Bilinguals in Style: Linguistic Practices and Ideologies of Cantonese-English Codemixers in Hong Kong.

Abstract

The trilingual (Cantonese, Putonghua and English) and multicultural setting of Hong Kong makes it a language contact zone in which different patterns of code-mixing occur. Previous studies of Hong Kong code-mixing mostly focus on the major pattern commonly found among locally educated ethnic Chinese; little has been done on the coexistence of different code-mixing patterns and their social significance. This research employs Irvine’s (2001) conception of ‘style’, and the associated Irvine and Gal (1995) semiotic processes of language ideologies, to investigate two code-mixing patterns found in Hong Kong and to explore how they are used indexically to construct distinct social and linguistic identities. The code-mixing style commonly used by the local younger generation, using Muysken’s (2000) typology, is insertional, in that individual English lexical items are inserted into a base language of Cantonese at an intra-sentential level. In contrast, another code-mixing style, which correlates with speakers who have overseas and/or international school experience, is structurally much more complex. It has a combination of insertion (Cantonese insertion into English sentences and vice versa), alternation between the two languages, and the use of discourse markers at switch points. For the local younger generation, most of whom went through Hong Kong’s bilingual education system, use of the local code-mixing style is a way to identify and interact with people of shared commonalities. It also provides a means to distinguish “outsiders” who use or prefer a different style of language mixing (or non-mixing). This research reveals how overlapping and fuzzy the linguistic and social boundaries between Hong Kong locals and returnees are, yet social participants essentialize the relationship between speech and speakers, using such knowledge to construct, negotiate, and (re)position their identities, make decisions about whether or not to cross perceived social group boundaries; and maneuver in their local social contexts and beyond. This research demonstrates that, to understand language and its speakers as social beings, linguistic structures must be studied in conjunction with their contextualized use as well as the mediating ideologies, i.e. the three components Silverstein (1985) defines as constituting a “total linguistic fact”.