Speaking of gender: Language genderedness and its association with gender differences in personality across 48 languages

Previous studies have suggested that personality assessments can be influenced by the language people speak. Therefore, gendered ways of thinking, feeling and behaving may be associated with gendered structures encoded in different languages. Here, we study the association be-tween gender differences in personality traits as measured by the IPIP-NEO (n = 755,307; rep-resenting 48 languages from 122 countries, with English being the language of assessment and therefore held constant) and the genderedness of languages as rated by experts, estimated by word embedding models on large-scale text corpora from movie subtitles and Wikipedia, and rated by large language models such as ChatGPT, Gemini and DeepSeek. Consistent across all measures of language genderedness, more gendered languages were associated with stronger gender differences in personality traits compared to less gendered languages (r = 0.51 to 0.59), suggesting that language might influence people’s self-concept in terms of their gender.