Cantonese-English code-mixing: a New Zealand case study

Abstract

It is frequently observed in many bilingual situations that two languages are used back and forth in a speech event or even within a single sentence. Studies of this phenomenon, especially in the past two decades, have been growing rapidly. While different terms have been used to refer to these language patterns, two terms are most frequently used to address them, i.e. "code-switching" and "code-mixing". Most recent studies have used these two terms to refer to the different patterns, i.e. "code-switching" for the switches that take place across the sentence boundary, and "code-mixing" for switches that take place within a sentence. This is the distinction which is adopted in the present thesis in studying the Cantonese-English code-mixing among the Cantonese speakers in New Zealand. While some researchers, especially the early scholars, have tended to regard code-switching and code-mixing as random or irregular, most recent researchers consider them as rule-governed, and have made attempts to formulate the syntactic rules or constraints governing these linguistic patterns. Based on the data collected among the selected Cantonese speaking subjects who were living in New Zealand, the present study attempts to study the phenomenon of Cantonese-English code-mixing through the integration of formal and functional approaches. It sets out to explore the linguistic features of the Cantonese-English code-mixing among these subjects, addresses the issue of whether code-mixing is random or rule-governed, and further discusses the probable constraints, on the assumption that there are underlying rules. It also analyses the probable factors influencing the subjects' choice of the mixed code, and the main functions this kind of code-mixed utterance performs. It was found that Cantonese-English code-mixing was not taking place at random; on the contrary, there were sets of constraints governing it. The code-mixed forms had their own linguistic characteristics, especially phonologically and syntactically. Though some individual factors, such as English proficiency, the situation and the topic, were observed to have particular bearing on the subjects' choice of code, Cantonese-English code-mixing was mostly influenced by a combination of several factors. It is concluded that it is the speakers' response to their perceptions of the factors present, together with their communication goal, that determines whether code-mixing is possible and to what extent it should be used.