Emotion as System: A Foundational Architecture for Affect, Meaning, Perception, and Action

Affective computing has historically focused on systems that recognize, express, respond to, or influence emotion. This paper extends that lineage by proposing a foundational systems architecture for describing emotion as structural behavior, operable at individual, collective, and computational scales. Within this architecture, affect operates as an energetic substrate. Emotion configures a global field of sensitivity and responsiveness. Meaning introduces mass that deforms the field. Perception traces trajectories through it. Action follows gradients shaped by its geometry. Field, mass, trajectory, and gradient form a structural vocabulary intended to be operationalized computationally, tested empirically, and extended through design — work the paper frames as openings rather than results. Recent interpretability work on large language models recovers the same primary axes of valence and arousal, and the same clustered geometry, that organize human affective experience. This provides convergent evidence that emotion has computable structure. The contribution here is architectural: it supplies a system definition that makes emotion available as an object of formal reasoning across affective science, affective computing, design, and AI systems.