Differential Governance: Coordination through Change among Systems Sharing No Common Ontology

Abstract

Coordinated action on an interdependent environmental field requires that parties describing the field in different terms act together on the same processes. The usual route to such coordination is agreement about states: a shared account of what a quantity is, and a shared target for it. This paper argues that agreement of that kind is expensive in representational commitment and frequently unavailable, and proposes a primitive that is cheaper. The method is conceptual and definitional, developed through a graded account of what agreement of successive kinds requires. The paper's central observation is that agreement about the direction and unusualness of a change demands far less shared representation than agreement about the identity of the quantity that changed, so that parties who cannot describe a system in common terms may nonetheless recognise, report, and act on correspondences among changes. Three consequences are developed. Differentials are proposed as an ontology-light coordination primitive, with the conditions under which correspondences among changes are identifiable and the conditions under which they mislead. The argument is grounded in a dynamical result and not in convenience alone: transitions induced by the rate at which a system is forced are not detectable from its state, so a governance arrangement addressed only to states is blind to a class of transition in principle. And the limits of the primitive are stated at length, including its blindness to a system already in an unacceptable state whose rate of change is zero, from which follows a division of labour between constraints stated over states and governance conducted over changes. The paper concedes the philosophical treatment of difference and repetition, the structural analysis of relational value, existing rate-based regulatory instruments, and derivative action in control engineering, and identifies what remains unclaimed. It reports no empirical study and establishes no formal result.