Cross-language Activation In Compound Nouns Recognition: An ERP Study

Research on the psycholinguistics of bilingualism is mainly concerned with two questions: How the two lexical systems are represented in a bilingual speaker’s mind and how the two lexical systems interact in language processing.Underlying these questions,there is a more basic issue: whether bilinguals are able to selectively activate one of their languages,i.e.,the language selectivity debate.Studies of bilingualism have often used a mix of first-(L1)and second-language(L2)words,thereby creating an artificial “duallanguage” context to test whether cross-language activation occurs or not.However,relatively less attention has been paid to noncognate languages and compound words in either L1 or L2.What’s more,the neural signatures of the compound words recognition in bilinguals remain unclear.Using event-related potentials,the present study attempted to investigate the crosslanguage activation with compound nouns in Chinese-English bilinguals.A one-factor two-level(translated words,e.g.,“黑板”–“blackboard”vs.unrelated words,“毛衣”–“database”)experiment was designed.Each target word pair consisted of a two-character word in Chinese and a compound word in English,respectively.Participants were selected with intermediate L2 proficiency and were required to decide whether the target word pairs were translation equivalents or not as accurately and fast as possible during the experiment.Comparisons were drawn between translated words and unrelated words.Both behavioral performances,including reaction times(RTs)and accuracy rates(ACCs),and electrophysiological data were recorded.The following findings have been revealed based on the analyses of the behavioral and ERP data:(1)Although the ACCs produced a null effect for translation,RTs revealed a facilitation effect,for participants responded much faster to translated words than unrelated words.(2)Inconsistent with the predicted N400 effects,ERP results showed that unrelated word pairs elicited larger left anterior negativity(LAN)from 300 to 450 ms than translated word pairs.A tentative explanation is that the LAN observed is more likely to mark the morphosyntactic mismatch in the experiment,indicating lexical anomalies in compound words recognition in two languages.An ERP collapse was found between the two conditions from 500 to 800 ms,termed the late positive component(LPC).It suggested the process of late conflict resolution and a recheck of the decision to reject unrelated distractors.Taken together,the present study provides an account that both languages of a bilingual were automatically activated in compound words recognition.In accordance with the assumption of the BIA+ model and the Multilink,the present study demonstrates electrophysiological evidence for non-selective lexical access,and extends the mechanism of cross-language activation to the domain of compound words recognition.This research contributes to bilingual lexical access and compound words recognition from the following perspectives:(1)Empirically,the present study offered empirical evidence for the pattern of cross-language activation in the field of compound words.(2)Theoretically,by asking how cross-language activation functions in compound words recognition,the present study enriches the understanding of the interaction between L1 and L2 in bilinguals then validates the BIA+ model and the Multilink.(3)Methodologically,the use of advanced ERP technique,prior to the traditional behavioral method,will reveal inner cognitive activities precisely as the processing unfolds,which is more likely to provide neat and faithful data for the bilingual language processing.