Examining non-task language effect on production-based language switching: Evidence from Tibetan-Chinese-English trilinguals

Aims and objectives/purpose/research questions: This study examined the non-task language effect in production-based language switching tasks by trilinguals using two of their languages. Non-task language refers to one of the languages of the trilinguals that is not used in a two-language switching task. Design/methodology/approach: Tibetan-Chinese-English participants (11th-grade high school students with a median age of 17) were recruited to perform picture naming tasks in three experiments, each with a different language pair of the trilinguals’ three languages (i.e., L1–L2, L2–L3, L1–L3). For each experiment, there were two groups of participants (30 for each group) with comparable proficiency in the two task languages but with significantly contrasted proficiency levels of the non-task language. Data and analysis: Each participant performed 216 trials and all together over 10,000 trials were used in the analysis of either the response times or the error rates in each experiment. Data of response times were analysed using linear mixed-effect models and data of error rates were analysed using logistic mixed models. Findings/conclusions: The response times and the error rates data across the three experiments revealed no main effect of non-task language proficiency, indicating a null effect of the non-task language. We proposed that a proactive language control mechanism might be used to overcome the interference from the non-task language. Originality: The non-task language effect in trilingual language switching is a rarely discussed topic. Significance/implications: This study is meaningful in evaluating the degree of generalizability of previous studies and can help gain insight into the underlying mechanisms of trilinguals’ language control.