The Cross-Layer Theory of Complexity: A General Framework for Understanding Emergence, Constraint, and Agency in Multi-Layered Complex Systems

This paper proposes a general theory of complexity grounded in a four-layer ontological architecture. The theory advances a central claim: genuine complexity emerges if and only if a system spans multiple ontological layers, where each layer is characterised by a distinct governance mode, agency degree, and thermodynamic regime. We identify four abstract layers—Elemental, Substantive, Cooperative, and Organized—that manifest at multiple scales from individual organisms to civilisational systems. Nineteen principles govern inter-layer relations, organised into six categories: layer structure, vertical relations, interfaces, temporality, energy dynamics, and agency. The framework generates a complexity typology distinguishing dissipative, adaptive, intentional, reflexive, and supra-intentional complexity based on layer span. We demonstrate the framework's analytical power through contrasting case studies (hurricane versus stock market) and address the question of whether artificial intelligence may constitute a fifth emergent layer. The theory provides predictive capacity absent from existing frameworks, including Donella Meadows' leverage points and David Snowden's Cynefin, by grounding the simple-complicated-complex distinction in ontological layer structure.