| |
Abstract:
In the theory of grammatical encoding, conceptual factors are
thought to only influence the assignment of syntactic functions,
but not in the creation of linear order (Bock & Warren, 1985).
But there is some evidence that this claim depends on the syntactic
options that a particular language allows (Prat-Sala &
Branigan, 1999). If a language has relatively free word order, then
we might expect that conceptual accessibility might be allowed to
have a strong influence on word order. In this study, we examined
whether conceptual factors would influence word order in a fairly
free-word order language (Japanese) that mainly uses particles to
mark syntactic functions.
We manipulated animacy in three constructions: dative,
transitive, and conjuncts. The dative sentences had three arguments
(marked with nominative, accusative, and dative particles) and the
verb. The animacy of the dative argument was set to either be a
human (animate) or an organization (inanimate). The transitive
sentences had two arguments (marked with nominative and accusative
particles) and the verb. The animacy of the nominative argument was
set to be either a human (animate) or a thing or event (inanimate).
The conjunct sentences were made up of two phrases, each marked
with the particle mo (which generally means also), and an adjective
or verb which described both arguments. One argument was varied so
that it was either a human (animate) or thing (inanimate).
The phrases that composed each sentence were separated, and
presented to subjects in four boxes on a computer screen. Subjects
were told to create a sentence by clicking on the boxes. After each
click, subjects could see the sentence that they were building in
separate box. Each of the sentences types alternated with fillers
and animacy was counterbalanced between subjects.
The dependent measure was the percentage of sentences which
placed the manipulated argument before a non-manipulated argument.
For the datives, we found a trend for dative arguments to precede
accusative arguments if the dative argument was animate rather than
inanimate (significant by subjects, but marginal by items). But for
the transitives and the conjuncts, we found no effect of animacy on
the ordering of phrases.
For the view that conceptual factors only influence syntactic
functions, the finding that internal arguments in datives are
sensitive to conceptual factors is difficult to explain. For the
view that conceptual factors can directly influence linear order in
some languages, our results suggest even when a language has a
variety of word order configurations, it seems that conceptual
factors are not strongly guiding the selection of linear
orders.
Bock, J. K., & Warren, R. K. (1985). Conceptual
accessibility and syntactic structure in sentence formulation.
Cognition, 21, 4767.
Prat-Sala, M., & Branigan, H. P. (1999). Discourse
constraints on syntactic processing in language production: a
cross-linguistic study in English and Spanish. Journal of Memory
& Language, 1-15.
|
