Language Presentationand Language Switchingof Profieient Mandarin-Cantonese And Cantonese-Mandarin Bidialectals: Evidence From ERP Studies

The first chapter IntroductionBackground and purposeBilinguals refer toindividuals who can speak or use two languages.Traditional bilingual studies primarily focused on bilinguals with two languages from different systems,such as English and French or Chinese and English.But in addition to such type of bilinguals,there is another special type of bilinguals,that is bidialectals who can speak two dialects.Dialect is a kind of variation formof a language,such as Cantonese,a dialect in Chinese.Generally,different types of dialects in a language share a same writing system but differ in phonology systems.Thus as the special characteristic of dialects,an interesting question arose as to how bidialectal speakers organize their two types of dialects in brain.Previous research focused on two issues of bilingual brain mechanism: the bilingual language representation and control mechanism.And the similarity of two dialects was higher than two languages of bilingual.Therefore,the conclusion of previous bilingual related research might not adapt to two dialects.Whether the two dialects shared the same brain regions and brain mechanisms in language forms and semantic?and whetherthe two dialectswas same representationor independent representationin the brain?Mandarin-Cantonese is one of the typical phenomenonof dialect.They werewidely usedin Guangdong,Guangxi,Hong Kong,Macao and southeast of Asia.The aphasia patients among these peoplewere rather major,butthe brain mechanism researcheswere less for both proficient bilingual and proficient Mandarin-Cantonese aphasia.However,these patients with aphasia needed for rehabilitation therapy,and thepresent aphasia recovery pattern was still controversial.It was still not clear how to choose a dialect for effectiverehabilitation treatment.So,if we wanted to study the brain mechanismof Mandarin-Cantonese aphasia patients,we must understand premorbid brain mechanisms with them.But we could not get it.Thus,we must discussthe brain mechanisms of both normal Mandarin and Cantonese.Moreover,most of present studies recruited Cantonese-Mandarindialects(the first language was Cantonese,and the second language was Mandarin),butits were less for Mandarin-Cantonese dialects(the first language was Mandarin,the second language was Cantonese),especially for proficient Mandarin-Cantonese dialects.Therefore,this study recruited proficient Mandarin-Cantonese and Cantonese-Mandarinbidialects to explore the language/dialect representation and language/dialect switching mechanisms of two dialects by event-related potentials(ERP)technique.This could not only make up for the inadequacy of bidialect brain mechanism,but also provide the evidence for the selection of rehabilitation training language with bidialectsaphasia.It would be an important theoretical and practical significance.The second chapter Language Presentationof Profieient Mandarin-Cantonese and Cantonese-Mandarin bidialectals:Evidence from ERP StudiesObjective The present study adopted a repetition prgiming paradigm to investigate the bidialectal(bilingual)representation of speakers with different native dialects by event-related potential(ERP)technique.Proficient Mandarin-Cantonese and Cantonese-Mandarin bidialectals participated in the study.Methods Thirty undergraduate students took part in the experiment,including 15 Mandarin-Cantonese bidialectals and 15 Cantonese-Mandarin bidialectals The two groups of participants(Mandarin-Cantonese and Cantonese-Mandarin bidialectals)have no significant difference in Mandarin and Cantonese listening,speaking,reading and writingaccording to their self-evaluations based on a seven-point self-rating scale.However,their use frequency of Mandarin and Cantonese were significantly different.The self-evaluation of Cantonese use frequency for Mandarin-Cantonese bidialectals was higher than that of Mandarin use frequency,whereas for Cantonese-Mandarin bidialectals,the self-evaluation of Cantonese use frequency was lower than that of Mandarin use frequency.The experimental materials consisted of 120 Mandarindisyllabic words,120 Cantonese disyllabic words and 120 filler words.Specifically,both Mandarin and Cantonese disyllabic words included 60 biological words and 60 non-biological words.The filler words included 60 Mandarindisyllabic words and 60 Cantonese disyllabic words.They were both written in simplified characters,and were matched in stroke numberand use frequency.We adopted an unmaskedrepetition priming paradigm with four types of repetition priming conditions in the experiment,including 1)Mandarin(prime)-Mandarin(target)(MM);2)Cantonese(prime)-Cantonese(target)(CC);3)Mandarin(prime)-Cantonese(target)(MM)and4)Cantonese(prime)-Mandarin(target)(CM).Each condition contained 30 biological word pairs and 30 non-biological word pairs.In addition to priming conditions,there were two types of filler conditions in the experiment.They were Mandarin filler condition(M)and Cantonese filler condition(C)respectively,and each has 60 words.The stimuli were presented by E-Prime 2.0.The EEGwas recorded by a 64-channel(Ag–Ag Cl)Brain Products(BP).In the data analysis,we first calculated the reaction time(RT),accuracy(ACC),P200 and N400 of repetition priming conditions and filler conditions,and then compared the RT and ACC between repetition priming conditions and their corresponding filler conditions(based on the dialect type of target word,e.g.,MM vs.M)by t-tests to confirm that the repetition priming conditions did elicit priming effects.If the priming effects did exist,we would further calculate the repetition priming effects by subtracting the RT,ACCand/or P200,N400 of repetition priming conditions from that of their corresponding filler conditions.At last,the repetition priming effects would beanalyzed by 2(native language: Cantonese/Mandarin)×4(priming condition: MM/MC/CC/CM)mixed repeated ANOVA with native language as a between-subject factor and repetition type as a within-subject factor.Results In the behavioral results,the t-tests of RT between repetition priming conditions and filler conditions showed significant difference between the repetition priming conditions and their corresponding filler conditions respectively.In addition,no difference between filler conditions for both groups of participants was found.The resultsconfirmed that the four types of repetition priming conditions did elicit repetition priming effects of RT and ACC.The further two-way mixed measure ANOVA of repetition priming effects in RT and ACC showed that the main effect of priming condition was significant by participants.The post hoc test showed that the priming effectsof RT in MM was significantly larger than that in CMcondition,and the priming effects in CC condition was significantly largerthan that in MC conditionand CM condition(ps<0.05).The priming effects of ACC in MM condition was significantly larger thanthat in CC and CMcondition(ps<0.05),the priming effects in MCcondition was significantly larger than that in CCcondition and CM condition(ps<0.05).There were no significant differences between the priming effects of other priming conditions.The interaction between native language and priming condition was not significant(ps>0.05).In the ERPresults,the t-tests of P200 and N400 between repetition priming conditions and filler conditions suggested that the four types of repetition priming conditions did elicit repetition priming effects of P200 and N400.The further ANOVA results of P200 and N400 repetition priming effects revealed a significant main effect of priming condition.The repetition priming effectsof P200 in MMwas significantly larger than that in MC and CM,and the repetition priming effects in CC was significantly larger than that in MCand CM.The repetition priming effects of N400 in MM was significantly larger than that in MCand CM,and the repetition priming effects of N400 in CC was significantly larger than that in MC and CM(ps<0.05).There were no significant differences between the repetition priming effects of other priming conditions(ps>0.05).The interaction between native language and priming condition was significant.Simple effect test of P200 revealed that for native Mandarin,the repetition priming effects in MM was significantly larger than that in CC and CM,and MC was significantly larger than CM(ps<0.05).However,for native Cantonese,the repetition priming effects in MM was significantly smaller than that in CC,and MC was significantly smaller than CM and CC(ps<0.05).Simple effect test of N400 revealed thatfor native Mandarin,the N400 repetition priming effects in MM was significantly larger than that inthree other priming conditions(ps<0.05),and that in CC was significantly larger than that in CMand MC(ps<0.05).No significant difference was found between the repetition priming effects of other priming conditions(ps>0.05).While for native Cantonese,the repetition priming effects in MM was significantly larger than that in MC and CMconditions(ps<0.05),while the repetition priming effects in CC was significantly larger than that in MC,CMand MM(ps<0.05).There were no significant differences between the repetition priming effects in other priming conditions(ps> 0.05).Conclusion The results clearly showed an asymmetric pattern for the representation of first and second dialect,which was consistent with the asymmetric hypothesis in Revised Hierarchical Model.Moreover,the results also revealed a significant role of native language(dialect)on bidialectal representation,though the use frequency of native language(dialect)was lower than that of second language(dialect).The third chapter Language Switchingof Profieient Mandarin-Cantonese and Cantonese-Mandarin bidialectals:Evidence from ERP StudiesObjective On the basis of the first study,this study also recruited proficient Mandarin-Cantonese and Cantonese-Mandarinbidialects to explore the language/dialect switching neural mechanism of two dialects with a delayed picture naming task by event-related potentials(ERP)technique.Methods Twenty-eight undergraduate students took part in the experiment,including 14 Mandarin-Cantonese bidialectals and 14 Cantonese-Mandarin bidialectals.It was the same with the first study that the two groups of participants(Mandarin-Cantonese and Cantonese-Mandarin bidialectals)have no significant difference in Mandarin and Cantonese listening,speaking,reading and writing,and their use frequency of Mandarin and Cantonese were significantly different.The experimental materials consisted of 60 black-and-white line drawings of concrete objects with a size of 15 cm×15 cm,each target picture was named by a two-character word.There was no significant difference inword frequency.10 additionalpictures were used in the practice session.No cognates werepresent for both practice and experimental stimuli.The stimuli were presented by E-Prime 2.0.The EEGwas recorded by a 64-channel(Ag–Ag Cl)Brain Products(BP).In the data analysis,the RT,ACCand/or P200,N400 would beanalyzed by 2(native language: Cantonese/Mandarin)× 4(switching condition: Mandarin-Mandarin<MM>/Mandarin-Cantonese<MC>/Cantonese-Cantonese<CC>/Cant onese-Mandarin<CM>)mixed repeated ANOVA with native language as a between-subject factor and switching type as a within-subject factor.Results In the behavioral results,the accuracy of naming pictures were 98.86%(SD = 2.46%),99.08%(SD = 1.78%),1.78%(SD = 3.53%)and 3.53%(SD = 3.76%)respectively.The two-way mixed measure ANOVA of accuracy wasn't shown significant difference between the native languages or switching conditions(ps>0.05).The reaction time naming pictures were1050.3ms(SD=401.79),1028.9ms(SD=416.11),1119.6ms(SD=402.05),1086.7ms(SD=406.23)respectively.The statistical result showed significant difference between the switching conditions.The results of post hoc test showed the reaction time was faster in non-switching conditions(MM/CC)than switching conditions(MC/CM)(ps<0.01),and the reaction time of MC was significantly longer than that in CM condition(p<0.05),but there was no significant difference between the reaction time of CC and MM conditions(p>0.1).In the ERP results,we rendered the clues presenting for zero time to analyze the mean amplitudes of P200(150 ~ 250 ms)and N400(300 ~ 500 ms).The statistical results of P200 and N400 weren't both shown significant main effect of switching conditions(ps>0.2).However,in the pictures presenting,the two-way mixed measure ANOVAof P200 and N400 were both shown significant main effect of switching conditions,and the results of post hoc test showed the mean amplitude of P200 and N400 were larger in switching conditions(CM/MC)than non-switching conditions(CC/MM)(ps<0.05).P200:For native Cantonese,the further simple effect test revealed that the mean amplitude of P200 in MC was significantly larger than that in CM(p<0.01).While for native Mandarin,it was opposite(MC<CM)(p<0.05).N400: For native Cantonese,the further simple effect test revealed that the mean amplitude of N400 in CM was significantly larger than that in MC(p<0.05).While for native Mandarin,it was opposite(CM<MC)(p<0.05),and there was also significant difference between CC and MM in non-switching conditions(CC>MM).Conclusion In the process of language switching,proficient bidialects existed language switching costs effect.In addition,the ERP results of two groups were both found asymmetric effect of language switching.Among them,the analysis of P2 showed that when the target language was native dialect,subjects needed more cognitive efforts to extract suppressed information of native dialect,which supportedlanguage-unspecific selection theory to a certain extent.While for the analysis of N400,the result indicated that native dialect lexical semantic information processing had certain advantages in the process of language switching,even if the subjects with use frequency of L2 was higher than native dialect.