語碼轉換的構詞句法機制：一個理論與實證的分析, 特以漢英語碼轉換為例

The study of code switching, i.e., the use of more than one language in the production of a sentence, has shown that this phenomenon follows regular grammatical patterns (Poplack 1980, 1981, Di Sciullo, Muysken & Singh 1986, Belazi, Rubin & Toribio 1994, Mahootian & Santorini 1996, MacSwan 1997, 2000). The goal of this project is to investigate the mechanism that generates and constrains the morphosyntax of code switching, with special reference to the facts of Chinese-English code switching. The project was conducted in two consecutive parts: the first part focused on the theory and the second part focused on testing the formulated hypotheses against empirical facts by collecting judgments of Chinese-English bilinguals. The major findings include: (i) The Free Morpheme Constraint by Poplack (1980, 1981) only applies to inflectional but not derivational morphemes. (ii) The mechanism of code-switching must incorporate the concept of the matrix language frame, but allows the lexical items from different languages project based on their own properties. (iii) The facts of Chinese-English code-switching do not support the claim that all Chinese nouns are mass nouns as proposed by Chierchia (1998) and Borer (2005).