Active Inference and Predictive Processing for Skill Learning: Implications for Skill Acquisition in Coaching Practice

This paper explores the potential of the Active Inference Framework (AIF) and Predictive Processing (PP) to advance our understanding of skill and motor learning with a key focus on introducing these concepts to coaches and highlighting their practical implications. Conceptual frameworks such as Information Processing (IP) and Ecological Dynamics (ED) have contributed significantly to the field of skill acquisition. Despite recent efforts to reconcile these perspectives, however, their fundamentally different theoretical foundations concerning perception, cognition and learning processes pose substantial challenges for integration. Although ED has made valuable contributions through its focus on perception-action coupling and direct interaction with the environment, it involves adopting a radical ontological stance that emphasises rejecting traditional cognitive constructs that can complicate practical applications in coaching. Conversely, IP's reliance on concepts such as stored motor programs and the view of the brain as a computer-like processor can limit its ability to account for the dynamic, emergent nature of skill acquisition in real-world contexts. We suggest that AIF offers a parsimonious perspective, accommodating the dynamic, embodied and enactive nature of skill acquisition without the necessity of committing to either framework's specific assumptions. This perspective provides practitioners with a coherent and flexible framework for understanding and designing skill learning interventions, offering a practically useful account of skill acquisition for coaching practice.