- RNP
- Motor Learning
- Predict, err, recalibrate: how the system learns
- Ecological dynamics versus predictive coding
SeriesPredict, err, recalibrate: how the system learns5 of 9
Ecological dynamics versus predictive coding
Two major schools of thought on how movement is learned, and how RNP practice puts them to work together.
In short. This tension refers to the fundamental opposition between two paradigms of motor learning. The ecological approach (Gibson, Bernstein, Newell, Davids, Rob Gray, Frans Bosch) is built on direct perception and perception-action coupling, with movement emerging from constraints. The predictive approach (Friston, Clark) treats prediction as the core mechanism: the brain works to minimize the gap between what it expects and what it perceives. RNP does not pick a side. It combines the two in a workable sequence.
Two paradigms, two ways of thinking about movement
On one side is ecological dynamics. In this lineage, movement arises from the direct coupling between athletes and their environment. They perceive affordances, opportunities for action, with no conscious calculation in between, and coordination self-organizes under the influence of constraints. This is the legacy of Bernstein and his famous phrase, repetition without repetition. Today it is carried forward by Davids, Rob Gray and Frans Bosch, and by teaching approaches built on motor variability.
On the other side is predictive coding. For Friston and Clark, the brain is above all a prediction machine. It constantly anticipates its sensory inputs and acts to reduce the gap between prediction and reality, a process known as active inference. From this angle, movement becomes a way of confirming a prediction. The tension is real: one school downplays internal representation, while the other puts it at the center.
RNP's workable sequential approach
RNP rejects both forced reconciliation and dogmatic choice. It places the two registers in sequence over time. Before the movement, the system prepares and anticipates. Prediction describes this well, and on the postural side it matches the anticipatory work of Massion and Berthoz. At contact, perception-action coupling takes over, fed by calibrated variability that keeps the system adaptable. You predict, you act, you recalibrate on the error. This fits both the sensorimotor loop and the test–stimulation–retest cycle.
This approach has a major consequence for how learning happens. As the LabO RNP Level 2 manual puts it: "Action is not the starting point of motor learning. It is a consequence of it." In other words, you do not impose a movement. You shape the conditions so that it can emerge.
The RNP training space, free for life
Get access to RNP training, no strings attached.
The RNP perspective
This topic is central to motor learning. This article gives its operational angle: when to activate each register. The main article sets out the full scientific lineage of both paradigms. For human movement professionals, the point is not to choose one school over the other. It is to know which register serves which moment.
→ Related terms: Active inference (Friston) · Clark: Predictive Processing and Surfing Uncertainty (2013-2016) · Perception-action coupling · Affordance · Bernstein 1967 · Differential learning
Scope: a functional, non-medical approach. It complements healthcare, it does not replace it.