A Developmental Study On The Effect Of Semantic Transparency Of Chinese Two-Character Words

In recent researches on lexical processing, the researchers focused on the processing and information access of muti-morphemic words. There were three theories about multi-morpheme words, such as the lexical decomposing theory, the whole word representation theory and the mixed representation theory. The controversies focused on these theories were how the lexical information and lexical structure were represented in mental lexicon, and what was the processing unit or access unit in identifying the multi-morpheme words, and how the detailed processing worked.In this study, the mental representation of multi-morpheme words was discussed with Chinese two-character words as the experimental materials, by means of controlling the frequency of the first morpheme, and changing the whole word frequency and semantic transparency. Eyelink II eye tracker was used to record participants' eye movements when they read the single sentences embedded with target words. Different eye movement index were analyzed, the aim was to investigate the role of semantic transparency and word frequency during the processing of Chinese two-character words, and their developmental characteristic in Chinese readers. Further, the representation and processing characteristic of Chinese two-character words were discussed.The participants were twenty four 8th graders, twenty three 11th graders and twenty five sophomores. The experimental design was 3 (grades: 8th graders, 11th graders, sophomore)×2(transparency: transparent, opaque)×2(frequency: high, low) design.The experimental results showed as follows:(1) The restraining effect of semantic transparency was found during the processing of Chinese two-character words. The lexical access for opaque words was earlier than for transparent words. (2) The accelerating effect of word frequency was found during the processing of Chinese two-character words. High word frequency could accelerate the processing of Chinese two-character words.(3) The results suggested that both the morpheme representation and whole-word representation for Chinese two-character words existed in the mental lexicon. The mixed representation theory was supported by the results.(4) The ability for lexical identifying was improved as grades. But the mixed representation model could account for the structure of the access representation for all the participants with different ages.