The book contributes to the limited sources on the bilingual situation in Hong Kong. The speech community -- students at the University of Hong Kong -- serves as a testing ground for evaluating the current approaches to sociolinguistic research in studying the factors that influence the students' mixing and selection of codes (dialects or languages).

Gibbons writes in a clear and concise style, providing useful references and introductory background on each approach before applying the methodology to the Hong Kong case. Four different sociolinguistic approaches are presented with one chapter devoted to each: 1) sociology of language approach, using census techniques to obtain statistics on code-switching and code choice; 2) the ethnographic approach, using recordings of situated speech for linguistic analysis; 3) the secular linguistic approach à la William Labov, using recordings of structured interviews to determine sociodemographic factors for linguistic variation; and 4) the social-psychological matched guise approach associated with W. E. Lambert and others, using bilingual speakers to obtain subject's attitudes toward English versus Cantonese speakers. Gibbons points out the strengths and weaknesses of each approach, but rightly argues that the approaches complement rather than conflict with each other in contributing to a better understanding of bilingual behavior in a speech community.

Gibbons has a second objective -- to build a more coherent model of code choice based on his findings. The model, offered in a speculative tone, attempts to include in stages all the various factors that interacted in influencing code choice in the Hong Kong situation. The overall model, however, could have been presented more succinctly.

A third and final objective of Gibbons is to provide more information on the speech behavior of his subjects. One chapter, occupying one-quarter of the book, is devoted to the linguistic descriptions of MIX, based on Cantonese with admixtures of English. The chapter is, however, unduly complicated by the use of Yale romanization, a phonemic inventory, and a phonetic transcription. Explanations of diacritics for tones, stress, and segments are buried in references and endnotes. The comparison of MIX with their English counterparts would have been more meaningful if both sets of data were in phonetic transcription instead of only the MIX forms. A morpheme-by-morpheme, or word-by-word, gloss of his examples would have increased their usefulness for the majority of readers unfamiliar with Cantonese.

Aside from various typographical errors in the transcriptions, there are also some errors that may have stemmed in part from his arbitrary reduction of Samuel Hung-nin Cheung's (not Yat-sing Cheung's) (1972) phonetic inventory to a phonemic one. Cantonese has [e], for example, which only occurs in the diphthong [ei], while [ε] occurs in other contexts. From examples such as Cantonese [k'εm] for English /kem/ "chem(istry)" and /kæmp/ "camp," phonetically [k'εm] and [k'æmp] respectively, Gibbons claims that "the English phonemes /e/ and /æ/ have been replaced by Cantonese /ε/ which falls between them" (p.46). He seems to have confused phonemics with phonetics. Cantonese simply lacks [æ] and replaces it with [ε]. (Note: "ε" is the epsilon-symbol, if one's web browser does not display it.)

Some facts that Gibbons finds puzzling in the Cantonese pronunciation can be explained with reference to phonotactics. The English syllable /fi/, for example, is pronounced [fei] rather than [fi] in Cantonese because in Cantonese, [i] only occurs after a high front glide or a sibilant, while [ei] occurs elsewhere, including after [f]. A few cases of [fi] in low tone are correlated with an unstressed [fi] syllable in English.

Although Cantonese constitutes eighty-five to ninety percent of the syllables in MIX, Gibbons contends that it is a separate code, an incipient koiné (the fusion of languages or dialects, such as standard Greek or English). The author does not fully address the issue that MIX is limited primarily to terms connected with student activities and academic affairs, with mixtures of trendy jargons and expressions coined by undergraduates. Nor does the author seriously reflect on the future status of Cantonese and English when Hong Kong is returned to China in 1997. The emergent role of Mandarin as the future symbol of political power is already present in new mixed Cantonese-Mandarin expressions in Hong Kong today.

In light of the scarcity of Chinese sociolinguistic works, the study should attract the interest of sociologists, anthropologists, linguists, as well as those involved in bilingual education or teaching English to Cantonese students.