You've done it, even if you'd never call it that. You showed up ten minutes late to see whether they'd be annoyed. You said something deliberately unlikeable and watched their face. You went quiet for a whole session to find out if they'd fill the silence or let you drown in it. You mentioned, too casually, that you'd been thinking about quitting — and then studied them for a flicker of relief.

You weren't being difficult. You were running an experiment. And the hypothesis you were testing wasn't about your therapist at all. It was about you: that you're too much, that people leave, that anyone who gets close enough will eventually want out. You built that hypothesis a long time ago, in a house where it might have been true. Now some part of you keeps checking whether it's still the law of the universe or just an old weather report.

The belief you're trying to disprove

In the 1970s and '80s, the psychoanalyst Joseph Weiss, working with Harold Sampson and colleagues at what became the San Francisco Psychotherapy Research Group, proposed something that broke with a lot of the thinking of his era. The prevailing view held that patients were driven by impulses they couldn't control, and that they resisted change because some part of them wanted to stay sick. Weiss listened to hundreds of hours of session recordings and heard something else entirely.

He heard people trying to get better.

The theory he built — Control-Mastery Theory — starts from the idea that we develop what he called pathogenic beliefs: convictions formed in childhood from real experience, usually experiences we had no power over. If I need too much, I'll be abandoned. If I'm angry, I'll destroy someone. If I succeed, I'll leave people behind and they'll hate me for it. My feelings are a burden. These beliefs weren't stupid. They were the best inference a small person could draw from the available evidence. If your mother's mood collapsed whenever you were upset, then my sadness hurts people was a reasonable conclusion. It kept you safe.

The crucial move in Weiss's thinking is this: you didn't just absorb that belief. You've been trying to escape it ever since. He argued that people arrive in therapy with an unconscious plan — not a strategy they could articulate, but a direction — to disconfirm the beliefs that are running their life. And the way you go about disconfirming them is by testing.

What a test actually looks like

A test is a small, usually invisible experiment in which you create the conditions of the old situation and wait to see if the old outcome repeats.

Weiss and Sampson described two basic shapes. In the first, you play yourself. You take the role you had as a child and see whether the therapist takes the role your parent had. You need something, and you make the need visible, and you watch. You disagree with an interpretation and brace for the withdrawal of warmth. You bring in something shameful and wait to be judged. The child who learned that asking was greedy will, twenty years later, ask for a schedule change with her whole body tensed against the no.

In the second shape — which Weiss called turning passive into active — you flip it. You do to the therapist what was once done to you, and you watch how they handle it. If you grew up with someone who was cold and unreachable, you may find yourself unaccountably cold in session, watching to see whether your therapist survives the cold, whether they get anxious or wounded the way you did. You are not being cruel. You are asking, from the safety of the other chair: was this survivable? Did I fail to survive it because it was unsurvivable, or because I was seven?

That is the strange and moving logic of it. You hand the therapist your childhood, and you watch them try to hold it.

Passing, failing, and what happens next

What makes this more than an elegant story is that it was studied. George Silberschatz and colleagues in the same research group developed methods for identifying tests in session transcripts and then examining what happened in the minutes that followed. When the therapist responded in a way that disconfirmed the patient's pathogenic belief — held steady, didn't retaliate, didn't collapse, didn't confirm the fear — patients tended to become measurably bolder afterward. They spoke with less anxiety. They went deeper. They brought up material they'd been circling for weeks. When the belief was confirmed instead, the opposite: the conversation flattened, the patient retreated to safer ground.

The therapist rarely knows a test is happening in the moment. Neither do you. What you both notice is the aftermath — that the session cracked open, or that it went strangely dead.

This is why the accusation people make against themselves — I'm self-sabotaging, I'm pushing them away, I'm being manipulative — is almost exactly backwards. Testing is not sabotage. It's the opposite of sabotage. It's your mind refusing to accept the verdict it was handed as a child, and gathering fresh evidence, at some risk to itself, in the only laboratory available.

Why it matters outside the room

Here's the part that should make you sit up: your therapist is not the only person you test.

You test the partner who says they love you, by being slightly worse than usual and seeing if it holds. You test the friend who says they want to hear about it, by giving them a small piece of the hard thing and reading their face for boredom. You test the boss who says the door is open. Most of us are running these experiments constantly and never once registering the data, because when someone passes we barely notice, and when someone fails we say see — as if we hadn't designed the trial ourselves.

The difference in therapy is that the experiment is happening somewhere it can be looked at. Someone is paying attention. And crucially, the person you're testing is trained not to take the bait — which means, for perhaps the first time, the old outcome doesn't repeat, and the belief that has been quietly organizing your life gets a genuine counterexample.

But a counterexample only counts if you catch it.

Your next moves

  • Name one pathogenic belief in a single sentence. Not a diagnosis — a sentence, in the voice of a child. If I need too much, people leave. If I'm angry, I'm dangerous. My feelings ruin things. Write it down. Most people find it takes three or four attempts before one sentence makes their chest tighten. That's the one.
  • After your next session, ask: did I set a trap this week? Look for the small things — the arriving late, the flat tone, the throwaway maybe I don't need to come anymore. Write down what you did and what you were secretly watching for. You're building a record of your own experiments.
  • Track what happened right after. Note whether the ten minutes following that moment went deeper or shallower. Boldness after a test is the tell. If you opened up, they passed — and your belief just took a hit you probably didn't register.
  • Bring the test itself into the room. Try, out loud: I think I was waiting to see if you'd get sick of me. This is one of the most useful sentences in therapy, and almost nobody says it. You don't have to be sure it's true. Saying it turns a hidden experiment into a shared one.
  • Find the same test somewhere else this week. One relationship outside therapy. Notice the moment you flinch and pre-emptively disappoint someone. Then don't. See what happens.

The record is the point

The trouble with tests is that they're only legible across time. One late arrival is a late arrival; the same flinch showing up in session after session, always right before you were about to say something real — that's a pattern, and patterns are where the belief lives. But almost nobody remembers session to session what they were secretly checking for. The insight lands, the week eats it, and the experiment runs again next Tuesday with no results recorded.

Sesh exists for that gap. It's a private place to write down what you tested, what your therapist did, and what you did next — so that the counterexamples accumulate instead of evaporating, and so you walk into your next session already knowing what you're really asking. What happened in therapy shouldn't stay in therapy. If you'd like somewhere to keep the evidence, Sesh is here.