| |
Abstract:
Using a slow presentation rate (120 words/min) Anderson and
Holcomb (1996) found that gender agreement violations between
pronouns and their antecedents elicited a P600-like effect at
posterior electrode sites, and a negative-going effect at anterior
sites that continued into the epoch of the following 3 words. To
determine if these findings were due to a slow presentation rate
the current study recorded ERPs to words in sentences presented at
near normal reading rates (200 words/min). In each sentence, the
pronoun referred either to a single proper name (50%), or to one of
two proper names (50%). Further, the pronouns and their antecedents
were either of the same gender (50%) or of different genders (50%),
the latter resulting in grammatical violations. The current study
replicated the findings of Anderson and Holcomb, with violating
pronouns producing larger posterior P600s and larger anterior
negativities. This suggests that the results found by Anderson and
Holcomb at the pronoun itself were not due to unnatural strategies
adopted by participants. However, the negativity found for the
words following the gender-mismatching pronoun had a wider
distribution across the scalp suggesting the possibility that
faster rates do differentially affect the processing of subsequent
words. The implications of these findings for the theories of
pronoun processing will be discussed.
|
