Towards a Generative-Relational Taxonomy of Governance

Governance is commonly classified through institutional forms, modes of coordination, jurisdictions, policy domains, or normative traditions. Such classifications illuminate important differences among governance arrangements, while providing a more limited vocabulary for distinguishing the structural objects upon which governance interventions operate. This paper develops a generative-relational taxonomy of governance organized around the layers through which relational systems generate, sustain, and transform their trajectories. The proposed framework distinguishes four analytically separable layers: state and explicit-rule governance, dynamical-process governance, relational-structural governance, and generative-background governance. Each layer contains multiple governance mechanisms. Dynamical-process governance includes event-based intervention, feedback governance, perturbative governance, tangent-space governance, and criticality governance. Relational-structural governance encompasses network, coupling, boundary, modular, distributed, and polycentric forms of coordination. Generative-background governance concerns structures that shape the conditions under which trajectories and relations become possible, accessible, stable, or costly, including generative conditions, structural flows on manifolds, background geometry, fields, symmetry structures, gauge-like connections, and emergent backgrounds. A formal representation is developed using dynamical systems, relational structures, differential geometry, and related mathematical languages. The taxonomy treats governance interventions as operators acting upon different components of a generative system and allows a single institutional intervention to be decomposed across multiple structural layers. This layered representation supports comparison among governance theories that employ different conceptual vocabularies and clarifies the relation between immediate intervention, trajectory regulation, relational reconfiguration, and modification of the conditions of generation. The paper presents the taxonomy as a foundational framework for subsequent theoretical and empirical development. Its broader objective is to provide a common analytical language for studying governance in heterogeneous, nonlinear, evolving, and partially observable relational systems while preserving the possibility of plural and revisable forms of order.