A Conceptual Explication of Role Hierarchy in LLMs: The Inclusive Structure of Generation, Operation, Self-Modification, and Teleology

This concept paper argues that current discussions of LLM use often conflate generation, operation, self-modification, and teleology, obscuring important distinctions about autonomy, agency, and responsibility. To address this, it proposes a theory of role hierarchy composed of at least four layers: direct content generation, operation on generated outputs, modification of operational rules and execution mechanisms, and teleology, which determines goals, preservation criteria, and desirable directions of change. The paper’s central claim is that these layers differ not only in function but in the very objects they act upon: what is a finished product at one layer becomes material at the next. On this basis, it offers a foundational framework for clarifying human–LLM role division, the limits of self-modifying systems, and the locations of sovereignty and responsibility.

Unreviewed conceptual preprint. PDFs in English and Japanese. Generative AI tools were used for language refinement and translation support; the author takes full responsibility for the final text, references, and claims.

Keywords
LLM, role hierarchy theory, generation, operation, self-modification, teleology, autonomy, agency, operability filtration, foundations of intelligent systems