Style Over Story: A Process-Oriented Study of Authorial Creativity in Large Language Models

TL;DR

The paper advances a process-oriented view of LLM creativity by applying narratology to computational authorship. It introduces constraint-based authorial decision making, deploying a library of 200 narrative constraints and three system personas to elicit and analyze how models allocate emphasis among Style, Character, Event, and Setting. Across six model families, the study finds a robust preference for Style over other narrative elements and reveals distinctive model-, event-, and persona-specific profiles in both constraint selection and the reasoning provided. The framework enables reproducible analyses of AI authorship, with implications for controllability, bias audits, and collaborative creative tasks in NLP.

Abstract

Evaluations of large language models (LLMs)' creativity have focused primarily on the quality of their outputs rather than the processes that shape them. This study takes a process-oriented approach, drawing on narratology to examine LLMs as computational authors. We introduce constraint-based decision-making as a lens for authorial creativity. Using controlled prompting to assign authorial personas, we analyze the creative preferences of the models. Our findings show that LLMs consistently emphasize Style over other elements, including Character, Event, and Setting. By also probing the reasoning the models provide for their choices, we show that distinctive profiles emerge across models and argue that our approach provides a novel systematic tool for analyzing AI's authorial creativity.

Figures (1)

• Figure 1: Overview of the study workflow. A library of narrative constraints(four elements, with five categories per element and ten constraints per category) is presented via a standardized user prompt, and six system-prompted LLMs conduct repeated runs, selecting exactly 20 constraints from the pooled list.