FROM THE BLOG: The Smallest Change and the Predictive Mind: Why Did Erickson’s Methods Work?
I became interested in the systems-theoretical idea of the smallest change in the 1980s. The idea immediately struck me as significant: if a system is in balance, it does not necessarily have to be changed by force. Sometimes a small, precisely placed change is enough to force the whole system to reorganize itself. Around the same time, a handout arrived at my workplace, and Milton H. Erickson was mentioned in its references. I began to investigate how he was connected to this way of thinking, and that was really where my interest in him began.
I have never been especially interested in Erickson as a person or as a therapeutic legend. I have been much more interested in the deep structure of his methods and language. What was happening in his speech? Why could a seemingly small, strange, or indirect instruction change a client's experience? Why could amplifying a symptom, accepting resistance, or moving the problem into a new form produce change where a direct attempt at change had failed?
Decades of theoretical study and later professional immersion in the world of hypnosis gradually led me to see recurring structures behind different schools of therapy, NLP, Erickson's methods, suggestions, and interactional phenomena. Different approaches used different language, but often they seemed to be dealing with the same basic phenomenon: how human experience organizes itself, how it defends itself, and how it can be changed. At that point, the predictive mind model began to function for me as a kind of Swiss Army knife. I now use it to examine therapies in general and hypnosis in particular — also when I follow the most popular hypnosis YouTubers and try to distinguish, beneath the surface technique, what is really happening in the interaction. The same model also helps in ordinary everyday conversations, where people are constantly responding to one another's expectations, roles, and small shifts in direction.
Today I would describe this from the perspective of the predictive mind. A human being does not merely react to what is. We are constantly living on the basis of what we anticipate will happen. A person has a model of themselves, their body, their limits, their possibilities, their failures, and what they might become. A problem is not merely a symptom or a thought. It is an entire predictive construction in which certain weightings guide perception, emotion, bodily readiness, and action.
In therapy, a contradiction always arises between what is and what the person wants to become. The client wants to be calm, but their system predicts restlessness. They want to be confident, but their body has learned to prepare for failure. They want to be free of the symptom, but the predictive model built around the symptom keeps returning them to the same position. Desire alone is therefore not enough. It collides with the weightings that maintain the problem.
This is where imagery work becomes interesting. The NLP-style method of producing an image of the desired self in an imagined situation can be seen as the deliberate construction of such a contradiction. The person sees themselves in a situation as different from how they currently experience themselves: calmer, freer, more confident, more flexible, more capable of acting. This is not merely "positive thinking" if it is done in a sufficiently vivid experiential way. It produces a cognitive and bodily contradiction between the current self-image and a possible self-image.
When this is repeated, the contradiction can grow. The mind cannot indefinitely maintain two completely incompatible models without one of them beginning to change. In the predictive mind framework, this means that the old weighting model is forced to correct itself. If a person has lived, in their imagination, enough times through a situation in which they act differently, feel differently, and are seen differently by others, this does not remain mere imagination. It begins to compete with the old predictive model.
The problem, however, is that moving directly toward the goal can activate resistant weightings. If a person has spent years trying to be brave, calm, or free, that very goal may already be part of the problem's structure. "I must change" can be a sentence that activates the prediction of failure. In that case, imagery work itself can become another performance: I should be able to see myself succeeding, but I cannot even manage that.
This is where Erickson's logic becomes especially interesting. He did not always move directly toward the goal. He could move in the direction of the symptom. He could amplify it, limit it, time it, give it a task, or change its place within the system. On the surface this looks strange: why would a therapist strengthen the very thing the client wants to get rid of? But at the level of the system, this may be precisely the smallest decisive change.
If the problem construction is an old predictive model, it does not necessarily have to be refuted with a direct prediction of success. Often, a smaller deviation is enough: a detail, observation, or image that no longer fits the model's previous structure. It does not yet have to correspond to the goal. It only has to unsettle the old model's self-evidence precisely enough for the mind to begin reorganizing the experience.
If a symptom is deliberately amplified, it is no longer the same symptom. If it is asked to occur at a specific time, it is no longer uncontrollable in the same way. If the client is given permission not to change, not changing is no longer the same compulsion. If resistance is accepted and brought into the work, it loses its opposing force. If fear is asked to increase, it may shift from an experience into a task, and becoming a task changes its position within the whole system.
This is the essential difference between direct correction and systemic disruption. Direct correction says: "Be different." Systemic disruption says: "Let us make a small change in how this same problem is organized." The first may collide with already activated resistant weightings. The latter may bypass them, because the system does not recognize it as the usual attempt at change.
From the perspective of submodalities, the same phenomenon can be described as changing the structure of an image. An image, memory, or inner representation linked to a problem is not a neutral picture. It has size, distance, brightness, movement, sound, bodily feeling, point of view, and emotional weight. When a contradiction is introduced into any of these, the meaning of the whole representation may change. The problem does not necessarily have to be "removed." It may be enough that its inner structure no longer remains a pure version of the old recording.
The "contamination" is a descriptive word, even though it sounds ugly. An old problem construction is often strong precisely because it is internally coherent. It tells the person the same story again and again: I am like this, this always happens to me, there is no way out of this. If something is added to it that does not belong to the story, the structure begins to leak. The addition may be an image pointing toward the goal, but it may also be paradoxical, exaggerating, playful, or temporally limiting.
What matters is that the addition does not activate the same old rejection. If a person resists the idea "I am calm," it may not be useful to push them immediately toward an image of calmness. But if they are asked to make the restlessness in their mind absurdly large, place it somewhere strange, give it a voice, or assign it a specific time of appearance, the old structure has to deal with something for which it has no ready-made defense. Then change may happen from the side.
In this sense, Erickson's methods were not random oddities. They were often very small but precise disruptions in a system that was holding itself together. He did not always try to change the client's content; he changed the client's relationship to the content. He did not always remove the symptom; he changed the symptom's position. He did not always answer the client's expectation; he joined it so precisely that he could turn its direction from within.
This resembles the systems-theoretical idea of the smallest change. When a system is in homeostasis, everything in it tends to restore the old balance. If the change is too large, the system defends itself. If it is too familiar, it is absorbed into the old structure. But if the change hits the right point, even a small difference may be enough. The old balance does not collapse chaotically; it shifts into a new position.
In therapy, this means that change does not always arise from a great insight. It can arise from a small addition that makes the old model impossible to preserve completely unchanged. A new feature is added to an image. A new task is given to a symptom. Resistance is brought into the material of the work. Failure is made temporarily intentional. The desired self is seen in an imagined situation so many times that it begins to demand a place beside the old self-image.
In all of these, the same structure is present: the current model comes into contradiction with a new experience. If the contradiction is too great, it is rejected. If it is too small, it has no effect. If it is precisely placed, the system begins to correct itself.
For this reason, therapy can be seen as the art of unsettling beliefs. The client's idea of themselves, their problem, and their future does not change merely because they are told a new truth. It changes when the old predictive model encounters an experience that it can no longer fully explain on its old terms. This may be a new image, a new bodily experience, a new interactional arrangement, or a paradoxical task.
Erickson's language interests me precisely for this reason. He did not use language merely to transmit information. His language did things. It built expectations, shifted weightings, changed roles, gave permission, created uncertainty, limited symptoms, and opened new routes without the client always noticing that they were in a direct attempt at change. The deep structure of his language was often systemic: it did not attack the problem, but changed its conditions.
From this perspective, NLP imagery work, changing submodalities, Erickson's paradoxical tasks, and the theory of the predictive mind are not separate things. They all describe the same basic phenomenon in different languages. Human experience is held together by predictive models. A problem is one such model that holds itself together. Change occurs when a difference is introduced into the model that it can no longer absorb into itself without correction.
In the end, the decisive question is not whether the intervention is large or small, direct or indirect, NLP, hypnosis, or therapy. What matters is whether it produces a living contradiction at the point where the problem maintains itself. If it does, even a small addition may be enough. If it does not, even a large explanation remains only another layer around the old problem.
Perhaps this is why Erickson stayed in my mind already in the 1980s. There was something in him that connected directly with the systems-theoretical idea of the smallest change. He did not seem to be looking for the largest possible intervention. He looked for the point where the client's system was already moving, stuck, or in contradiction with itself — and added something that changed the whole arrangement.
That is still, to me, the most interesting aspect of his work. Not the legend, not the personality, not the guru. The deep structure of the methods: how a small, precisely placed linguistic or experiential change can force an old predictive model to correct itself.