Abstract：Abstract: A self-paced reading experiment was conducted to investigate whether there were differences in the processing sentences containing anaphoric and cataphoric pronouns in the second language. Reading time was measured on the sentences containing two clauses in which, respectively, located a proper name and a pronoun, either congruent or incongruent in gender. After reading the whole sentence, participants were required to answer a yes-no question concerning the reference of the pronoun. The results of reading time showed that cataphoric pronouns were resolved more rapidly than anaphoric pronouns in the first clause, but were resolved more slowly than anaphoric pronouns in the second clause. The results of co-reference judgment were in par with the reading time. We posit that the differences originated from the different lexical and discourse characteristics of pronouns and proper names. Furthermore, L2 readers adopted different processing strategies for different pairs of referents (i.e., name-pronoun vs. pronoun-name): bonding and evaluation for anaphoric pronouns and active search for cataphoric pronouns.
Key words: anaphoric pronoun, cataphoric pronoun, pronoun resolution, the Discourse Prominence Representation Model
