Contrasting Statistical Indicators of Maori Language Revitalization: Conversational Ability, Speaking Proficiency, and First Language

Is it possible to track the revitalization of the Maori language statistically? Different large-scale statistical collections (censuses and surveys) in New Zealand effectively have different definitions of speaker because they ask different questions. This paper compares trends in numbers of Maori speakers as estimated from responses to questions about conversational ability, first language, and level of speaking proficiency, with particular reference to the 2013 Census and Te Kupenga (Maori social survey) 2013. One might expect estimates based on these responses to align closely, but they do not. This paper explores the relationships between the different estimates for different birth cohorts. Data on first language from at least four surveys provide strong evidence of a resurgence in intergenerational language transmission, which is not clearly apparent from the other indicators. Patterns of response to conversational ability and speaking proficiency questions are found to vary according to first language and birth cohort. It is argued that the apparent inconsistencies between the indicators reflect the real complexity of revitalization processes, as well as varying interpretations of the language questions, and that the New Zealand census language question on conversational ability is of questionable value as an indicator for tracking Maori language revitalization.