Creative Convergence or Imitation? Genre-Specific Homogeneity in LLM-Generated Chinese Literature
arXiv:2603.14430v1 Announce Type: new Abstract: 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 tha
arXiv:2603.14430v1 Announce Type: new Abstract: 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.
Executive Summary
This study employs a novel theoretical framework combining Proppian narratology and narrative functions to analyze Large Language Model (LLM)-generated Chinese literature. The research reveals that current LLMs produce structurally homogenized stories due to an inability to comprehend narrative functions and adhere to rigid narrative generation paradigms. The findings suggest a need for LLM development to incorporate more nuanced understanding of narrative structures and functions. The study contributes to the field of natural language processing by providing a systematic framework for analyzing narrative logic in LLM-generated texts. The implications of this research extend to the development of more sophisticated narrative generation capabilities and the potential to create more engaging and diverse digital literature.
Key Points
- ▸ The study employs a novel theoretical framework combining Proppian narratology and narrative functions to analyze LLM-generated Chinese literature.
- ▸ Current LLMs produce structurally homogenized stories due to an inability to comprehend narrative functions and rigid narrative generation paradigms.
- ▸ The research contributes to the field of natural language processing by providing a systematic framework for analyzing narrative logic in LLM-generated texts.
Merits
Strength
The study provides a systematic and theoretically grounded framework for analyzing narrative logic in LLM-generated texts, which contributes to the advancement of natural language processing research.
Demerits
Limitation
The study's focus on Chinese web literature may limit the generalizability of the findings to other languages and genres, and further research is needed to validate the applicability of the framework across different contexts.
Expert Commentary
This study makes a significant contribution to the field of natural language processing by providing a systematic framework for analyzing narrative logic in LLM-generated texts. The findings of the study highlight the need for more advanced LLMs that can comprehend narrative functions and generate diverse narratives. However, the study's focus on Chinese web literature may limit the generalizability of the findings to other languages and genres. To overcome this limitation, future research should aim to validate the applicability of the framework across different contexts and languages. Furthermore, the study's implications for the development of AI-powered content creation tools and the potential impact on the creative industries and cultural landscape warrant further consideration. Overall, this study provides a valuable contribution to the ongoing conversation on the potential and limitations of AI-generated content.
Recommendations
- ✓ Recommendation 1: Future research should aim to develop more advanced LLMs that can comprehend narrative functions and generate diverse narratives.
- ✓ Recommendation 2: Policymakers should consider the potential impact of AI-generated content on the creative industries and the cultural landscape, and develop guidelines for the responsible development and deployment of AI-powered content creation tools.