Relationship Of Prototypicality And Language Transfer In SLA: From The Perspective Of Passive Voice

Language transfer has always been a controversial topic in second language acquisition (SLA) (Ellis, 1994; Odlin, 2001). In the past decades, its role in SLA has been reevaluated several times with the changes of the views toward it. After the behaviorist's exaggerating and the mentalist's rejecting of its role, it became one focus in SLA study again with the development of cognitive science (Richards, 2001). A major problem in language transfer research is to predict under what conditions language transfer may occur (Cai, 1998). In Kellerman's words (1979), what is transferable? In order to answer this question, Kellerman (1977, 1979) proposed that prototypicality can be used to predict the occurrence of language transfer. In Kellerman's framework, some linguistic structures perceived by learners to be language-neutral are potentially transferable and some language-specific structures are potentially non-transferable. However, because Kellerman's studies mainly focused on the language transfer between languages from the same language family, more evidence is needed as for the situation in which native and target languages involved are typologically different. Though there have been some researches on prototypicality and language transfer at home, most of them are just literature reviews while the empirical studies are quite rare.On the basis of Kellerman's theory, from the perspective of language transfer on the grammatical level, three research questions are addressed in this thesis in order to prove the universal predictability of prototypicality in language transfer and to specify the relations between language transfer and prototypicality:(1) Can prototypicality still be used to predict language transfer from Chinese to English, two typologically different languages?(2) What are the manifestations of language transfer in Chinese beginners' production of the passive in English?(3) What are the specific relations between prototypicality and language transfer?The experiment is carried out among 80 Chinese of English in their first year in the senior high school. They are required to finish three tasks: two judgment tests and a translation test. The former is to test the transferability and prototypicality of the 15 Chinese passive sentences, while the latter is to find out the manifestations of the language transfer in learners' production of passive voice. The results suggest that prototypicality is predictable in language transfer between languages of different language families. There are generally two major kinds of manifestations of language transfer: the positive and negative. The latter includes over-use, under-use, mal-form use and other errors. Based on the analysis of these manifestations, it is found that as for the relations of prototypicality and language transfer, if the structures are perceived by learner to be prototypical, they will be transferred as a kind of positive transfer, which promotes the acquisition of basic concept of the corresponding structure in the second language. However, the deeply rooted prototypical construction in the first language may cause some errors in the expression of the corresponding form in the second language if there are some differences syntactically between the two languages, which often leads to negative transfer. As to the situation of non-prototypicality perceived by learners, because the results are quite inconsistent, no conclusions can be made yet and further investigation is in need.