| |
Abstract:
It is a matter of debate whether during sentence
comprehension the on-line resolution of local structural
ambiguities is achieved by syntactic information only, or whether
other kinds of information (e.g., lexical, thematic, or pragmatic)
are used as well. In Dutch a subject-initial structure is preferred
over an object-initial structure. In a previous experiment
event-related potentials (ERPs) showed early and late positivities
for the object-initial structure. The results indicate that
syntactic processes were involved in resolving the local ambiguity
(Lamers, 1995). To address the question whether thematic/pragmatic
information can play a role in disambiguating a sentence with a
local structural ambiguity the animacy of the first NP was
manipulated. Sentences starting with an animate NP were
disambiguated by the case-marked pronoun
hij
("he") (1a); sentences starting with an inanimate NP were
disambiguated by thematic/pragmatic information at the verb (1b).
The sentence starting with an animate NP (1a) can be parsed
without any problems until the case-marked pronoun
hij
("he") , in contrast to the sentences starting with an inanimate
NP (1b). The sentences of both conditions will be initially parsed
as subject initial sentences, which is the preferred structure. At
the verb it turns out that the inanimate NP can not be the subject
of the sentence, while the animate NP is fine. The inanimate NP
does not fit the thematic role of the external argument of the
verb. This problem can be resolved by reanalysing the sentence as
an object-initial structure, in which the verb can assign the role
of theme to the internal argument. Although thematic/lexical
information is involved in resolving the local ambiguity, a
positivity like the P600 was expected at the verb for the sentences
starting with an inanimate NP (1b) compared to the sentences
starting with an animate NP (1a), since the problem lies in
building the appropriate structure rather than in semantic
interpretation.
At the verb, sentences starting with an inanimate NP showed both
early and late positivities compared to the sentences starting with
an animate NP. This held in both subject- and item-analyses. The
early positivity starts as early as 200 ms after onset of the verb,
and is found over posterior electrode sites. A sustained later
positivity occurs in the time frame of the P600 and is more broadly
distributed. No differential negative shifts were found. The
results complement results in the literature (cf. Hagoort, Brown,
& Groothusen, 1993; Mecklinger, Schriefers, Steinhauer, &
Friederici, 1995; Osterhout, Holcomb, & Swinney, 1994), and
indicate that thematic information can be used early in resolving a
local structural ambiguity, and that the underlying processes are
more related to syntactic processes than to semantic processes.
References
Hagoort, P., Brown, C., & Groothusen, J. (1993). The
syntactic positive shift as an ERP-measure of syntactic processing.
Language and Cognitive Processes,
8, 439-483.
Lamers, M. J. A. (1996). Parsing Dutch sentences: Ambiguity
resolution. In R. Jonkers, E. Kaan, & J. A. Wiegel (Eds.).
Language and Cognition 5. Yearbook 1995 of the Research Group
for Theoretical and Experimental Linguistics of the University of
Groningen.
Groningen.
Mecklinger, A., Schriefers, H., Steinhauer, K., & Friederici,
A. D. (1995). The processing of relative clauses varying on
syntactic and semantic dimensions: An analysis with event-related
brain potentials.
Memory and Cognition,
23, 477-494.
Osterhout, L., Holcomb, P. J., & Swinney, D. A. (1994). Brain
potentials elicited by garden-path sentences: Evidence of the
application of verb information during parsing.
Journal of Experimental Psychology: Learning Memory and
Cognition,
20, 786-803.
|
