What question did this study set out to answer?

The aim is to formalise the Recursive Admissibility Guard (RAG) to enhance stability in self-referential reasoning.

February 14, 2026Open Access

The Recursive Admissibility Guard (RAG): A Stability Constraint for Self-Referential Reasoning within the Paton System

Key Points

The aim is to formalise the Recursive Admissibility Guard (RAG) to enhance stability in self-referential reasoning.
Developed RAG as a machinery-layer component within the Paton System.
Implemented calibration frequency and observable re-anchoring.
Enforced external constraint injection for stability.
Established termination protocols upon reaching stability within tolerance bounds.
RAG successfully stabilises recursive reasoning processes.
Maintains structural viability without altering existing models.
Functions as a non-interfering overlay applicable across various domains.

Abstract

This paper formalises the Recursive Admissibility Guard (RAG) as a machinery-layer component within the Paton System. RAG stabilises recursive or self-referential reasoning by enforcing calibration frequency, observable re-anchoring, external constraint injection, and termination upon stability within tolerance bounds. The construct operates as a non-interfering admissibility overlay applicable across domains without modifying underlying physical or theoretical models. Positioned alongside Paton Compass, Paton Assist, Structural Honesty Gate, and Viability Permission Gate, the RAG preserves recursive integrity while maintaining domain neutrality and structural viability.

The Recursive Admissibility Guard (RAG): A Stability Constraint for Self-Referential Reasoning within the Paton System

Key Points

Abstract

Cite This Study