International Journal of Education

Lexical errors are a common issue in the oral production of second language learners, yet existing research has largely been based on classroom or test corpora, with insufficient attention paid to the characteristics of errors made by advanced learners in natural communicative contexts. This study draws on the first two episodes of the ninth season of the cultural talk show Informal Talks as its data source, identifying 31 instances of oral lexical errors. Based on the four-type error classification proposed by Lu Jianji (1994) and adapted to the actual corpus, an analytical framework comprising five categories— “overrepresentation, omission, overinclusion, misordering, and others” —was constructed to describe and analyze the error cases. It was found that overrepresentation errors accounted for the highest proportion, and the causes of these errors were closely related to negative transfer of the mother tongue, overgeneralization of target language rules, and the communicative pressure of impromptu performances in the variety show. This study provides insights for advanced-stage oral instruction and error research based on natural corpora.

Chinese language learning; Error analysis; Informal Talks; Spoken vocabulary