The Transparency Effect In The Processing Of Chinese Two-Word-Compounds

The Transparency Effect in the Processing of Chinese Two-Word-CompoundsSince the concept of morpheme was introduced into the field of psycholinguistics by Taft and Foster in 1975, the role of it in the processing of linguistic cognition had attracted the psycholinguists' attention, who deduced different theories through researches. The controversy focuses on what the basic unit of polymorphemic words is in mental lexicon representation, and the nature of it is the role of morpheme plays in the representation and processing of polymorphemic words. At present, most psycholinguists have agreed that morpheme plays a certain role in the representation and processing of polymorphemic words. However, they have not reached a consensus on to what degree, at which stage and at which level it works. There are three views about this question according to the researches of the alphabetical languages: the morpheme representation viewpoint characterized by Lexical Decomposing Model, which believes polymorphemic words represent in the form of decomposing morphemes; the whole words representation viewpoint characterized by Whole Words Representation and Access Model, which believes polymorphemic words represent in the form of whole words; the mixed representation viewpoint characterized by AAM model, which believes either morphemes or the whole words can possibly be the unit of representation and access.Two-word-compounds are dominant in Chinese language, so the processing of it attaches great importance to the psycholinguistic study. The theories put forward with the alphabetical languages as raw materials are undoubtedly valuable. However, Chinese polymorphemic words are different from those in the alphabetical languages. So the theories can not be followed blindly and the access and representation of the Chinese two-word-compounds must be explored with the characteristics of the Chinese language itself.Three different ideas are put forward by some scholars who borrow the theories used in alphabetical languages and carry on the experiments with Chinese as the materials: the viewpoint of decomposing storage by Zhang Biyin; Multi-Level ClusterRepresentation Model by Zhou Xiaolin and Maslen-Wilson; and Inter/Intra Connection Model by Peng Danling.The semantic relationship between morphemes and the whole words is seldom concerned in the early study of polymorphemic words, so the interpretation of it is influenced negatively. Semantic transparency is introduced in the field of lexical recognition to explain the phenomenon of semantic opaqueness. And it refers to the degree to which the meaning of the compounds can be inferred from that of its individual rnorphemes, and its operational definition is the correlation degree between the whole word and its morphemes. Researches show that semantic transparency can affect the recognition of the polymorphemic words, especially the opaque ones.Therefore, some certain questions still need solving in the study of semantic transparency: 1. Some materials are legal non-words and some are inflectional words and derivatives in the experiments in the alphabetical languages, while the storage of the Chinese polymorphemic words should be studied with the compounds as the experimental materials. 2. Though both transparency and word frequency can affect the storage of the polymorphemic words, how can they work? 3. Besides direct lexical decision tasks, lexical decision tasks on the condition of repetition priming and semantic priming are frequently used in lexical recognition. Different conclusions are drawn because of the different SOA (Stimulus Onset Asynchrony) between priming stimuli and target stimuli. So how can transparency affect lexical recognition if the SOA is different?In this dissertation, the mental representation of the polymorphemic words is discussed with Chinese two-word-compounds as the experimental materials by means of manipulating surface frequency with various SOA and employing the repetition priming technology and lexical decision tasks. Three questions are discuss...