Proceedings of the Annual Meeting of the Cognitive Science Society

Cultural norms vary dramatically across social groups. Herewe use large scale data to examine the extent to which languageplays a role in shaping one such norm—the gender norm to as-sociate men with careers and women with family. We measurecross-cultural variability in this gender bias using previously-collected estimates from the Implicit Association Task (IAT; N= 663,709). We then try to predict bias variability by the waythat gender is encoded in language semantics and grammar.We quantify gender bias in semantics using word-embeddingmodels trained on different languages. Our data suggest thatthe linguistic encoding of gender predicts the degree of speak-ers’ gender bias in the IAT, pointing to a causal role for lan-guage in shaping gender norms.