This study formulates a unified, substrate-neutral theory of consciousness grounded in a universal operator framework for information dynamics, x(t + 1) = Σ(Wx(t) + b(t)), which enables an equivalent representation of neural systems, artificial neural networks, reaction–diffusion media, and distributed multi-agent collectives. Within this formalism, consciousness is defined as an emergent property of a causally closed recursive Self subsystem that maximizes integrated informational coherence, exhibitshigh causal density, and maintains a predictively valid model of its environment. Subjective experience is operationalized as a functional consequence of self-referential information processing, with pain interpreted as a highly prioritized interoceptive prediction error necessary for the optimization of regulatory policies within a higher-order control loop. The proposed three-axis model of consciousness—(i) depth of introspective recursion, (ii) complexity of the environment model, and (iii) integration and regulatory capacity—is mapped onto measurable quantitative variables, including integrated information Φ, spectral characteristics of the W matrix, causal density, and error-signal dynamics within predictive coding architectures. Application of the model to neural systems, advanced AI architectures, swarm systems, and aneural organisms (e.g., Physarum polycephalum) provides evidence for strong substrate neutrality and enables a reinterpretation of the “hard problem” of consciousness as a question of system architecture and informational thermodynamics. The resulting framework supports the definition of quantitative consciousness metrics (e.g., KEMI) and establishes formal-engineering prerequisites for deriving the moral status of artificial systems.