Narrative Interoperability: A Decision Procedure for AI Across Incommensurable Narrative Systems

This article makes a conceptual contribution at the intersection of philosophy and artificial intelligence. It introduces narrative interoperability, a discipline that specifies the conditions under which one and the same computational system can operate across several narrative universes without reducing one of them to an error or suspending the established results of another. It advances three theses. First, a narrative universe is characterized not by its content but by its limits, and these limits have a threefold origin: systemic, essential to the knowing subject, essential to the object. Second, the distinction between these origins is decided by a procedure of four tests—substitution, formalization, internal contradiction, transferability—which an AI system can apply without taking a position on the truth of any narrative. Third, the procedure is reflexive: it applies to itself, and its own limits are stated as part of what it asserts. The article situates narrative interoperability with respect to pluralistic alignment and to work on interpretive authority: where these frameworks ground their requirements in political or epistemic values external to the narrative, narrative interoperability derives its constraint solely from the status of the tool. The result is a decision procedure, not a doctrine.