Inhibitory Control In Chinese-English/Chinese-Japanese Bilinguals' Language Production And Comprehension

Bilinguals are individuals who actively use more than one language. According to language proficiency of bilinguals, they are divided into balance and unbalance ones while according to the age of their learning language, bilinguals can be classified into early bilinguals and late bilinguals (Grosjean, 2012). However, no matter what type of bilinguals they are, bilinguals are able to switch between two languages successfully. Some researchers hold that there must be a cognitive mechanism such as inhibitory control responsible for this successful switch (Green, 1998). Inhibitory control here refers to a non-linguistic cognitive control network in which bilinguals are able to activate the intended language and inhibit the unintended language with the assistance of situations (De Bruin et al., 2014). Currently, a general agreement cannot be reached upon whether inhibitory control has a role to play in language selection(Costa & Santesteban, 2004; Gollan & Ferreia, 2009; Gullifer, Kroll, &Dussias, 2013), which leads to two research directions: supporting-inhibition and non-supporting inhibition. Researchers with two directions put forward a variety of models of bilingual language processing to explain whether or how inhibitory control works in language selection and also design different experiments to examine the role of inhibitory control.Nowadays, great progress has been made in the study of bilingual inhibitory control, nevertheless, there are still some weaknesses in the previous studies: 1) as regards research subjects, the studies concentrating on bilinguals whose L1 is Chinese are in scarcity while the majority of the previous studies take bilinguals whose L1 is English as subjects; 2) with regard to research content, the previous studies are mainly investigating one specific point (phonological or lexical) of bilingual language production and comprehension where inhibitory control may work, but as for semantic and discourse levels, little work has been done to explore whether inhibitory control works or not; 3) with respect to research methods, behavioral research with reaction time is majorly adopted in the previous studies of bilingual processing while ERP methods with a high temporal resolution are scarcely adopted. Still worse,research paradigms used in the previous studies are usually task switching and tasks used are picture naming; 4) regarding theoretical construction, the previous theoretical models are mostly able to explain the findings at a phonological or lexical level,but these models may not provide a good explanation for the findings at a semantic or discoursal level.Considering the aforementioned limitations of the previous relevant studies,based on Green(1998)'s inhibitory control model,and Dijkstra and Heuven (1998)'s bilingual interactive activation, etc, in the present thesis, a new revised model of bilingual language production and comprehension is proposed that inhibitory control could occur at phonological, lexical, semantic and discoursal levels. Adopting behavioral and ERP experiments, the present study employs four experiments with different tasks (picture naming tasks, verbal fluency tasks, semantic relatedness judgment tasks and speech tasks) to examine whether or not inhibitory control is able to occur at phonological, lexical, semantic and discoursal levels.The results of the experiments show:l) When switching into L1 or L2,asymmetrical switch cost in voice onset time of stops was found that L1 voice onset time suffered more as compared with that of L2, which was believed to be a strong index of inhibitory control that worked at a specific point. Moreover,a strong correlation was obtained between inhibition abilities and VOT cost, which meant that better inhibition abilities led to a less cost of voice onset time. Hence, it is concluded that inhibitory control is able to occur at a phonological level; 2) Asymmetrical switch cost was also found in bilinguals' lexical retrieval in a fluency task, which meant that inhibitory control worked at bilinguals' lexical retrieval when switching into L1 or L2.Inhibitory ability was also found to be correlated with switch cost at the lexical level.Thus, inhibitory control at the lexical level was also observed; 3) Homographs could activate both of their meanings in two languages and lexical meaning of non target language (Japanese) was persistently inhibited; ERP components N200 (an electro-physiological negative wave reaching peak in the time window of 200 ms)and N400 (an electro-physiological negative wave reaching peak in the time window of 400 ms) that are usually used to index the existence of inhibition were obtained and N200 was observed to be more negative in Left frontal, central and posterior scalp site while N400 was more negative in the central and posterior sites. Therefore,Involvement of inhibitory control was identified at the semantic level. 4) In speech task switching, asymmetrical CAF (complexity, accuracy and fluency) cost was attained that the participants' L1 speech performance suffered more in a switch position as compared with L2 speech performance. Moreover, the participants'inhibition ability could impact some aspects of speech performance and asymmetrical impact was also found. Hence, it was proved that inhibitory control could work at a discoursal level.Overall, the present four experiments demonstrated that inhibitory control occurred phonologically, lexically, semantically, and discoursely. And the new revised model of bilingual language production and comprehension was confirmed. The findings of the present study are able to make some contributions to the research of bilingual language production and comprehension as well as second language learning and teaching. Regarding contributions to bilingual production and comprehension,firstly, the present study helped extend research subjects to unbalance or late bilinguals and thus provided more evidence for the generalization of inhibitory control in bilingual language processing. Secondly, research content of inhibition was extended to semantic and discoursal levels, which was able to enrich research dimensions of Green (1998)'s inhibitory control. Thirdly, a methodological contribution was made that a variety of research paradigms and tasks were adopted and ERP method was also employed to investigate the role of inhibition at different points. Fourthly, a theoretical contribution was made in the present study that a new revised model of bilingual language production and comprehension was put forward to explain how inhibitory control worked at the phonological, lexical, semantic and discourse levels. With regard to contributions to second language learning and teaching, the first one is to help change our attitudes towards errors or mistakes made by learners in language switching. Secondly, the present study contributes to helping realize the importance of individual cognitive abilities such as inhibition ability and working memory in language learning and teaching. Thirdly, it helps change our attitude towards language learning as a tool into regarding language learning as a method to improve inhibition ability of language learners.