English Pronoun Gender Processing In L2 Comprehension By Chinese Learners Of English

Based on previous findings, errors concerning gendered pronouns in English (e.g., he/she confusion) are recurrent among various levels of Chinese learners of English. Many investigations based on the second language (L2) speech production data (naturalistic data and elicited data) highlight Chinese learners’ problem with the acquisition of English pronoun gender agreement. So far very few investigations of L2 learners’ acquisition of pronoun gender agreement have been done based on L2 comprehension data. As some studies (Anton-Mendez,2010; Dong et al.,2014; etc.) explained, the recurrent pronoun gender errors in L2 speech production are probably due to the failure to automatically encode semantic gender information at the pre-linguistic conceptual level. So we are also interested in whether automatic processing of gender information exists in L2 comprehension, in particular, of Chinese learners of English.The present study reported two experiments to investigate the question of whether Chinese learners of English of two L2 proficiency groups could automatically process semantic gender information in their L2 comprehension. Experiment 1 used self-paced reading paradigm to test the automaticity of semantic gender processing at discourse level and Experiment 2 used semantic priming/lexical decision paradigm to test the automatic semantic gender processing at purely lexical level. More specifically, Experiment 1 investigated the gender mismatch effect with the two groups of participants:whether they would spend longer reading time for the critical pronoun (he/she) when the pronoun is mismatched with its antecedent in semantic gender than in the match condition. Experiment 2 investigated whether the automatic gender priming effect would be obtained with Chinese learners of English in a pronoun/non-pronoun lexical judgment task. The results of Experiment 1 obtained significant mismatch effects; that is, both the two groups of Chinese learners of English were automatically sensitive to the gender mismatch between the pronoun and its antecedent, suggesting that they could automatically process semantic gender in pronominal coreference processing. Results of Experiment 2 obtained the semantic gender priming effect, indicating that participants of the two English proficiencies could automatically activate and use the semantic gender information carried by gendered nouns and pronouns; this suggested that Chinese learners of English automatically processed the pronominal gender feature even when they comprehended individual pronouns in isolation without a discourse context.In sum, the present two experiments provided evidence for the automatic activation and use of semantic gender information in L2 comprehension of pronouns. Combined with previous L2 production data related to pronoun gender errors, the findings in our present study indicated that Chinese L2 learners have asymmetric performances on production and comprehension of the third-person singular pronouns in English. The probable interpretation may be that in English reading comprehension they have no problem in automatically activating the gender information encoded in linguistic forms and checking the pronominal gender agreement, but in English speech production they might have problem with automatically preparing the gender information necessary for English pronominal reference at the pre-linguistic conceptual level and this might be true even for learners of high English proficiency.