An Empirical Study Of Chinese EFL Learners’"Covert Nonnative-likeness":from The Perspective Of Domain Mapping Theory

"’Nornnative-likeness" can be defined as follows:foreign language (FL) learners produce FL expressions which bear the impress of first language (LI) and differ from those of target language. It includes "overt nonnative-likeness" and "covert nonnative-likeness". Recent studies mainly focus on the latter, which is touched upon at lexical level and validates the existence of it through empirical experiment. However, there exist two deficiencies in the previous studies:on the one hand, the previous theoretical explanation of the cognitive mechanism underlying "covert nonnative-likeness" is from the perspective of conceptualization mode and hence no further theoretical exploration of it from the perspective of FL learners’language competence is given; on the other hand, the conclusion that "covert nonnative-likeness" exists is drawn from two active tasks of FL production and no evidence supports the existence of "covert nonnative-likeness" in passive ones.The current research adopts the empirical study by way of reaction time and accuracy rate to explore the occurrence of "covert nonnative-likeness" in passive task of FL production among Chinese English as a Foreign Language (EFL) learners. Three research questions are put forward:1) Is there any preference for Chinese-&-English-shared domain mapping filler (SF) rather than English-specific domain mapping filler (EF) in passive task of FL production among Chinese EFL learners, namely "covert nonnative-likeness"?2) If there is, what are the features and differences of "covert nonnative-likeness" between passive task and active task of FL production?3) Is there any difference in processing between Chinese-versioned domain mapping filler (CF) and English-versioned domain mapping filler which is SF in the current thesis, namely, two languages—English and Chinese?Through a quantative study, a2×2factorial design is adopted in the current study, in which the two independent variables are the priming conditions and the filler types, the former referring to non-priming condition and priming condition and the latter referring to two pairs of filler types—SF and EF, CF and SF. And the two dependent variables are reaction time and accuracy rate. Totally, there are twenty-eight postgraduates majored in English Language and Literature from first, second and third year of Sichuan International Studies University (SISU) chosen to participate in the experiment. The stimuli, programmed by E-Prime2.0, are presented at random on the19inch wide screen. The subjects are required to carry out the sensibility judgment by pressing the button upon understanding the sentences with blanks, followed by fillers which are presented to be judged by subjects whether they are suitable for filling in the previous blanks. The behavioral data of reaction time and accuracy rate are put into the statistic software SPSS. The experiment is completed in the Key Lab of Cognitive Neuroscience&Foreign Language Learning in SISU.The research obtains the following three findings:1) Under non-priming condition, there is significant distinction between SF and EF in accuracy rate (1=5.07, p=0.000) but no significant distinction in reaction time (1=0.79, p=0.439).2) Under priming condition, there are significant distinctions between SF and EF in both reaction time (1=3.94, p=0.001) and accuracy rate (1=7.60, p=0.000).3) In allusion to the same experimental content, there are significant distinctions between CF and SF in both reaction time (1=10.823, p=0.000) and accuracy rate (r=-3.602, p=0.001).To sum up, the conclusions are as follows:1)"Covert nonnative-likeness" occurs in both active task and passive task of FL production;2) Compared to active task of FL production, the occurrence of "covert nonnative-likeness" in passive ones is more invisible than that in active ones and requires the pre-induction of L1conceptualization mode;3) The existence of significant distinctions between CF and SF shows that the occurrence of "covert nonnative-likeness" is closely related to FL conceptual competence being weak and L1conceptual competence being strong.