Creative Convergence or Imitation? Genre-Specific Homogeneity in LLM-Generated Chinese Literature
This addresses the issue of repetitive and stereotypical story generation in LLMs for Chinese web literature, which is incremental as it applies an extended theoretical framework to analyze a specific domain.
The paper tackles the problem of structural homogeneity in LLM-generated Chinese literature by proposing a novel theoretical framework based on Proppian narratology, revealing that LLMs fail to comprehend narrative functions and rely on rigid paradigms, leading to severe homogenization.
Large Language Models (LLMs) have demonstrated remarkable capabilities in narrative generation. However, they often produce structurally homogenized stories, frequently following repetitive arrangements and combinations of plot events along with stereotypical resolutions. In this paper, we propose a novel theoretical framework for analysis by incorporating Proppian narratology and narrative functions. This framework is used to analyze the composition of narrative texts generated by LLMs to uncover their underlying narrative logic. Taking Chinese web literature as our research focus, we extend Propp's narrative theory, defining 34 narrative functions suited to modern web narrative structures. We further construct a human-annotated corpus to support the analysis of narrative structures within LLM-generated text. Experiments reveal that the primary reasons for the singular narrative logic and severe homogenization in generated texts are that current LLMs are unable to correctly comprehend the meanings of narrative functions and instead adhere to rigid narrative generation paradigms.