| |
Abstract:
What does it actually mean when it is claimed that a pronoun
has been assigned to its antecedent? Is it possible to determine
whether pronoun assignment has occurred if comprehension is
measured at only one location in a sentence? Research suggests that
text representations are dynamic (e.g. Greene, McKoon and Ratcliff,
1992) but has failed to investigate them in a dynamic way.
Measurements of comprehension that take place only at the referring
expression dont allow us to determine whether or not pronoun
assignment has started or ended. Previous research has implicitly
accepted the immediacy assumption (Just and Carpenter, 1980) that
each word in a discourse is immediately interpreted as deeply as
possible. Eye fixations are therefore maintained until processing
has finished ( the eye-mind assumption). Consequently, researchers
have accepted processing differences at referring expressions as
evidence for immediacy in reference resolution (e.g. Ehrlich, 1980,
Hielscher and Musseler, 1990, Sichelschmidt and Gunther, 1990).
Other evidence suggests that resolution may be delayed until post-
pronominal information is read (Ehrlich and Rayner, 1983, Garrod,
Freudenthal and Boyle, 1994, Hirst and Brill, 1980, Stevenson and
Vitkovitch, 1986).
In two word-by-word self paced reading tasks we manipulated
shifts in the focus of attention at two different sentence
locations. Two characters were introduced into each sentence, one
character occupied a thematic role that led to a focusing bias
towards that character (Arnold, 1997; Stevenson, Crawley and
Kleinman 1994; Wilson and Stevenson, 1998). This focus was either
shifted or maintained by a pronominal reference (with a gender cue
in expt. 1; no cue in expt. 2) and at a subsequent verb (which
could only be sensibly interpreted in relation to one of the
characters). This region either maintained the reference
established at the pronoun or produced another reference shift. The
results indicate that the first reference shift to the non focused
character (whether at the pronoun or at the verb) increased
processing time. A second shift incurred no processing cost. This
suggests that the processes involved in the resolution of pronouns
may be initiated before a pronoun is encountered. Focus may be
conceived of as a type of pragmatic cue which allows readers to
predict the most likely candidate for resolution in conjunction
with the degree of markedness within the referring expression
itself (e.g. Ariel, 1988, Givon, 1983). Furthermore the results
suggest that the resolution process is not completed when the
pronoun is fixated. The resolution of unambiguous pronouns
(experiment 1) could have been completed when the pronoun was
fixated. The results suggest that information after the pronoun
affected the resolution process. In the case of ambiguous pronouns
(experiment 2) although resolution could not take place at the
pronoun, verb reading times suggest readers still preferred to
assign the pronoun to the focused individual. Example of materials:
John bought the car from Anne and he/she decided to drive/walk
home through the countryside.
Ariel, M. (1988) "Referring And Accessibility." Journal of
Linguistics, 24 65-87
Arnold, J. (1997) " The Role Of Frequency In Reference
Processing." Paper presented at the Stanford Linguistics
Colloquium, 5th June 1997
Ehrlich, K. (1980) "Comprehension of Pronouns." Quarterly Journal
of Experimental Psychology, 32 247-255
Ehrlich, K. & Rayner, K. (1983) "Pronoun Assignment and
Semantic Integration During Reading: Eye Movements and Immediacy of
Processing." Journal of Verbal Learning and Verbal Behaviour, 22
75- 87
Garrod, S., Freudenthal, D. And Boyle, E. (1994) "The Role of
Different Types of Anaphor in The On-Line Resolution of Sentences
In A Discourse." Journal of Memory and Language, 33 39-68
|
