RAPID LARGE SCALE INTRA-NATIONALITY LANGUAGE SHIFT IN HONG KONG 1949-1971

According to Tsou (1978), post-1949 Hong Kong has evolved from a multi-dialectal society to a Cantonese-speaking society, and that not only is the scale of the shift relatively big, its pace is also unusually fast. In addition there appears to be significant differences in the rate of shift among the four dialect groups identified in the study. However a validation of Tsou's observations and a more precise estimation of the pace and scale of the shift have been hampered by (1) the inherent constraints of census data used by Tsou, (2) the loose and inconsistent use of dialect-group labels and (3) the absence of a direct study of the groups concerned. This paper reports the findings of our attempts to address these issues via documentary (census and historical) evidence, and a sociolinguistic survey conducted from 2005-07 among members of the groups concerned. Specifically, findings from our study of the documentary evidence largely corroborate Tsou's observations as they indicate that (1) Cantonese-speakers' share of the population increased from somewhere between 51.8% and 57.2% circa 1949 to 88.2% in 1971; (2) a significant differential rate in the groups' shift to Cantonese is in evidence albeit the number of dialect groups involved should be seven instead of the aforementioned four. Findings of our sociolinguistic survey of six of the seven groups indicate that (1) bilingualism is in evidence among most immigrant (G-0) interviewees, suggesting the shift is already in progress among them; (2) the shift to Cantonese with the corresponding loss of native-tongue proficiency is relatively complete among most G-1 and G-2 interviewees but (3) there are sufficient difference in native-tongue retention among the groups to rank them according to this ability in descending order as follows: Weitou, Kejia, Minnan, Chaozhou, Siyi and Shanghainese. Moreover the survey data indicate that (4) the shift took place with a relative absence of linguism, and that (5) it might have been facilitated by an apparent multiplicity in the number of factors that the survey participants use to define their nationality. Initial analysis of the data also indicates that the groups' settlement patterns might contribute to the differential rate of shift. It is further observed that Cantonese is not indigenous to Hong Kong, and that this shift may develop in tandem with the emergence of the Hong Kong identity. The shift may also be unique in the context of Modern China as it involves the