Toward a Science of Full-Body Intelligence: A Falsifiable Framework Integrating Embodied Cognition, Interoception, Predictive Processing, Neurovisceral Regulation, and Sensorimotor Grounding

Full-body intelligence is proposed as a falsifiable, organism-level construct integrating research on embodied cognition, interoception, predictive processing, active inference, allostasis, neurovisceral regulation, and sensorimotor grounding. Rather than treating cognition as exclusively brain-bounded, the framework examines how adaptive cognition and self-regulation may depend on dynamic interactions among neural, visceral, autonomic, sensorimotor, neuroendocrine, and environmental processes. Five candidate dimensions are proposed: interoceptive discrimination and metacognitive calibration; adaptive autonomic and cardiorespiratory regulation; flexible interoceptive–exteroceptive precision weighting; recovery and adaptive reorganization following perturbation; and, as an explicitly exploratory dimension, pre-stimulus physiological differentiation. The framework distinguishes convergent empirical evidence, supported theoretical models, integrative hypotheses, contested findings, and long-horizon extrapolations. It further specifies falsification criteria, competing explanations, multimodal measurement strategies, and a prospective experimental program for determining whether full-body intelligence has explanatory value beyond established cognitive and physiological constructs. The research program originated in experiential and hypothesis-generating inquiry by Daphne Garrido and was developed through iterative literature synthesis and AI-assisted research using ChatGPT (OpenAI) and Grok (xAI). AI outputs were used for ideation, organization, critique, and literature discovery and are not treated as scientific evidence.