- RNP
- Motor Learning
- Constraints and self-organisation
- Constraints-led approach (CLA): why the best constraint in the world can trap your athlete
SeriesConstraints and self-organisation2 of 8
Constraints-led approach (CLA): why the best constraint in the world can trap your athlete
The constraints-led approach (CLA): what it is, how it drives learning without instruction, and the three pitfalls that make it ineffective.
In short. The constraints-led approach (CLA) means organizing the context (organism, task, environment) so that certain movement solutions become more likely than others, without prescribing a form. You don't tell the athlete what to do: you sculpt the landscape in which the system self-organizes.
You set down two cones, narrow the space, add a sound cue that triggers the start. The athlete changes direction differently, without you saying a word about technique. You didn't correct anything. You changed the ground under their feet, and the movement shifted on its own. That's the constraints-led approach. It's also exactly where things can go wrong.
The constraints-led approach, plainly put
The principle fits in one sentence: movement isn't programmed, it emerges. You don't reproduce a movement someone shows you; you solve a problem posed by the context.
The model doesn't tell you to stop giving instructions, nor to let anything go. It asks something else of you: organize the context so that certain solutions pay off better than others. You shift from director to designer. From the one who demonstrates to the one who creates the conditions. The right movement is no longer dictated; it becomes the most likely outcome of a landscape you have shaped.
To dig deeper into the theoretical framework, see the constraints model and how constraint beats instruction.
Three families of constraints, never in isolation
The model distinguishes organism, task and environment constraints. A crucial detail: they never act alone. It's their interaction that shapes learning. Change just one of them, and the whole solution can tip over.
Organism constraints are morphology, strength, mobility. But also, and this is almost always overlooked, perception, tolerance to error, emotional state, sensory hierarchy. Two biomechanically identical athletes can perceive the same situation in opposite ways: one reads it as stable, the other as threatening. That difference in perception weighs as much as a difference in strength.
Task constraints define the goal, the rules, the cost of error. Environmental constraints speak directly to the body. Change the surface, and the loading pattern under the foot changes. Change the space, and the movement strategy reorganizes. The environment isn't a backdrop; it's an active source of information, the language the system understands best.
The full breakdown lives in the organism–task–environment triad, and reading the context as information continues in affordance.
The RNP training space, free for life
Get access to RNP training, no strings attached.
Why it works: you move the attractor, not the form
When the system finds a solution that is stable, economical and safe, it returns to it spontaneously. That's an attractor. Real learning isn't about correcting the visible form; it's about moving that attractor.
You know the runner who keeps sitting into their stride despite ten technique sessions. Their solution is stable but not optimal: it persists because it feels perceptually reassuring. As long as you go after the form, nothing changes. The day a well-placed constraint makes another solution more attractive, the landscape reconfigures and the movement changes by itself.
That's exactly what a task that forces calibration works on: it makes error informative and the right solution worthwhile, without a single word of explanation.
The three pitfalls that make CLA fail
This is where many practitioners get burned. CLA isn't magic, and when it's misunderstood, it blocks learning instead of driving it.
First pitfall: everything becomes a constraint. You change everything, all the time, with no clear intent. The result: you generate noise, you prevent stabilization, you wear the system out. A constraint has to be targeted, prioritized, dosed. You don't change for the sake of changing. You change to force a specific exploration, with intent.
Second pitfall: ignoring the state of the system. A constraint that's perfect for an available athlete becomes toxic for an athlete in defensive mode. You can build the smartest design in the world: if the sensorimotor loop is saturated, if the hierarchy is locked, your constraint won't create exploration, it will create rigidity. Before you design, read the state of the system. Otherwise, restore that capacity first.
Third pitfall: confusing exploration with wandering. Exploration is structured and produces information. Wandering is costly and produces nothing but noise. Without a fine-grained reading, you manufacture variability for its own sake and mistake it for learning. It isn't learning. It's confusion. The quality of your relevant information design makes all the difference between the two.
Where RNP comes in: a constraint creates nothing in an unavailable system
Not every system explores freely. A system that is defensive, sensorily overloaded or hierarchically unbalanced explores little, poorly, or always in the same direction. And there, no constraint, however brilliant, triggers learning.
That's the layer the RNP reading adds. Before you sculpt the landscape, check that the athlete can explore it. RNP identifies why the system can't explore; the constraints-led approach creates the conditions for it to explore. Taken separately, each has a blind spot. Together, they form a complete approach to motor learning.
Without reading the state of the system, a constraint can either educate or trap. It all depends on whether the system is available to explore the landscape it's offered.
In one sentence: the constraints-led approach drives learning by sculpting the context rather than dictating the movement, provided the system is available to explore and the constraints are targeted, dosed and read in light of its state.
Field test. Take a constraint you already use (surface, space, cue, target). Ask yourself three questions before you run it: is it aimed at a specific exploration, or am I changing for the sake of changing? Is the athlete available to explore, or in defensive mode? Is what I'm provoking structured exploration, or wandering? If you hesitate on even one, your constraint will trap, not educate.
A constraint teaches what an instruction explains, but only once you've checked that the system is able to hear it. That dual requirement is what I pass on in our training programs.
Scope: a functional, non-medical approach. It complements healthcare follow-up; it does not replace it.