Challenging implicit gender stereotypes in Italian news through language

This paper investigates gender stereotypes in Italian news by analyzing statistical representations provided by Word Embeddings of a sample of occupational names and political titles. The findings demonstrate that journalistic language encodes gender associations, which reflect both societal gender inequalities and socio-cultural biases. Further analysis explores the role of Italian grammar, indicating that the Italian gender-marked grammatical system can contribute to mitigating stereotypes, although it cannot completely eliminate them. Based on these findings, the article highlights the importance of computational linguistics tools in identifying latent biases in news language and recognizes journalism's fundamental role in promoting gender-fair language.