Cognitive behavioural therapy (CBT) is one of the most popular forms of talking therapy. It is the treatment of choice for depression and anxiety and is a staple of public healthcare systems, such as the NHS and the Australian Medicare system. CBT’s understanding of mental illness and therapeutic techniques is already part of the mainstream – accusations of “catastrophising” and pleas to “reality check” beliefs can be found everywhere. As a Washington Post article put it: “For better or worse, cognitive therapy is fast becoming what people mean when they say they are ‘getting therapy’.”
One of the reasons for CBT’s runaway success is the reams of evidence that support its effectiveness at treating a host of mental health disorders. Although there is evidence for CBT’s effectiveness, the evidence for its theory, particularly its understanding of mental illness, is far more mixed. To put it another way, we know that CBT works, but we are not sure how or why it works.
CBT’s cognitive model of mental illness, originally developed by Aaron Beck in the 1960s, hypothesised that disorders such as depression were characterised by certain patterns of thought that give rise to the negative emotions and behaviour typical of mental illness. These patterns of thought are referred to as “cognitive distortions” or “negative automatic thoughts”.
But what exactly is wrong with these thoughts? What makes them “distorted”? Generally, vague answers are offered in response. For example, the American Psychological Association describes these thoughts as being “faulty” or “unhelpful”. Looking at lists of distortions, offers clues.
Most distortions focus on faulty reasoning, where someone “jumps to conclusions”, makes a poor inference (“overgeneralising”), is biased in how they perceive a situation (“black or white thinking”), or, more straightforwardly, when they believe something false or inaccurate. CBT then goes on to suggest that if this faulty reasoning was resolved, the “unhelpful” negative emotions and behaviour will change.
Three reasons to doubt the model
There are three reasons to doubt the cognitive model and the association of mental illnesses with errors in reasoning.
First, the sort of issues CBT draws attention to – bias, false beliefs, poor inferences – are all relatively common, even in mentally healthy people. As a great deal of psychological research has shown, we are all prone to poor reasoning. And even with mental disorders that seem to involve obvious faulty thinking, such as schizophrenia or psychosis, it is very difficult to nail down the difference between a delusion and a strange belief. For example, what distinguishes delusions from the sorts of beliefs associated with conspiracy theories or belief in the supernatural? “Faulty” thinking does not obviously correlate with mental illness.
Second, although CBT researchers have studies showing that mental disorder has something to do with cognitive distortions, there is a problem with the tests or measures used in this research. Many of these tests ask questions that have nothing to do with poor reasoning. They often ask people to answer questions that are simply about how they feel (“I’m so disappointed in myself”, Automatic Thoughts Questionnaire), need a great deal more information, maybe about population-level data to answer (“I do few things as well as others”, Inventory of Cognitive Distortions), or seem to be about moral or practical issues rather than poor reasoning (“Taking even a small risk is foolish because the loss is likely to be a disaster”, “To be a good, moral, worthwhile, person, I must help everyone who needs it”, Dysfunctional Attitude Scale).
Finally, there is research suggesting that it is mental health rather than mental illness that is related to poor reasoning. The “depressive realism hypothesis”, shows that depressed people more accurately: predict how much control they have over outcomes, evaluate their performance and recall feedback.
Mentally healthy people, on the other hand, succumb to an “illusion of control” and tend to recall their own performance and feedback in an excessively rosy light. Although most of this research has been on depression, there are studies suggesting that schizophrenia may be associated with better theoretical reasoning and autism is sometimes characterised by enhanced logical and theoretical reasoning.
Not backed by research
Not only is there contrary evidence showing problems with reasoning are widespread as well as potentially associated with mental health rather than mental disorder. But the evidence in favour of CBT’s take on mental illness is tainted because the tests used in these studies do not even track problems with reasoning. CBT provides a compelling story about mental illness – mental illness is associated with “faulty” reasoning, and in resolving this, negative behaviour and emotions are addressed. Unfortunately, research doesn’t quite back up this story.
We might wonder whether it matters. After all, CBT seems to work, so why should we care how it works or whether it is wrong in its story about mental illness?
It matters ethically. It is one thing to point out that certain patterns of thinking are “unhelpful” or bring about negative emotions and behaviour, quite another to suggest that someone is irrational or reasoning poorly when the evidence for this is shaky. It is what the philosopher Miranda Fricker terms “epistemic injustice”, where a member of a disenfranchised group (that is, the mentally ill), is told their claims are plagued by errors or cannot be taken at face value. Even worse, with CBT they are told this when they come seeking help. Troubling, at best, unethical at worst.