Chapter 34: Embodiment of Concepts and Predictive Processing

Embodiment of Concepts and Predictive Processing from The Oxford Handbook of 4E Cognition provides a comprehensive theoretical analysis of how human concepts are acquired, represented, and utilized, ultimately proposing an integration of embodied cognition with the predictive processing framework. The text begins by outlining the fundamental debate regarding concept acquisition, contrasting nativist views—which posit that infants are born with innate core systems for understanding objects and agents—with empiricist perspectives that argue concepts are learned through sensorimotor experience and associative mechanisms. It further examines the representational format of concepts, distinguishing between the classical cognitivist view of concepts as amodal, arbitrary symbols in a language of thought, and the embodied view where concepts are grounded in modality-specific brain systems involved in action and perception. The authors also discuss hybrid theories and the semantic hub hypothesis, which suggest that conceptual processing is flexible and context-dependent, utilizing both embodied simulations and abstract representations based on the level of embeddedness in the environment. A central theme of the chapter is the application of the free energy principle to conceptual theory, arguing that the brain functions as a prediction machine. In this model, concepts act as high-level generative models within a hierarchical network, producing top-down multimodal predictions to minimize surprise and prediction error signals. This perspective reconciles the role of priors with Bayesian learning, explaining how concepts enhance perceptual precision and are dynamically updated through recurrent error processing during interactions with the world.