A CRITIQUE OF LANGUAGE USE SURVEYS IN SINGAPORE

Abstract

This dissertation presents a critical account of various studies to acquire data on linguistic behavior in Singapore. The data will be discussed within a framework of the limitations to the validity of various kinds of performance data on language while acknowledging within these limitations the contributions which the surveys have made. This dissertation reviews surveys which have sought to collect and analyze information on language and ethnicity in Singapore. The dissertation then examines the field and analytic techniques used in the surveys. This examination is undertaken for two reasons: First, to help arrive at an appraisal of what themes and areas are understood; and which, by implication remain to be illuminated. Second, to attempt to contribute to subsequent surveys both by summarizing difficulties notable in those under review and, where possible, by suggesting how the difficulties may be avoided or overcome. Three sources of limitation affect the validity of various kinds of performance data on language. One is what utterances are properly called and constitute samples of a language. This difficulty is discussed. Another is the extent to which data collection and analysis reflect the investigators rather than their topic; in any single study these two are necessarily interwoven, accordingly, a comparative review, like this one, has some hope of augmenting more empirical work. The remaining source of limitation is the kind(s) of performance data employed: eg. text, speech, self-reports, transcriptions, entirely public language or more domestic utterance. Not withstanding the limitations of the surveys considered, we should recognize their strengths. These surveys present certain data which is consistent and reliable. They give us a fair understanding of generational change and why people do not speak a language even though they can.