Getting to Know Ziji: Preliminary Findings from an ERP Study on Chinese Reflexive in L1 and L2
Journal
Handbook of Chinese Language Learning and Technology
Start Page
197-221
ISBN
9789819759293
9789819759309
Date Issued
2025
Author(s)
Abstract
Anaphoric expressions are central in human communications in that they help bind relevant identities to form a coherent discourse. Being able to accurately and reliably link these anaphors to their intended referents is therefore critical for efficacious language use. Unlike the majority of languages, the reflexive ziji in Mandarin can be interpreted as binding with entities outside of its governing category. This special feature is thus a challenge for L2 learners of Mandarin. To better understand the mechanisms involved in comprehending Mandarin reflexive ziji, the present study examined online Event-Related Potential (ERP) responses as well as offline behavioral judgments to ziji in local and long-distance reference biasing contexts from native speakers of Mandarin and L2 learners of Mandarin whose native languages do not allow long-distance referencing. Our results of offline behavioral reference selection showed comparable accuracy for both local (LOC) and long-distance (LD) references from L1 participants. However, there was a strong preference for local reference among L2 participants regardless of whether they were aware of the semantic implausibility of a local reference. These offline measures thus provide strong evidence to corroborate the influence of L1 transfer in the L2 learning of ziji. For the online ERP data, L1 participants showed a robust P600 grammaticality effect to ziji in the ungrammatical (UG) condition relative to the grammatical LD and LOC conditions. Critically, there was also a significant response difference between the two grammatical conditions, with more positive responses to LD ziji than the LOC ziji during the P600 time window. These data provide new insights into the processing mechanisms for ziji among native speakers, indicating extra processing costs for L1 speakers for linking ziji with long-distance than local referents despite similarly accurate reference selection in the offline reference selection task. For the L2 learners, we did not observe any systematic brain response differences at the group level. There was a numerical trend for an association between better verb knowledge and a larger P600 response difference between UG ziji and LOC ziji. In contrast, no systematic relation was found between verb knowledge and response difference between UG ziji and LD ziji, indicating that L1 influence for L2 learners on interpreting Mandarin reflexives persists despite increasing L2 proficiency. Overall, our findings advance our understanding of L2 learners’ processing of Mandarin reflexives and highlight the importance of explicitly teaching long-distance reflexives during teaching Mandarin as a second language.
Publisher
Springer Nature Singapore
Type
book part