The Cognitive Model

What it is

Epictetus, in the fifth chapter of the Enchiridion, wrote that people are disturbed not by things but by the views which they take of them. For eighteen centuries that was a consoling remark. In 1955 Albert Ellis began calling himself a rational therapist and took the line literally, as an instruction about what to do in a consulting room: if the view is what disturbs, then the view is the thing to work on, and it can be identified, written down, argued with and replaced. He quotes Epictetus directly, and Reason and Emotion in Psychotherapy, his major statement of the approach, appeared through Lyle Stuart in 1962 with the quotation at page 54.

Aaron Beck arrived at the same structure by a different route. Working at the University of Pennsylvania in the late 1950s, he set out to find empirical support for the psychoanalytic account of depression and found the data pointing somewhere else, towards a characteristic style of thinking rather than towards buried conflict. He published the result in Archives of General Psychiatry in 1963 as the first of two papers under the title "Thinking and Depression", naming the cognitive distortions the field still uses, and set out the full clinical model in Cognitive Therapy and the Emotional Disorders in 1976. Between them, Ellis and Beck give the twentieth century the same architecture twice: event, thought, feeling, with the middle term as the point of intervention.

The architecture has three moving parts and each carries a distinct claim. That a thought intervenes between an event and a feeling. That the thought is available to inspection, which is not obvious and not always true. And that changing it changes the feeling downstream. Only the third has been tested at any scale, because it is the one that turns into an outcome measure. Cognitive behavioural therapy built on this model is among the most heavily evaluated psychological treatments there is, and by the standards of the field it works.

Which is where the argument on this site begins, because a working therapy establishes something weaker than the doctrine it inherited. Showing that adjusting a thought changes a feeling does not show that thinking is what produces the suffering. A thermostat can be adjusted without being the reason the room is cold. The model earns its clinical keep as a lever. Whether the lever is attached to the cause is a separate question, and the design that would answer it, a trial that strips the treatment into components and tests them apart, has been run on cognitive therapy once, with an awkward result. See Dismantling Design.

The inheritance also runs further back than Epictetus. The Stoic route to tranquillity through judgment, and the Epicurean route through the management of desire, both relocate the problem out of the world and into the person standing in it. See Ataraxia. Ellis borrowed a line from that tradition, not evidence from it, and the therapy's results are not retrospective support for the philosophy.

In effect

Research: taking cognitive therapy apart (Jacobson and colleagues, 1996)

Neil Jacobson and colleagues randomly assigned 150 outpatients with major depression to one of three treatments: behavioural activation alone, behavioural activation plus work on automatic thoughts, or the full cognitive therapy package including work on core schemas. If the cognitive component is the active ingredient, the full package should have won. It did not. The behavioural component on its own performed as well as the complete treatment, and the result held at six-month follow-up. Published in the Journal of Consulting and Clinical Psychology in 1996, it is the study that put behavioural activation back on the map as a stand-alone treatment, and it is the clearest evidence inside the model's own literature that the part credited with the effect may not be the part producing it.

Everyday life: the move itself

Someone does not reply to a message. The feeling that follows arrives so fast that it seems to have come from the silence. The cognitive move is to slow the sequence down enough to find the sentence in the gap, usually something like "they have decided I am not worth answering", and then to ask what else would explain the same silence. Anyone who has done this knows two things about it. It often works, in the sense that the feeling loosens. And it does not feel like discovering the cause of the feeling so much as finding a handle on it. That ordinary experience is the thermostat point, felt from the inside.

Commerce: the model industrialised

England commissioned this model at national scale. The Improving Access to Psychological Therapies programme, piloted in Doncaster and Newham in 2008 and now called NHS Talking Therapies for anxiety and depression, was built on the economic case made by Richard Layard and the clinical case made by David M. Clark, and it delivers cognitive behavioural therapy through a trained workforce to very large numbers of people. The reason it could be commissioned that way is that the model is manualised: it specifies steps, which makes it trainable, auditable and costable in a way that most psychotherapy is not. That is a real achievement and it is also a selection effect worth naming. What gets deployed at scale is what can be written down, which is not necessarily what is doing the work.

What it does not say

It does not establish that thoughts cause suffering. Efficacy is not mechanism. A treatment can move an outcome reliably while the account of why it moves it is wrong, and Jacobson and colleagues' component analysis is the reason to take that possibility seriously in this particular case rather than as a general caution.

It does not validate the Buddhist or Epicurean claim it descends from. Ellis borrowed a sentence from Epictetus, not a result. The therapy's evidence base is evidence about the therapy.

It does not say that thoughts are always available to inspection. The model assumes the middle term can be found and named, and there is a substantial literature arguing that people have poor introspective access to the processes generating their own responses. See The Self-Report Problem.

It does not say cognitive behavioural therapy is the best available treatment, or the right one for any particular person. Comparisons with other active treatments are a live research question, effect sizes vary by condition and by outcome measure, and none of that is settled by the model being coherent.

Nothing on this page is clinical advice, and it is not a reason to start, stop or change any treatment. It is an argument about what a body of evidence establishes, which is a different question from what anyone should do about their own mental health. That is a conversation for a qualified clinician.

---