What CBT Gets Wrong About Change | iAST

A couple finishes twelve sessions of cognitive behavioral therapy. They have new communication skills, a list of cognitive distortions they can name on sight and a shared vocabulary for what goes wrong between them. They still fight every Sunday night about the same thing they fought about before session one.

The practitioner didn’t fail. The model reached its limit.

CBT’s theory of change locates the source of psychological problems in cognition. Distorted thinking produces distorted emotional responses. Fix the thinking, fix the problem. For presentations where the cognitive model fits (specific phobias, mild depression, performance anxiety), cognitive restructuring paired with behavioral activation produces measurable improvement.

The difficulty starts when the problem isn’t cognitive.

A parent can’t set limits with a teenager. A partner absorbs blame to avoid conflict. A client’s chronic headaches vanish during solo vacations and return within hours of coming home. In each case, the client’s thinking is downstream of the actual problem. The problem lives in the structure of an interaction: who does what, in response to whom, with what effect.

Cognitive restructuring asks the client to identify and replace the thought. The client replaces it. The pattern continues, because the thought was tracking reality. The parent’s belief that “my child won’t listen to me” is accurate. The partner’s conviction that “speaking up will make things worse” reflects how conflict functions in that specific household.

When a client’s cognition accurately reflects a structural reality, challenging the cognition is clinically imprecise. The client learns to think differently about a situation that hasn’t changed. That produces a person who understands their problem and still has it.

Strategic therapy reads these same presentations through a different question: what pattern of interaction is maintaining this problem?

The parent who can’t set limits exists within a triangle: the child, the other parent (or a grandparent, an ex-spouse) who undermines the limit, and the client. The child’s acting out keeps two adults focused on the child instead of on their own unresolved conflict. Restructuring the triangle resolves the case. Changing the parent’s cognition about the child doesn’t touch it.

A directive targets the structure. The practitioner assigns a specific behavioral task designed to alter the relational pattern. The parent is told to do something specific, at a specific time, in the presence of the person who usually undermines the limit. The directive doesn’t require the client to think differently first. Behavioral change produces a structural shift, and the shift produces new cognition as a downstream effect.

This is the distinction CBT’s model misses. In structural problems, the sequence runs from behavior to structure to cognition. Reverse it and you get insight without resolution.

CBT’s theory of change is incomplete for a specific category of problems. The cognitive model fits cognitive problems. Structural problems, where the cognition tracks an interaction pattern without generating it, sit outside that model’s range.

The 26.2% average dropout rate from CBT¹ and the finding that the most common reason clients give for leaving is “desiring alternative treatment” suggest that a significant portion of clients encounter this limit in practice. They know something isn’t working. They often can’t articulate what. The answer is usually that their problem lives in a pattern.

For the practitioner, the question is practical. When cognitive restructuring stalls, when the client does the homework and nothing shifts, what are you looking at? The client’s thoughts, or the structure their thoughts describe?