected in the subject noun's morphological
form, and (2) the conceptual number of the referent of the subject
noun. Previous research examining the coordination of these two
sources of information during the encoding of subject-verb
agreement in language production has produced contradictory
results. Bock and colleagues (e.g., Bock & Miller, 1991) have
shown that conceptual number has no effect on verb agreement;
whereas Eberhard (submitted) and Vigliocco and colleagues (e.g.,
Vigliocco, Butterworth, & Garrett, 1996) have demonstrated that
conceptual information significantly influences verb agreement.
Eberhard (submitted) interpreted this discrepancy as a difference
in the imageability of the phrases. Following Bock and Warren
(1985), Eberhard hypothesized that the effect was due to speakers'
ability to access the referential features during the functional
level of grammatical encoding. However, Eberhard was unable to show
that an immoderately accessible discourse context (e.g., a picture)
would strengthen the influence of conceptual information. The goal
of this research is to re-examine the influence of a discourse
context when confounding factors are controlled, and to delimit the
influence of accessible conceptual information during grammatical
encoding.
In each experiment, speakers were presented pictures followed by
complex noun phrases that identified referent(s) in the pictures.
Speakers used the phrases as beginnings of sentences that they
completed based on the pictures. Similar to Eberhard (submitted),
the noun phrases in the first experiment were grammatically
singular complex noun phrases with singular head nouns followed by
prepositional phrases (PP) that contained plural local nouns which
manipulated conceptual number. Controlling for the felicitous use
of the post-modifying PP, the pictures depicted not only the
referent(s) of the phrases, but also another possible referent of
the head noun. Consistent with Eberhard's results there was
significantly more plural agreement following conceptually plural
phrases.
In the second experiment, discourse contexts inconsistent with
all sources of grammatical information were examined by presenting
noun phrases headed by mass nouns and followed by singular object
nouns. Two pictures were presented with each phrase; one with a
singular and one with a plural referent. More plural agreement was
produced following phrases that accompanied conceptually plural
pictures. The results of these two experiments will be discussed in
terms of Garrett's model of language production.
|
