Innovators of language shift in the Chinese community

of Northeast England

David Walshaw & Li Wei

Fri. 2:00-4:05 C

This paper investigates the innovation process underlying the process of language shift in a minority Chinese community in an English speaking country. The community in question are the Cantonese-English bilinguals in the Northeast of England, centered around Newcastle upon Tyne. This group have been the subject of extensive study in recent years. The change in relative language use in conversational dialogue, from predominantly Cantonese among older members, towards an English bias among the children, is well documented (Li Wei, 1994). Other effects have also been identified, for example a clear tendency for males to use English more than females.

Here we attempt to identify the social factors underlying the innovations of the language shift. While it is tempting to claim that the social groups which use more English are responsible for driving the process, the issue is complicated by variations in relative language ability. For example, male members of the community tend to have higher competence in spoken English than females, which is partly responsible for increased usage.

Despite this, a clear element of choice remains in a bilingual speaker's use of language, which can be clearly observed in this community. For example, individual speakers use differing amounts of English with distinct interlocutors. We argue that it is this residual element of true choice that should be utilized in terms of explaining the innovation of language shift, while the relative language abilities serve as a gauge of the extent to which the shift has already occurred.

By making use of detailed measures of spoken ability in both Cantonese and English, we isolate the residual element of choice among 165 members of the community. This is done separately for several distinct groups of interlocutor. We are then able to explore the innovation of language shift, and how it relates to the social variables of the speaker and the interlocutor. We employ a range of analytical and statistical techniques, ranging from simple correlations, through to log-linear models. These more sophisticated models are needed to identify the interactions between speaker variables which influence the distinct categories of residual choice.

Reference

Li Wei (1994) Three Generations, Two Languages, One Family. Clevedon: Multilingual Matters 104.