Social salience, the association of a social category with linguistic variables, has been hypothesized to be an important driver of language change. This hypothesis has not been rigorously tested due to the lack of a reliable measure of social salience. In this paper, we present Salience Categorization Test (Sali-CAT), a new approach to measuring the association of word variants with social categories across multiple lexical variables. The approach includes a customized experimental paradigm (three alternative forced choice) and a statistical method to establish the baseline Salience Ratio (Sali-RAT) score for word variants that do not have a bias in usage with respect to the social categories. We demonstrate the approach by testing the association of multiple variables with different generations of speakers in the Gurindji speech community.
