Tell, Don't Show: Leveraging Language Models' Abstractive Retellings to Model Literary Themes

TL;DR

This paper introduces Retell, a novel method that uses language models to generate summaries of literary passages, enabling more accurate topic modeling of complex texts by translating narrative details into higher-level themes.

Contribution

The paper presents a new approach combining language model retellings with LDA to improve topic modeling in literature, addressing limitations of traditional lexical methods.

Findings

Retell improves the precision of topic modeling in literary texts.

LDA on retellings yields more informative themes than direct LDA or LMs.

Method demonstrates effectiveness in a case study on cultural identity in literature.

Abstract

Conventional bag-of-words approaches for topic modeling, like latent Dirichlet allocation (LDA), struggle with literary text. Literature challenges lexical methods because narrative language focuses on immersive sensory details instead of abstractive description or exposition: writers are advised to "show, don't tell." We propose Retell, a simple, accessible topic modeling approach for literature. Here, we prompt resource-efficient, generative language models (LMs) to tell what passages show, thereby translating narratives' surface forms into higher-level concepts and themes. By running LDA on LMs' retellings of passages, we can obtain more precise and informative topics than by running LDA alone or by directly asking LMs to list topics. To investigate the potential of our method for cultural analytics, we compare our method's outputs to expert-guided annotations in a case study on racial/cultural identity in high school English language arts books.