| |
Abstract:
Fodor (1998) suggests that the Human Sentence Parser has a
tendency to equalise the length of sister constituents. The aim of
this poster is to investigate the effect of constituent length on
attachment preference of relative clauses to complex noun phrases
(N1 of N2). In French, experiments in which the length of
constituents was not manipulated showed a preference for high
attachment (Zagar, Pynte and Rativeau, 1997; Frenck-Mestre and
Pynte, in press). However, in a series of eye-tracking experiments
manipulating the length of the relative clause, Pynte and Colonna
(in press) observed a tendency to attach high with long relative
clauses and to attach low with short relative clauses.
Here we report an eye-tracking experiment in which we
manipulated the length of NP1 and NP2 by adding an adjective either
to N1 or N2 (examples a-b).
(a) Il aime la souriante soeur des Anglais qui arrive(nt)
à l'agence de voyages.
(He loves the smiling sister of the British who arrive(s) in the
travel agency)
(b) Il aime la soeur des souriants Anglais qui arrive(nt) à
l'agence de voyages.
(He loves the sister of the smiling British who arrive(s) in the
travel agency)
If Fodor is right, we should observe a preference for low
attachment in (b), and a preference for high attachment in (a).
Alternatively, Thornton, Gil and MacDonald (1998) claim that a
pragmatic constraint (the felicity of NP modification) influences
the resolution of structural attachment ambiguities. Accordingly,
readers should tend to avoid the presence of two modifiers
(adjective and relative clause) within the same NP, in which case,
the reverse interaction should be observed, i.e. a preference for
low attachment in (a), and a preference for high attachment in
(b).
The results showed no clear effect on the disambiguating region,
but first pass gaze duration for the following region (end of
sentence) revealed the interaction predicted by Fodor
(F1(1,28)=4,58, p<.05; F2(1,20)=4,23, p<.05), i.e. a
preference for low attachment in (b) (674 vs. 708 ms for low and
high attachment respectively), and a preference for high attachment
in (a) (736 vs. 649 ms for low and high attachment respectively).
The results will be further discussed in the context of Fodor's
proposals.
Zagar, D., Pynte, J., & Rativeau, S. (1997). Evidence for
early closure attachment on first pass reading times in French. The
Quaterly Journal of Experimental Psychology, 50A (2), 421-438.
Pynte, J., & Colonna, S. (in press). Decoupling syntactic
parsing from visual inspection: The case of relative clause
attachment in French. In A. Kennedy, R. Radach, D. Heller, & J.
Pynte (Eds.), Reading as a Perceptual Process. Oxford:
Elsevier.
Thornton, R., Mariela, G., and MacDonald, M. C. (1998).
Accounting for cross-linguistic variation: A constraint-based
perspective. In D. Hillert (Ed.), Syntax and semantic, sentence
comprehension: A croos-linguistic perspective. Academic Press.
Fodor, J. D. (1998) Learning to parse? Journal of
Psycholinguistic Research, 27(2), 285-319.
Frenck-Mestre, C., & Pynte, J. (in press). Romancing
syntactic ambiguity: Why the French and the Italians don't see eye
to eye. In A. Kennedy, R. Radach, D.
Heller, & J. Pynte (Eds.), Reading as a Perceptual Process.
Oxford: Elsevier.
|
