Embodied predictive processing: A unifying generative principle for cognitive development

Human cognitive development unfolds through complex, non-linear interactions among the brain, body, and environment, producing diverse developmental trajectories across individuals and contexts. Capturing such variability requires a generative account that explains how cognition emerges and reorganizes over time, rather than models that describe isolated functions or static developmental stages. In this chapter, we propose Embodied Predictive Processing as a unifying generative framework for cognitive development. Building on predictive processing as a principle of brain function, we extend it to encompass bodily dynamics and social interaction, conceptualizing development as distributed prediction error minimization across brain-body-environment systems. Through integrative evidence from developmental studies and constructive robotic models, we demonstrate how brain-body co-development, body-grounded multimodal integration, and socially coordinated interaction jointly give rise to both developmental change and individual diversity. These findings position embodied predictive processing not only as a theoretical framework for understanding cognitive development, but also as a mechanistic and empirically testable generative account of developmental dynamics.