Does ‘mental illness’ exist? I have taken as my title one of my least favourite questions. Although often posed to critics of psychiatric practice like myself, it actually makes very little sense. In unpicking it, I hope to show that we have better ways forward than the current, largely unchallenged understandings of emotional distress which do not reflect reality – either in terms of the evidence, or in terms of people’s lives.

The question really needs re-phrasing in two parts. If we framed the first part as ‘Do people really experience extreme forms of distress such as suicidal despair, hearing hostile voices, crippling anxiety and mood swings?’ then of course the answer is yes. As a clinical psychologist who has worked in the field of mental health for over three decades, and as a human being who is not immune from distress myself, I know this very well. But my answer to the implied second part ‘Are these experience best understood as “mental illnesses”?’ is a definite no. The concept of ‘mental illness’ obviously exists, as do the concepts of witches, ghosts and God – but the idea that the very real experiences subsumed under this term are best explained as medical disorders does not have, and has never had, any evidence to support it.

The concept of ‘mental illness’ obviously exists, as do the concepts of witches, ghosts and God – but the idea that the very real experiences subsumed under this term are best explained as medical disorders has never had any evidence to support it.

This is likely to come as a surprise to most members of the general public. We can approach the debate by asking whether the psychiatric labels that are applied to people are proper scientific categories. Any science – in this case medicine - needs to be able to demonstrate that it is based on a reliable and valid classification system, in order to develop testable hypotheses and hence the general laws that constitute a body of scientific knowledge. There is no issue of greater importance or greater controversy in mental health, since if this cannot be established, the whole model breaks down and all psychiatry’s other functions – indicating treatment, research and so on - will be fundamentally undermined. In the words of Peter Breggin, psychiatry would then become ‘something that is very hard to justify or defend – a medical specialty that does not treat medical illnesses.’

‘Reliability’ describes the likelihood that when faced with the same patient and an agreed list of criteria, clinicians will come up with the same diagnosis. In relation to psychiatric diagnosis, reliability is extremely low – which is one of the reasons why people typically collect a whole list of labels in their journey through the mental health system. But an even more important issue is the validity of such categories. There are different meanings of validity, but essentially it is about whether the categories actually describe something in the real world. Are there, for example, established patterns of chemical imbalances, genetic flaws or other bodily malfunctions which correspond to these labels, and are causally linked to the ‘symptoms’, as we would expect if we were diagnosing pneumonia, or kidney failure, or breast cancer?

So-called ‘symptoms’ are not examples of bodily dysfunction, but a ragbag of social judgements about people’s thoughts, feelings and behaviour.