1. RNP
  2. Motor Learning
  3. Constraints and self-organisation
  4. Constraint beats instruction: why repetition changes nothing

Motor LearningSeries 4/8

SeriesConstraints and self-organisation4 of 8

Constraint beats instruction: why repetition changes nothing

Constraint beats instruction: changing the environment and the task brings out the solution better than telling someone what to do. An action principle drawn from the constraints model.

LabO RNP·5 min read·reviewed on

In short. Constraint beats instruction. Changing the environment, the task or the target brings out the motor solution more effectively than telling someone what to do. A constraint acts directly on the available information, so it does not depend on understanding.

You say "push more horizontally" for the tenth time. The athlete nods, sweats, goes again, and the ground contact stays exactly the same. You feel the frustration building, yours and theirs. The problem isn't that they aren't listening. The problem is that your words never reach the place where the movement is decided.

What instruction cannot reach

Instruction acts on cognition. It assumes understanding. Above all, it assumes the athlete already perceives what you're talking about. If perception is faulty, your cue gets misread or simply ignored. It becomes noise laid over a system that has no way to use it.

Think of explaining a flavor to someone who has never tasted it. You can describe bitterness in perfect words. As long as the mouth has perceived nothing, your description creates no reference point. A motor cue works the same way. You can say "balance yourself" a thousand times to a child on a beam. The word stays empty as long as nothing in their body latches onto concrete information.

And the bill gets heavier with every repetition. The more you repeat a cue that isn't landing, the more cognitive load and frustration you pile onto a system that has no way to turn it into movement. You're not correcting. You're cluttering.

Constraint speaks the system's language

Constraint works differently. It changes the available information directly and alters the cost of each solution, without ever going through words. The system doesn't need to understand, focus or apply a cue. You change the landscape, and adaptation happens.

Go back to the sprinter. Instead of repeating your sentence, you change the push-off angle, the distance to cover or the target to reach. The motor solution changes on its own, without a single verbal correction. The ear received nothing new. The system, however, received new information through the body: through the ground contact and the resistance felt under the foot.

That's why constraint is more powerful than instruction. It doesn't depend on understanding, attention or willpower. It reshapes the terrain the system operates in, and self-organization does the rest. This mechanism follows directly from the constraints model and the logic of the constraints-led approach. You don't prescribe a shape. You sculpt the conditions that make a solution likely.

The RNP training space, free for life

Get access to RNP training, no strings attached.

Welcome! Introduce yourself

SecureNo cardNo password

By continuing, you agree to our Terms of Use and our Privacy Policy.

Three concrete levers, not a stance

Constraint is not an abstract idea. It works through three families of levers you can manipulate directly, as detailed in the article on organism, task, and environment constraints.

On the task side, you redefine the goal, the target or the success criterion. In physical therapy, telling an anxious person to "put more weight on it" triggers their fear and locks everything up. Changing the environment works better: lower an obstacle, move a support. That brings out spontaneous loading, without the sentence that sets off the defense. Instruction wakes the guard. Constraint goes around it.

On the environment side, you change the surface, the space or the tempo. Change the surface, and the ground-contact pattern changes. Change the space, and the motor strategy changes. The environment isn't scenery. It is an active source of information that speaks to the system more clearly than any correction. That is exactly what the notion of affordance describes: the invitation to act that the environment sends directly to the body.

On the organism side, the most overlooked point, you also act on perception, tolerance of error and the person's current state. A good task constraint makes certain errors costly and certain solutions rewarding. It teaches without explaining. The error provides clear information, instead of that information coming from your mouth.

Why it truly shifts the movement

You want to change the movement in depth, not get a nice shape for three minutes. This is where the concept explored in Attractor: why your correction holds for two repetitions then fades comes in. An attractor is the preferred solution the system spontaneously returns to because it is stable, economical and reassuring.

A runner keeps sitting in their stride despite all your technical work. This attractor is stable but not optimal, and it persists because it feels perceptually safe. You've tried everything through cues, and nothing moves. The reason is simple: you never changed what makes that attractor stable. The shape is only the visible part. As long as you correct the shape with words, you leave intact whatever pulls the system back to the same place.

Constraint shifts the cost. It makes another solution more rewarding, more accessible and more natural than the old one. The system isn't ordered to change. It discovers that somewhere else, things are cheaper.

The condition nobody reads

There's a trap, and it's a serious one. A constraint cannot create learning if the system isn't available to explore. You can design the smartest constraints in the world. But exploration won't happen if the system is in defense mode, if the sensorimotor loop is saturated or if the sensory hierarchy is stuck. In that state, the constraint will only make the rigidity worse.

On an unstable surface, a client either explores or stiffens up depending on their current state, not on the quality of your exercise. A constraint that suits an available system becomes toxic for a defensive one. This is exactly where the RNP reading comes in. Before designing a constraint, you read whether the system can explore. Otherwise, you're building on sand.

Even when the system is available, keep your eye on the intention. You don't change things just to change them. Unstructured motor variability produces costly wandering, not learning. This caution connects with the work on guidance and feedback dependency and on Internal versus external focus of attention: why talking less about the body frees movement. Too much verbal help, poorly dosed, traps people where you thought you were making things easier.

"Constraint teaches what instruction explains. And in most cases, teaching is more powerful than explaining." Adapted from the LabO RNP manual, level 2.

In one sentence: instead of saying what to do, change the task, the environment or the target so the right solution becomes the most rewarding one. This only works if the system is available to explore it.

Field test: the next time you're about to repeat a cue for the third time, stop. Ask yourself which single constraint you could change so the movement emerges without you saying a word. It could be the surface, the distance, the target or the tempo. Then watch what changes before you open your mouth.

When I stop prescribing the shape and start sculpting the landscape, I almost always find that the system already knew. It was just waiting for the right conditions.

Scope: a functional, non-medical approach. It complements medical care; it does not replace it.

€0Your free RNP PathsStart