A Study Of The Processing Of Compound Words By Chinese EFL Learners

The processing of morphologically complex words has long been a fascinating research subject in psycholinguistics since it can provide valuable insights into the structure and working mechanism of the mental lexicon.Previous studies on morphological processing have mostly focused on derived words or inflected words,suggesting the existence of morphological decomposition as well as holistic representation and processing depending on the interplay of a variety of factors,semantic transparency and frequency in particular.In contrast,compound processing has received relatively less research concern and furthermore,the processing of L2 compound words is a still less researched area.In addition,most of the Europe-or North America-based researchers have typically used balanced bilinguals or high-proficiency L2 learners as their subjects,who can hardly represent the overwhelming majority of Chinese EFL learners.The purpose of this dissertation then,is to investigate how Chinese EFL learners,mostly medium and low-proficiency learners,process compound words in both English and Chinese,with an emphasis on English processing.From a cross-language perspective,this dissertation provides an opportunity to examine the role of lexicality,or lexical status of the translated compound words in L1,in the processing of L2 compound words.Lexicality effect will bear witness to both compound decomposition and cross-language activation in processing.This dissertation also addresses a foundational debate regarding the role of semantic transparency in L2 compound processing.Semantic transparency effect will provide evidence for the choice of processing routes.In addition,the present study tests the hypothesis of cross-language priming asymmetry on compound words by examining whether L1/L2 is activated in the processing of L2/L1.Up to now,cross-language priming asymmetry has only been established for monomorphemic words and to some extent,this dissertation represents the first attempt in this regard with compound words.Finally,the time course of processing compound words is analyzed and discussed.Three online psychological experiments were designed and conducted to fulfill the research purposes.Using a lexical decision task,Experiment 1 demonstrates lexicality effect of English compound words regardless of semantic transparency.The target words in Experiment 1 were transparent and opaque English compound words whose translated compounds are either real words or nonwords in Chinese,and the results showed significantly higher accuracy rates for the real-word condition.However,Experiment 1 failed to identify the semantic transparency effect which has been found in many previous studies of L1 compound processing.This may be explained by the view that morphological decomposition in L2 compound processing takes place on the lexical form level but the semantic representation is not activated.Compound decomposition and cross-language activation got further evidence in Experiment 2,but not in Experiment 3,which are both cross-language constituent priming experiments.Experiment 2 presents English compound word as the target and the Chinese translation of the second constituent of the compound as the prime,whereas Experiment 3 presents similar prime-target pairs in the reverse direction,with English translation of the second constituent priming the Chinese compound word.In Experiment 2 there was a significant L1-L2 constituent priming effect in terms of both accuracy and response speed,but similar effects were not discovered in the L2-L1 constituent priming of Experiment 3.The results of Experiments 2 and 3 point to a clear asymmetry in the cross-language priming between L1-L2 and L2-L1,supporting the prediction that in the bilingual lexicon,the connection between L1 and the conceptual system is stronger than that between L2 and the conceptual system.Experiments 2 and 3 also feature the tricky variables of semantic transparency and prime durations.The compound targets are divided into transparent and opaque groups,and the priming tasks were conducted under three different prime durations—100 ms,250 ms,and 400 ms.There was a significant transparency effect in response time in the L1-L2 constituent priming of Experiment 2.Considering the participants' different performances in recognizing transparent and opaque compounds,the dual-route model is more capable of accounting for the data presented here.Last but not least,Experiment 2 demonstrates stronger priming in terms of response speed under longer prime durations,and also shows that the activation of semantic representation seems to occur at the later stage of processing,since transparency effect was the most robust under the prime duration of 400 ms,but was negligible under shorter prime durations.Finally,regarding the processing of L1 compound words,Experiment 3 failed to yield a significant semantic transparency effect,which implies a possibility that the whole-word route is preferred for Chinese compounds.However,caution needs to be exercised in drawing conclusions in L1 compound processing based on the present study since the experiment design was made chiefly to investigate Chinese EFL learners' processing of English compound words.The findings together argue that Chinese has a role to play in the processing of English compound words by Chinese EFL learners,and that compound decomposition is an inevitable stage in processing though the whole-word route is more likely to take over in the case of opaque compounds.Compound words are represented in the bilingual lexicon in a way different from the way they are represented in the monolingual lexicon mainly because of the intervention of L1 between L2 and the semantic representation.The findings of the current study hold implications not only for the theoretical models of L2 compound processing,but also for English vocabulary acquisition and instruction in Chinese EFL learners.However,with its limitations in the choice of subjects and test items,as well as in the experiment design,this dissertation only represents a small step toward investigating the processing of L2 morphologically complex words and understanding the mechanism of the bilingual lexicon.