Big Data, Actually: Examining Systematic Messaging in 188 Romantic Comedies Using Unsupervised Machine Learning

Abstract

Popular films within the romantic comedy genre set social expectations for romantic relationships, sexual activity, and gender roles, especially for young audiences. Nevertheless, academic research on the genre is scarce and mostly limited to manual content analyses conducted on relatively small samples of films. To systematically examine trends in thematic content and social messages in romantic comedy scripts over time, we harness a novel computational method, the ANalysis of Topic Model Network to analyze 188 scripts of top-grossing romantic comedies in the United States between 1980 and 2019. We estimate the dynamic prevalence of themes in the dialogue and demonstrate the increasing focus on romantic relationships at the expense of other life aspects in recent decades. We demonstrate how this trend systematically resulted in an increase in the prominence of misleading and detrimental messages about romantic relationships. The theoretical, social, and practical implications of our findings are discussed.