Take our Do I Need Therapy?

Find a therapist near me

People actively test with others to see if their prior expectations still hold.

These tests often happen outside awareness.

Missed responses can confirm old beliefs.

Change depends on experiencing safety and disconfirmaion of old beliefs.

This post is the second in a two-part series. You can read Part 1 here: Why We Don’t Change—Even When We Know What’s Wrong

In my previous post, I discussed a central problem in psychological change: People often understand their patterns clearly, yet still find themselves repeating them. Insight, on its own, is rarely enough.

One influential framework—predictive processing—helps explain why. It suggests that the mind is constantly generating expectations based on past experience and tends to prioritise predictability over accuracy. This makes change difficult, even when new experiences contradict what we believe.

But this raises a deeper question: If these beliefs are stabilising—even when they are painful—why would anyone move toward situations that might challenge them? Why risk giving up a model of the world that, however limiting, is at least familiar?

Control–Mastery Theory (CMT), developed by Joseph Weiss and colleagues at the San Francisco Psychotherapy Research Group, offers an answer: People don’t just wait for change to happen. They actively move toward it. And they do so by testing.

Enter Control–Mastery Theory

But to see how, it’s worth tracing the two terms—control and mastery—back to Freud, where both originate.

“Control,” as Weiss uses it, draws on Freud’s later work, where the term referred to the ego’s capacity to gain command over its own mental functioning. Weiss develops this into a motivational idea: We are, both consciously and unconsciously, oriented toward gaining greater control over our own internal processes: our thoughts, feelings, and the defences that organise them.

This includes becoming aware of defences that were previously outside awareness, and integrating them into overall functioning. In this sense, control is not about suppressing internal experience. It is about increasing the mind’s capacity to know what it is doing, and to function with more flexibility.

Central to this is what Weiss called the safety principle. Movement toward greater awareness and integration only happens when the system registers the environment as safe enough. That sense of safety may come from changes in external reality, or in therapy from the therapist’s responses. Without it, existing defensive organisation remains in place, even when it is costly.

“Mastery” also comes from Freud, specifically from Beyond the Pleasure Principle, where repetition was understood as an attempt to gain mastery over overwhelming experience. Weiss retains this idea but shifts its focus. Rather than mastering drives, mastery in CMT refers to working through trauma and the beliefs that formed around it.

Take our Do I Need Therapy?

Find a therapist near me

In this sense, mastery is both a motivation and a capacity. It involves making sense of traumatic experience without reinforcing the beliefs that developed in response to it, regulating the emotions and bodily states associated with it, and integrating what had previously been warded off into a broader sense of self. What was once split off becomes something that can be thought about and experienced without overwhelming the system.

Again, this depends on safety. Traumatic material cannot be approached—let alone integrated—unless conditions allow for it. Without that, the system continues to rely on defensive patterns rather than engaging in the work of mastery.

Control and mastery are closely related but not identical. Control refers to gaining greater command over one’s mental functioning in the present. Mastery refers to working through what has been traumatic so that it no longer organises experience in the same way.

CMT, thus, addresses what predictive processing leaves open: not just how beliefs change, but why individuals are motivated to move toward experiences that make such change possible. Why would anyone seek a process that might destabilise them? Why risk giving up a model that, however painful, is at least familiar?

CMT suggests that people are actively, though often outside awareness, working toward greater control over their internal world and toward mastery of what has been traumatic.

The therapy room is a testing ground

What CMT adds, and what predictive processing alone cannot account for, is motivation. People don’t simply wait for corrective experiences; they move toward them, often indirectly. They test.

Patient testing refers to the process by which patients elicit responses from the therapist to find out whether their expectations will hold. It is, in effect, hypothesis testing within the relationship. These tests tend to take recognisable forms. For example, in transference testing, patients recreate familiar relational patterns to see whether they receive the same response. In passive-into-active testing, they reverse roles, treating the therapist as they themselves were once treated. Both are asking: Does this still happen here?

Take a patient who grew up asking for care and guidance from parents and being told, with impatience and irritation, to figure it out herself. She arrives in therapy self-contained and careful about not needing too much. One day she opens a session by saying she doesn’t know what to talk about, and asks whether the therapist has any suggestions.

A therapist who turns it back—“What would you like to explore?”—may be following good general practice. But given the patient's history, that response doesn’t land neutrally. It feels familiar: another moment where she is left to manage on her own, rather than being met where she is.

A responsive therapist would recognise that this is a test. The question is not simply what to talk about, but whether someone will step in when she reaches out. Instead of turning it back, the therapist would offer some guidance.

From a predictive processing perspective, this is a moment when a mismatch can register. From a CMT perspective, something else matters just as much: The patient set this up, even though probably unconsciously. She created the situation in which something new could happen.

Predictive processing gives us the mechanism—how belief updating happens—while CMT gives us the direction: why patients move toward the experiences that might allow it.

The therapist’s task, reconsidered

This also makes it worth taking failed tests seriously. If a therapist responds in a way that feels dismissive, inconsistent, or too close to what the patient already expects based on their past experiences, the result is not just a missed opportunity. It may function as confirmation.

When patients disengage suddenly, or pull back just as something seemed to be shifting, it is often worth asking whether a test was missed, and what that may have signalled to the patient about whether it felt safe, or meaningful, to continue.

Predicting to control, testing to master

Predictive processing and CMT are, in different ways, describing the same system. Predictive processing emphasises how the brain reduces uncertainty: predicting in order to maintain a sense of control of the world. CMT highlights something alongside that: testing experience in order to master what has been traumatic.

Bringing these perspectives together gives a more complete picture of what happens in therapy. People are not only trying to maintain stability by predicting the world around them. They are also, often outside awareness, creating situations in which those predictions can be tested.

People are not simply waiting to change. They are also trying to do something with their experience: to understand it, to integrate it, and, where possible, to move beyond it.

Predicting, in order to maintain control. Testing, in order to achieve mastery.

In therapy or in real life, both processes are happening at once.

Note: The author would like to thank Professor Francesco Gazzillo for his helpful clarification of the concepts of control and mastery; the “Enter Control–Mastery Theory” section draws substantially on his explanation. This piece builds on the collaborative work of the original journal article, and the author would also like to thank the co-authors: James McCollum, Josh Krieger, Sophia E. Winter, Daniel Duane, and George Silberschatz for their contributions.

Li, E., McCollum, J., Krieger, J., Winter, S. E., Duane, D., & Silberschatz, G. (2025). Predict to control, test to master: Integrating predictive processing and control–mastery theory in understanding how psychotherapy works. Journal of Psychotherapy Integration. Advance online publication. https://dx.doi.org/ 10.1037/int0000386

There was a problem adding your email address. Please try again.

By submitting your information you agree to the Psychology Today Terms & Conditions and Privacy Policy

© Psychology Today