A sprinter places their foot on the ground for one hundred milliseconds. By the time the information 'the foot has touched' reaches the brain, the contact is already over. So how do they manage to run accurately?
Published on July 3, 2026
Hello to you, the movement professional,
Here is a question that seems trivial and dismantles half of the shortcuts we hear about movement. If the brain received its information and then sent its commands in real-time, no quick movement would be possible.
The feedback from a proprioceptive sensor takes one hundred to two hundred milliseconds to be processed, and a sprint step lasts about one hundred milliseconds. At the moment the foot touches, the nervous system does not yet know it has touched.
Therefore, movement is not commanded, it is predicted. And this idea, once grasped, changes the way we read a moving body. This is the whole purpose of this foundational article.
The sensorimotor loop is the basic unit of Neuro-Postural Reprogramming. Far from being a metaphor, it is a complete and measurable cycle through which a movement occurs.
It can be described in five stages. A sensory perception opens the cycle, the input. The nervous system then pools what it receives, which is central integration. Then comes the decision, the choice of a strategy. Then execution, when the muscles and joints obey. And the feedback, the reafference, which confirms or contradicts what was predicted.
This last stage is the most important, and it is the one most approaches forget. The loop has no end. The output of each gesture becomes an input for the next. The body does not just execute; it constantly compares what it predicted to what it obtained, and it recalibrates. Rather than a command descending, a movement is a conversation that quickly circles between what the body expects and what it receives. The question remains: who, in this loop, holds the pen?
Here is the distinction that organizes everything else. We readily imagine a brain that sends signals to the muscles like typing commands on a keyboard. The reality is more elegant and better documented.
Neuroscientist Karl Friston summarized it in 2011 in the journal Neuron: motor control does not function like a pilot correcting second by second, but like a system that anticipates the future state of the body and minimizes the gap with what it perceives.
Concretely, the nervous system builds an internal model, a kind of simulation of what will happen.
It deduces the sensations it should receive if the gesture unfolds as expected.
Then it compares this prediction to the actual feedback. The gap between the two, the prediction error, is the true fuel of motor learning.
The body progresses by reducing this gap, not by repeating for the sake of repetition. This is why the postural adjustments that precede a gesture are anticipated and not reactive: they are part of the prediction, not the correction. And this is also why working on the output without touching what predicts it amounts to correcting the same mistake indefinitely.
If movement is predicted, then the moment that truly counts occurs before the muscle moves. The motor decision is this act by which the nervous system chooses, in the one hundred to three hundred milliseconds preceding a gesture, the strategy best suited to the available sensory context.
It is made upstream of the muscle, based on what the body sees, feels, and perceives of its position in space.
This choice is prereflective. It precedes the awareness of having chosen. As early as 1983, the work of Benjamin Libet showed that preparatory brain activity precedes the feeling of deciding by several hundred milliseconds.
In other words, when an athlete "decides" to start, their nervous system has already made the decision. For the professional, this window is a goldmine.
In the two or three tenths of a second preceding a step, one can observe a micro-tilt of the pelvis, a pre-positioning of the shoulder, an orientation of the gaze, a tensioning of the deep back muscles. To the trained eye, this is visible. In frame-by-frame video, it can be measured. The decision is a readable signal, not a mystery, and it is the first place to intervene.
A precise mechanism makes this prediction possible. When the nervous system sends a command to the muscles, it simultaneously emits an internal copy, directed towards the structures that integrate the movement, notably the cerebellum. Physiologists Erich von Holst and Horst Mittelstaedt formalized it in 1950 under the name of the reafference principle, and it is also known as the efference copy. This copy is used to announce in advance the expected sensations of the gesture.
The most telling example lies in an everyday experience: no one can tickle themselves. The reason is mechanical. The brain predicts the sensation of its own hand and cancels it before it arrives because the efference copy had announced it.
An external hand, however, was not anticipated, so the surprise gets through. This small fact says something significant: the body constantly distinguishes what it has produced from what happens to it. When this sorting malfunctions, the gesture becomes costly and poorly calibrated. And to read this malfunction, one must first know on which body map the nervous system relies.
The nervous system does not control an abstract body; it controls a representation of this body. There are two, and mixing them leads to reading errors. The body schema is an unconscious map, dedicated to action, continuously updated by proprioception, touch, vestibular input, and vision.
It is what allows you to bring your hand to your mouth in the dark without thinking.
Its scientific lineage is old and solid, from Head and Holmes in 1911 to the systematization work of Shaun Gallagher, with the double dissociation demonstrated by Jacques Paillard in patients deprived of sensations.
The body image is something else. It is the conscious, perceived representation, charged with emotion, that one has of their own body. An athlete can have a very fine body schema, remarkable coordination, and a damaged body image, made of anxiety or rejection. The reverse also exists.
For the movement professional, the distinction is a safeguard. The body schema is worked on through sensory input and gesture. The body image, when it is suffering, requires cooperation with other professions. Confusing the two is to promise on a field that is not one's own.
Not everything is slow prediction. The body also has a short circuit. The reflex arc passes through the spinal cord, in two or three relays, with a latency of ten to twenty milliseconds. It is the patellar reflex, for example, the stereotyped, non-negotiable, non-contextual response.
The long loop, on the other hand, involves the brainstem, cerebellum, thalamus, and cortex, with a latency of one hundred to three hundred milliseconds, and a response adapted to the context. Charles Sherrington, as early as 1906, had already laid the foundations for this reading by arcs and loops.
Functional movement never uses one without the other. It makes the short, fast, and automatic circuit dialogue with the long loop, slower but intelligent. And it is precisely at the junction of the two that the first major calibration of the system occurs. This calibration has a name you already know.
Before knowing how to make fine decisions, the infant has ready-made motor programs. The primitive reflexes are these automatic responses of early childhood that serve for a time, then must integrate, meaning come under the control of the higher levels of the nervous system.
These are the first building blocks of the loop, the way the body learns to link a sensory input to an organized motor response.
When one of these reflexes does not fully integrate, it does not disappear; it becomes noise in central integration. Take the asymmetric tonic neck reflex, the ATNR. It is triggered by head rotation and is initially proprioceptive at the neck level, with the vestibular system accompanying the movement, as turning the head involves both cervical receptors and the labyrinth.
Residual in adults, it can disrupt coordination between the trunk and limbs as soon as the head turns. The observation goes beyond laboratory curiosity. A 2024 study published in Frontiers in Sports and Active Living reports a residual ATNR in a large majority of high-level athletes.
The correct reading is not "this reflex causes the problem," but "this reflex is a variable that weighs on the motor decision, among others."
Separating the trigger from what activates at the same time is what distinguishes serious reading from a shortcut.
A loop is only as good as the quality of its inputs. Postural control relies on three major informants, constantly articulated.
Vision provides information on the vertical, the horizon, and the flow of the passing scenery. The vestibular system, located in the inner ear, captures accelerations and orientation relative to gravity. Somesthesia, finally, includes the deep proprioception of muscles, tendons, and joints, and the sensitivity of the skin, especially under the soles of the feet.
This three-system grid is that of contemporary balance research, from Horak and Macpherson in the 1990s, and it has replaced the old models with five hierarchical sensors.
The key point is that the nervous system does not add them up; it weights them according to the context.
In full light on stable ground, vision takes the weight. On a moving surface, the vestibular system takes over. Eyes closed in an open field, somesthesia becomes king. Reading an imbalance often means identifying which system has become unreliable and which the body defaults to. You don't strengthen a muscle to fix this. You give the right sensor clear information.
Everything that precedes leads to a method, not a theory. The grid of Neuro-Postural Reprogramming is summed up in one sentence: we read an output, trace back to the input that feeds it, and recalibrate the phase of the loop that fails.
A gesture that is corrected ten times and returns ten times betrays a decision that has never been addressed, not bad will. Recalibrate the decision, and the gesture reorganizes itself.
This requires a work order. The majority of the result comes from the accuracy of the reading, a smaller part comes from the precision of the reprogramming.
In other words, read before intervening, reprogram before strengthening. And follow the order of system construction, from the deepest to the finest, without skipping steps. As long as the sensory base is not reliable, nothing stable is established above. It is a gardener's logic, not a mechanic's.
We must be honest about the edges of the map. The predictive model of movement is well-supported, but its fine neural implementation remains a field of research, not a closed fact.
The efference copy has been established since 1950; the details of the circuits that carry it are still debated. Similarly, a visible residual reflex is never, by itself, the cause of underperformance: it is a variable among others, to be crossed with the rest, never a sole culprit.
Anyone who sells you reflex integration or sensory work as a guarantee of results has left the field of rigor.
Here is a way to ensure we remain honest, a prediction that could contradict us.
If we correctly recalibrate a failing sensory input of an athlete, their motor output should improve without having worked on the gesture itself.
If the gesture does not change while the input has been made clear and reliable, the reading was wrong, and it needs to be redone. The loop is our safeguard as much as our tool.
👉 I want to learn to read the loop upstream, not to correct symptoms: labo-rnp.com/fr/pros
It is the complete cycle through which a movement occurs: sensory input, central integration, motor decision, muscular execution, then sensory feedback that confirms or corrects. The output of each gesture becomes the input for the next, allowing the nervous system to continuously recalibrate.
No. The processing delay of sensory feedback, from one hundred to two hundred milliseconds, is too long to guide a rapid gesture second by second. The nervous system anticipates the body's future state and compares this prediction to the actual feedback. The movement is predicted, then corrected, not commanded live.
It is the choice, made one hundred to three hundred milliseconds before the gesture, of the postural and motor strategy best suited to the sensory context. This act is prereflective, it precedes the awareness of deciding, and it is readable in the micro-adjustments that precede the movement.
The body schema is an unconscious map of the body dedicated to action, updated by the senses. The body image is the conscious and emotional representation one has of their body. One can have one intact and the other disturbed. The first is worked on through sensory input, the second often involves cooperation with other professionals.
They read the output, the gesture or posture, trace back to the sensory input that nourishes it, and recalibrate the phase of the loop that is failing, instead of correcting the visible symptom. The quality of the reading matters more than the power of the exercise.
Romain Katchavenda and the LabO RNP team
Discover how the brain anticipates movement according to Berthoz. A new perspective on perception and action that changes everything.
Discover how motor decision influences sports performance, integrating strategy and perception in 280 milliseconds.
