The Effect Of L2 Proficiency On L1-l2 And L2-L3 Switch Costs Of Tibetan-Chinese-English Trilinguals

As a social communication phenomenon,code-switching refers to a bilingual switching from one language to another.People will use longer time when switching between two languages than using only one language,a phenomenon that is called as switch cost.Switch cost could be symmetric or asymmetric: while the balanced bilinguals show similar switch costs when switching between two languages,the unbalanced bilinguals will spend longer time when switching to the less proficient language.This means that longer response time and higher error rate would be made by the less proficient language users in different language switching tasks due to different language proficiency.In a word,switch cost is closely associated with language proficiency.The previous study of Zhang & Cui(2008)shows that language switch cost is closely related to language proficiency,age of acquisition,usage of languages,experimental paradigm.It can be clearly seen from previous studies(Andra & Iring,2007;Xavier & Jean-Marc,2013;Daniel,2016;Michela & Kees,2017)both at home and abroad that many studies have focused only on languages of the Indo-European families that are typologically similar,with fewer studies dealing with typologically different languages.Specifically speaking,as typologically different languages have different orthographic characteristics and different language processing features,different retrieval time to languages with different proficiency would be made inlanguage switching tasks.What’s more,many previous studies on switch cost focused on language production while fewer studies examined switch cost in language comprehension.As language production and comprehension will involve different cognitive processing mechanism,switch cost might also be different.In addition,most switch cost studies were on bilinguals rather than trilinguals.Among those studies on trilinguals,few studies considered L2 proficiency effect on switch cost.People from ethnic minority areas in China have acquired their mother tongue(L1)and Chinese(L2)before they begin to learn English(L3)in Grade 3 or Grade 7.Meanwhile,they have different L2 proficiency due to language practice or language environment.Therefore,this study intends to explore the effect of L2 proficiency to Tibetan-Chinese and Chinese-English language switching among Tibetan middle school trilingual students in Gansu province.Based on the previous study(Zhang & Cui,2008),on the one hand,this study considers the effect of different L2 proficiency;on the other hand,participants in this study are middle school students.These two aspects are valuable to further verify switching mechanism and characteristics.In this study,Tibetan-Chinese-English trilingual middle school students with different L2 proficiency were involved in the behavioral experiments in which they were asked to do lexical recognition tasks in either Tibetan-Chinese or Chinese-English switching mode.The research questions are:Question 1: Does switch cost occur in the lexical recognition when Tibetan trilingual students with different L2 proficiency switch between Tibetan and Chinese? Ifit does,is the cost symmetrical or asymmetrical?Question 2: Does switch cost occur in the lexical recognition when Tibetan trilingual students with different proficiency switch between Chinese and English? If it does,is the cost symmetrical or asymmetrical?This present study adopted 2*3 mixed experiment design.The independent variables were languages and tasks,and the dependent variables were response time and error rate.Among the 60 participants,30 were Tibetan Mediated Curriculum(TMC)based students with higher Tibetan proficiency than Chinese and the other 30 were Chinese Mediated Curriculum(CMC)based students with higher Chinese proficiency than Tibetan.These two groups of participants were asked to do the lexical recognition by judging whether the items presented on the screen were real words or not as quickly as possible.The experiments were programmed and conducted by DMDX software,which recorded participants’ response times and error rates in the recognition tasks.The data were analyzed by SPSS-20.The research findings are as follows:First,as for Tibetan-Chinese switching,there was no switch cost to both Tibetan and Chinese for TMC students and CMC students.In a word,switch cost did not occur to both languages with balanced proficiency.Second,as for Chinese-English switching,switch cost occurred to both Chinese and English for both TMC students and CMC students,and the switch cost was asymmetrical for less proficient language in switching tasks,which also means that longer retrievaltime would be made by the less proficient language.In other words,longer response times and error rates were observed when switching to the less proficient language,and the more proficient language needed less retrieval time.In a word,the more unbalanced the languages’ proficiency,the much more switch cost there would be;and less switch cost occurs to languages with balanced proficiency.In conclusion,switch cost did not occur to language pairs with relatively balanced language proficiency but occurred to language pairs with unbalanced language proficiency.Meanwhile,switch cost to less proficient language was larger than to proficient language and this switch cost was asymmetrical in language comprehension.