A Psycholinguistic Investigation Of The Phonological Encoding In The Second Language Speech Production

Before uttering words,people need to encode what they want to express,and this whole process from getting concepts to articulation is called speech production.Specifically,after the retrieval of the syntax-assigned lemma from the mental lexicon,speakers then need to encode the phonological information for their utterances.This thesis aims to investigate the phonological encoding for Mandarin Chinese-English bilinguals,based on WEAVER++ model put forwarded by Levelt,Roelofs and Meyer(1999).Empirical studies have mainly focused on the issues of planning unit and its encoding manner at phonological encoding.However,there have been more studies that employed alphabetic languages as experimental materials,rather than Mandarin Chinese,a logographic language with writing system composing of radicals instead of phonemes.Thus,it would cause differences cross distinct languages.The present research focuses on the proximate unit during bilingual phonological encoding.Experimental evidence has suggested that the proximate unit is not restricted by the target language during bilingual speech production.For instance,for bilinguals with both alphabetic languages background such as Dutch and English,the first unit to be encoded is segment,the same form for both their L1 or L2,though.On the contrary,for bilinguals with cross language family,such as Mandarin Chinese and English,Japanese and English or Dutch and Cantonese,the proximate unit is independent of the languages they were asked to perform in experiment.Besides,the bilingual proficiency would also influence the naming process.Taken together,the proximate unit cross language family still needed to be discussed.The current issue is to clarify the proximate unit at bilingual phonological encoding for Mandarin Chinese-English bilinguals.To further examine the discrepancy which may caused by the bilingual proficiency,we divided participants into two groups in terms of their performance.In two picture-word interference experiments,highly and less proficient participants were tested respectively.A 4(relatedness between target and distractor: syllable overlapping;onset two phonemes overlapping;onset phoneme overlapping;unrelated)by 3(SOAs: 0ms,75 ms and150ms)design was conducted.In experiment,the English distractor word was superimposed on line drawing target picture.Participants were supposed to name the target picture in English while ignoring the distractor as soon and accurate as possible.The naming response was recorded simultaneously by the experimental software,Psychopy,and statistically analyzed by R.Behavioral data indicates that(1)the main phonemic effect is significant.(2)The segmental overlap effect was observed.(3)Highly proficient participants named target faster than less proficient bilinguals.The results indicate that(1)bilinguals with both highly and less proficiency would employ segment as the proximate unit to encode their L2,demonstrating a segmental priming effect,which is the same with L1 speakers.(2)Both proficiency bilinguals encode English phonological code rightward incrementally,which is not only compatible with previous study by Verdonschot(2013)about highly proficient bilinguals,but also supply a gap of less proficient data.(3)Although the proficiency level does not influence the selection of proximate unit,the difference still exist on behalf of the response time,revealing the proficiency advantages in naming pictures.In summary,the current study specifies the nature of proximate unit at phonological encoding for Mandarin Chinese-English bilinguals,with picture-word interference paradigm.Crucially,we add the less proficient bilingual group as comparison with previous studies.The results support the assumption that bilinguals phonologically encode L2 with the same proximate unit as the target language,which will provide significant insights for speech production and second language teaching.