Lexical Linking Patterns In Chinese Efl Learners’ Mental Lexicon

Lexical knowledge is a meaningful indicator of reading skills, and plays an essential role in the second language acquisition. Research on lexical knowledge basically focuses on the acquisition of the deep lexical knowledge, and the research in this field has been changed radically for the recent fifty years, whose focus has moved from the increase of vocabulary size to the organization features of mental lexical network, and from multidimensional decomposition of a word to the exploration of lexical representation and processing with research methods employed mainly in the other disciplines. Of all the research topics on mental lexicon, the relationship between the two languages has been a heated one. Since the access and the linking patterns of the lexical representations cannot be directly visualized, the studies based on RT gradually become the one of the best choices for both psychologists and linguists. Therefore, the present study investigates the linking patterns of Chinese EFL learners’ mental lexicon by means of the translation-recognition task with two RT experiments, from the perspectives of lexical variable effect, language effect and proficiency effect.Two RT experiments recruited84and68undergraduates in Jiangsu province as participants respectively. These participants were of two English proficiency levels. Experiment1involved49participants of high level and35of low level, while Experiment2included33and35participants, respectively. The experimental materials were made up of576English-Chinese word pairs, with384as test word pairs and192as fillers. These word pairs were ordered using Latin-Square design into four lists, with144word pairs per list. Each list contained word pairs of translation equivalent, form distracter, and meaning distracter as critical conditions, and unrelated word pairs as a baseline. Word pairs of each condition were further classified into three groups (high-frequency concrete, low-frequency concrete, and high-frequency abstract) according to the two lexical variables, frequency (high and low) and concreteness (concrete and abstract). The participants were asked to decide if the target words were the translation equivalents of the prior words. The software E-prime was used for presenting stimuli and collecting of behavioral data, accuracy and response time (RT). Repeated measure ANOVA was performed with SPSS for analyzing the results obtained.For the lexical variable effects on bilingual lexical links, it was found that both L2-L1and L1-L2links varied with word frequency and concreteness, that is, there were frequency effects and concreteness effects on these links. Chinese EFL learners generally responded to high-frequency and concrete words faster and more accurately than to low-frequency and abstract words. It was also found that frequency effect and concreteness effect were greater in L1-L2links than in the L2-L1links; compared with high-frequency and concrete words, low-frequency and abstract words were to a larger extent influenced by form distracters in L2-L1translation recognition, and were more likely to be influenced by meaning distracters in L1-L2translation recognition.In terms of language effects, it was obtained that L2-L1and L1-L2links varied with the target language. Additionally, the basic feature of lexical links of Chinese EFL learners was that orthography dominated L2-L1links, and semantics dominated L1-L2links, although L2words were represented in a way in which orthographic and semantic links co-existed in two directions.For the proficiency effects on bilingual lexical links, it was observed that there were proficiency effects on L2-L1and L1-L2links in terms of both proficiency effects on orthographic/semantic levels and the linking strength.There were robust proficiency effects on orthographic/semantic levels, namely,(1) for low-proficiency L2learners, semantic representation emerged at a very early stage, but the orthographic links were stronger; for high-proficiency learners, semantic links were stronger, but orthographic links did not disappear with the increase of proficiency. These findings were unexplainable within the existing frameworks which were mainly based on alphabetic bilinguals, because these models underestimated the early emergence of L2semantic representations and late reliance on the L1orthography.(2) Organization of bilingual lexical links was a dynamic process, where lexical and conceptual representations interacted. For the low-proficiency learners, orthographic links were stronger than semantic links. For the high-proficiency learners, the pattern was the opposite, that is, they had a much stronger semantic link as compared with an orthographic link. More importantly, the dynamic change was quantitative, and there was no qualitative shift from lexical reliance to conceptual mediation, as described in RHM modeled with alphabetic bilinguals.(3) Such mainstream models as RHM and BIA, modeled mainly with alphabetic bilinguals, were unable to account for some features of lexical links of Chinese EFL learners.There were also robust proficiency effects on L2-L1and L1-L2linking strength, that is,(1) High-proficiency learners showed stronger links than the low-proficiency group in both directions at both orthographic and semantic levels.(2) Both groups showed asymmetry in the linking strength, which meant their L2-L1links were stronger than the L1-L2links. This was contradictory to the prediction of RHM, which expected a trend of symmetry with the improvement of proficiency. Despite the proficiency effects on form-meaning links and the linking strength, there was no significant proficiency effects on lexical variables. Specifically, there was no significant difference in the magnitude of frequency effects between the two groups in either directions, which was contradictory to the Weaker Links Hypothesis. There was no significant difference in the magnitude of concreteness effects between the two groups in either directions, which was consistent with Distributed Conceptual Feature Model (DCFM).The findings of the present study shared theoretical, methodological and pedagogical significance. Theoretically, a Bilingual Form/Meaning Linking (BFML) model was established to accommodate the findings on the lexical links for the Chinese English learners. The present study also highlighted orthographic-semantic (or form-meaning) links, which enriched the dimension of deep lexical knowledge. Methodologically, the strictly constrained word pairs used in the present study were of referential significance, for their database-wide sampling, strict parameter settings and high operationality. Pedagogically, it was indicated that L2word acquisition should focus on establishing and refining L2semantic network; and compiling dictionaries and textbook word lists should facilitate the efficient organization of bilingual lexical links.