Cross-Language Priming Of Word Meaning In The Process Of Second Language Sentence Comprehension

Over the past decade, much research has been aimed at investigating cross-language priming effects with bilinguals. This thesis mainly investigates the benefit to English listening comprehension of priming word meanings with brief visual exposure to L1translation equivalents and tests whether cross-language effects can be utilized to reduce the inherent working memory (WM) load and facilitate meaning retrieval during L2listening comprehension by L2beginners of China. The study is relatively significant in aspect of dealing with the bottleneck of meaning retrieval in L2speech processing.Forty participants at Grade9in middle school take part in this experiment to evaluate the validity of the sentences only heard via earphones of the computer and without presented on the screen. Of the selected sentences, half of both valid and invalid sentences are synchronized visual exposure for300ms of L1translation equivalents, of which, one third of the primed sentences are preceded by an early single word, one third followed by a late single word and another with both. Half of both valid and invalid statements are unprimed, but blank is exposed in lieu of prime words for early and late words in the same proportions as for the primed sentences for300ms. RT and ERs are measured by DMDX3.2.6.1and analyzed by SPSS16.0. Results are submitted to One-way ANOVA analysis and correlation analysis, and major findings are revealed below.1) Whether cross-language priming can facilitate meaning retrieval during L2listening comprehension counts on the location of priming.2) Early priming that shows robust priming effects is outstandingly superior to late priming and remarkable difference exists. However, statistical difference isn’t found between the late priming and no priming, indicating no priming effects.3) Two-word priming doesn’t show an advantage over single-word priming and to some extent, has even interfered in evaluation leading to increased ERs and RT.4) Difference between valid and invalid statements in ERs and RT has reached the statistical level and correlation has been found in RT. There is no significant difference between male and female and no correlation is found in RT.