Narrative conflicts: a tri-modal computational analysis of antagonism in Shakespeare’s Julius Caesar

This study introduces a novel computational framework to analyze multi-modal antagonisms—semantic, emotional, and relational—in dramatic literature, specifically focusing on Shakespeare’s Julius Caesar. Employing natural language processing (NLP) techniques, text embeddings, emotion classifiers, and network-based character analyses, we systematically extract and quantify antagonistic relationships within the play. Semantic antagonisms are identified through hierarchical clustering and dimensionality reduction of character embeddings, revealing rhetorical groupings aligned closely with narrative functions. Emotional antagonisms, captured via emotion distribution profiles and variance analysis, illuminate characters’ affective dynamics and their alignment with dramatic roles. Relational antagonisms are explored through co-occurrence networks, highlighting unexpected centrality of minor characters as critical mediators of conflict. Integrating these modalities with Hegelian dialectics and Nietzschean interpretations, our tri-modal analysis provides fresh insights into ideological tensions, character motivations, and narrative structure. This interdisciplinary approach demonstrates the effectiveness of AI-driven tools in enriching literary criticism opening new avenues for exploring conflict dynamics in canonical texts.