required subjects to note an abstract relational similarity that required recognition
that several grammatical functions co-occur. Thus, we would expect verb sorts to
have an inherent advantage over constructional sorts.
Nonetheless, six subjects produced entirely construction sorts. Seven other sub-
jects produced entirely verb sorts, and four subjects provided mixed sorts. To include
the mixed sorts in the analysis, the results were analyzed according to how many
changes would be required from the subject’s sort to produce a sort entirely by verb
(VS) or a sort entirely by construction (CS). The average number of changes required
for the sort to be entirely by verb was 5.5; the average number of changes required for
the sort to be entirely by construction was 5.7. The difference between these scores
does not approach significance. That is, subjects were just as likely to sort by construc-
tion as they were to sort according to the single dimension of the morphological form
of the verb. If verbs provided equally good cues to overall sentence meaning, there
would be no motivation to overcome the well-documented preference for one-dimen-
sional sorts: Subjects would have no motivation to sort by construction instead of by
verb. Bencini and Goldberg (2000) hypothesize that constructional sorts were able to
overcome the one-dimensional sorting bias to this extent because constructions are
better predictors of overall sentence meaning than the morphological form of the verb.
A question arises about why constructions should perform at least as well as pre-
dictors of overall sentence meaning as verbs. The answer, we believe, stems from the
fact that in context, knowing the number and type of arguments conveys a great deal
about the scene being conveyed. To the extent that verbs encode rich semantic frames
that can be related to several different basic scenes, the complement configuration or
construction will be as good a predictor of sentence meaning as the semantically
richer but more flexible verb.
On the multiple-sense view, the reason instances of
throw
, for example, were put
into separate piles was that each instance represented a distinct sense that was more
similar in meaning to one of the senses of another verb than to the other senses of
throw
. The only way for subjects to discern which verb sense was involved, however,
was to recognize the argument structure pattern and its associated meaning. That is,
the proposed different verb senses all look the same; the only way to determine that a
particular sense is involved is to note the particular argument structure pattern that is
expressed and infer which verb sense must have produced such a pattern. Therefore,
at least from a comprehension point of view, the pairing of argument structure pattern
with meaning must be primary.
The most important contribution of the studies cited in this section are that they
provide a sufficiency proof that types of complement configurations play a crucial
role in sentence interpretation, independent of the contribution of the main verb. The
results suggest that constructions are psychologically real linguistic categories that
speakers use in comprehension.
3
Evidence for Constructions in Language Production
In this section we provide experimental evidence that units of the type and kind of
constructions are activated during the processes of language production. We present
three studies. These three studies build on a large body of work conducted primarily
by Bock and colleagues that employ a structural priming methodology (Bock 1986).
10
Do'stlaringiz bilan baham: