Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Revisiting Northrop Frye's Four Myths Theory with Large Language Models

Edirlei Soares de Lima, Marco A. Casanova, Antonio L. Furtado

TL;DR

This paper integrates Frye's four literary genres with Jungian archetypes to form a four-function, sixteen-role character framework for computational narratology. By grounding functions in Jungian psyche and specializing them to genre-specific roles derived from prototypical works, the authors enable structural character analysis across diverse narratives. They validate the framework through a multi-LLM study on 40 works (with 160 positive mappings) and 20 negatives (30 invalid mappings), achieving a mean balanced accuracy of about $82.5\%$ and substantial inter-model agreement ($κ=0.600$). The findings reveal both strong, genre-consistent patterns and expected ambiguities in romance and satire, highlighting implications for narrative analysis and interactive storytelling while noting limitations such as gender biases and prototype dependence. Collectively, the work advances computational narratology by providing a rigorous, testable bridge between classical theory and data-driven narrative generation.

Abstract

Northrop Frye's theory of four fundamental narrative genres (comedy, romance, tragedy, satire) has profoundly influenced literary criticism, yet computational approaches to his framework have focused primarily on narrative patterns rather than character functions. In this paper, we present a new character function framework that complements pattern-based analysis by examining how archetypal roles manifest differently across Frye's genres. Drawing on Jungian archetype theory, we derive four universal character functions (protagonist, mentor, antagonist, companion) by mapping them to Jung's psychic structure components. These functions are then specialized into sixteen genre-specific roles based on prototypical works. To validate this framework, we conducted a multi-model study using six state-of-the-art Large Language Models (LLMs) to evaluate character-role correspondences across 40 narrative works. The validation employed both positive samples (160 valid correspondences) and negative samples (30 invalid correspondences) to evaluate whether models both recognize valid correspondences and reject invalid ones. LLMs achieved substantial performance (mean balanced accuracy of 82.5%) with strong inter-model agreement (Fleiss' $κ$ = 0.600), demonstrating that the proposed correspondences capture systematic structural patterns. Performance varied by genre (ranging from 72.7% to 89.9%) and role (52.5% to 99.2%), with qualitative analysis revealing that variations reflect genuine narrative properties, including functional distribution in romance and deliberate archetypal subversion in satire. This character-based approach demonstrates the potential of LLM-supported methods for computational narratology and provides a foundation for future development of narrative generation methods and interactive storytelling applications.

Revisiting Northrop Frye's Four Myths Theory with Large Language Models

TL;DR

This paper integrates Frye's four literary genres with Jungian archetypes to form a four-function, sixteen-role character framework for computational narratology. By grounding functions in Jungian psyche and specializing them to genre-specific roles derived from prototypical works, the authors enable structural character analysis across diverse narratives. They validate the framework through a multi-LLM study on 40 works (with 160 positive mappings) and 20 negatives (30 invalid mappings), achieving a mean balanced accuracy of about and substantial inter-model agreement (). The findings reveal both strong, genre-consistent patterns and expected ambiguities in romance and satire, highlighting implications for narrative analysis and interactive storytelling while noting limitations such as gender biases and prototype dependence. Collectively, the work advances computational narratology by providing a rigorous, testable bridge between classical theory and data-driven narrative generation.

Abstract

Northrop Frye's theory of four fundamental narrative genres (comedy, romance, tragedy, satire) has profoundly influenced literary criticism, yet computational approaches to his framework have focused primarily on narrative patterns rather than character functions. In this paper, we present a new character function framework that complements pattern-based analysis by examining how archetypal roles manifest differently across Frye's genres. Drawing on Jungian archetype theory, we derive four universal character functions (protagonist, mentor, antagonist, companion) by mapping them to Jung's psychic structure components. These functions are then specialized into sixteen genre-specific roles based on prototypical works. To validate this framework, we conducted a multi-model study using six state-of-the-art Large Language Models (LLMs) to evaluate character-role correspondences across 40 narrative works. The validation employed both positive samples (160 valid correspondences) and negative samples (30 invalid correspondences) to evaluate whether models both recognize valid correspondences and reject invalid ones. LLMs achieved substantial performance (mean balanced accuracy of 82.5%) with strong inter-model agreement (Fleiss' = 0.600), demonstrating that the proposed correspondences capture systematic structural patterns. Performance varied by genre (ranging from 72.7% to 89.9%) and role (52.5% to 99.2%), with qualitative analysis revealing that variations reflect genuine narrative properties, including functional distribution in romance and deliberate archetypal subversion in satire. This character-based approach demonstrates the potential of LLM-supported methods for computational narratology and provides a foundation for future development of narrative generation methods and interactive storytelling applications.
Paper Structure (30 sections, 1 figure, 9 tables)

This paper contains 30 sections, 1 figure, 9 tables.

Table of Contents

  1. Introduction
  2. The Character Functions Approach
  3. Experimental Validation
  4. Validation Dataset
  5. Sample Generation and Validation
  6. Negative Sample Construction
  7. Evaluation Protocol
  8. Model Selection
  9. Prompt Design
  10. Data Collection and Processing
  11. Evaluation Metrics
  12. Results and Discussion
  13. Overall Performance
  14. Model Agreement and Consensus
  15. Performance by Genre
  16. ...and 15 more sections

Figures (1)

  • Figure 1: Diagram of Jung's psychic structure showing the four components (ego, persona, shadow, anima) and their integration through the individuation process Jacobi2013.