The Point Is to Change Things

Philosophers have merely interpreted the world in various ways, the point is to change it.

—Karl Marx (1818–1883), Theses on Feuerbach, edited and published posthumously by Friedrich Engels

I am grateful for the opportunity to respond to Mullen’s review of my book Madness Explained: Psychosis and Human Nature. As Mullen notes, my approach arises from a tradition that he describes as Anglo-empiricism, although I would prefer to say that it is inspired by a scientific, realist worldview. I believe that there are facts about psychiatric disorders that are waiting to be discovered; it is just that it is not always easy to discover them. I also believe that we can distinguish, on the basis of evidence, between theories that are poor and misleading and those that are better supported by the data and more helpful, although making this distinction is often difficult. It was in this spirit that I tried to examine the psychiatric literature with fresh eyes. The account of severe mental illness which emerged differs from conventional thinking in several important ways and is, I believe, more rational. In his critique, Mullen somehow fails to summarize or even critique my thesis so, for the benefit of potential readers, I summarize it herein.

My starting point is the observation that little progress has been made in the understanding and treatment of severe mental illness since the end of the nineteenth century (an argument I develop further in my more recent book Doctoring the Mind: Why Psychiatric Treatments Fail [Bentall 2009], in which I compare the modest advances in psychiatric care with the dramatic leaps that have been made in the treatment of cancer and heart disease). It seems unlikely that this regrettable state of affairs reflects a lack of resources—billions of dollars have been spent on psychiatric genetics and antipsychotic drug development with very little to show for it. Rather, it seems more likely that psychiatry has suffered from a poverty of ideas.

I trace contemporary approaches to psychosis to nineteenth century Germany and, in particular, the work of Emil Kraepelin, whose ideas influenced the American psychiatrists who designed the DSM diagnostic system. This system assumes that there is a clear dividing line between sanity and madness, and that psychiatric disorders fall into discrete categories, for example, schizophrenia and bipolar disorder. Most psychiatrists and psychologists have also assumed that the psychoses are diseases of the brain in much the same way that diabetes is a disease of the pancreas. Unfortunately, even those who are critical of this way of thinking often underestimate its harmful consequences. By far the majority of psychiatric research compares patients with one diagnosis, say schizophrenia, with individuals with no diagnosis (healthy controls) and perhaps another group of individuals with another diagnosis (say patients with depression). This strategy cannot hope to succeed unless the patients with the target diagnosis (schizophrenia in this example) have something in common of etiological significance, which the control participants lack. Examination of all of the available evidence shows that this condition is rarely, if ever, met. The diagnosis of schizophrenia, for example, does not consist of a meaningful cluster of symptoms but at least three independent clusters (positive symptoms, negative symptoms, and cognitive disorganization). Phenomenological, epidemiological, and genetic evidence suggests that, far from being separable, schizophrenia and bipolar disorder are overlapping constructs. Nor is there a clear dividing line between psychosis and normal functioning. In fact, on every possible test, psychiatric diagnoses are not fit for purpose; they do not even predict which kinds of treatments are likely to be effective for which patients. Research using diagnoses as an independent variable will therefore inevitably implicate many variables in the causation of psychiatric disorder but none firmly, which is exactly what we find. It is for this reason that I compare psychiatric diagnoses with star signs (another prescientific system of classification that is widely but erroneously held to describe individuals efficiently and to predict what will happen to them).

So what is to be done? As we obviously require some principle for describing individuals for the purposes of research we...