Conservative and innovative behaviour by female speakers

The focus of this article is the supposed ‘Gender Paradox’, proposed by Labov (1990, 2001), that women are both sometimes conservative and sometimes innovative in terms of linguistic variation and change. Here we explore the paradox from two perspectives: we both investigate its applicability to multilingual as opposed to multidialectal communities as well as question whether the paradox is methodological or real. Although much sociolinguistic research on the paradox has been on macro studies of men vs. women in monolingual multidialectal communities, this paper presents quantitative analyses supplemented by in-depth ethnographic observation and data collection in communities of practice in a multilingual Japanese-Palauan community of the Western Pacific. What is more, and perhaps somewhat surprisingly, the ‘conservatism’ and ‘innovation’ of women in the community under scrutiny is represented by the use of the very same language –Japanese. For older Japanese Palauan women, the greater use of Japanese represents adherence to their heritage language. Among younger Japanese-Palauan women (most of whom are bilingual in Palauan and English), however, the use of Japanese represents a change towards a language highly valued in the economy as essential for the promotion of tourism and trade. Since our results demonstrate that the effects of gender on language behaviour may appear in differences within sex groupings, we conclude that the paradox is methodological, rather than real, and is a result of the distilling of gender down to binary male-female categories of analysis, rather than investigating the complexity of gender more qualitatively; ethnographic investigation enables us to identify conservative or innovative linguistic behaviour in the local context as well as which men/women are members of which social groups on the basis of their social role and social practice in the community. Our ethnographic analysis of multilingual data from Palau presented here demonstrates that function as well as form is important in understanding seemingly paradoxical examples of language shift, as well as highlighting the need both for further research on the effects of gender in multilingual communities and the combination of qualitative and quantitative analysis in studies of language change. * We would like to thank: all those in Palau who kindly extended their warm hospitality during fieldwork trips in 1997, 1998 and 2000; Enam Al-Wer and the audience at the VIEW conference (14-16 September 2000) at Essex for useful comments on previous versions of this paper; as well as Phil Scholfield for his statistical assistance.