Archer and Friends: a mixed-methods approach to characterization in TV series.

Abstract

This study employs a comparative, computational approach to analyze the characterization of male and female characters in two popular television series, Archer and Friends. Utilizing telecinematic discourse analysis alongside advanced computational techniques including sentiment analysis, word embeddings, and Transformers-based sequence classification, this research explores how these series portray gender roles and stereotypes. Our analysis reveals that Archer, characterized by its satirical tone, subverts traditional gender norms more dynamically than the now-classic show Friends, which tends to reinforce conventional gender roles through its narrative and character interactions. Sentiment analysis indicates that Archer employs a more varied emotional language, reflecting complex character development, while Friends maintains a consistent emotional tone that aligns with its straightforward situational comedy. Additionally, embeddings-based characterization shows that Archer features more linguistically distinct characters than Friends, suggesting deeper narrative complexity. These findings highlight the influence of narrative style on audience perceptions of gender, demonstrating the usefulness of integrating computational methods with traditional media analysis to uncover nuanced representations of identity in television.