There is abundant evidence of brain disease causing what is defined as "mental illness"

The concept of causation is tricky even in the physical sciences. David Hume, for instance, denied that there can be such a thing as evidence of causation. He was wrong about that (because he was wrong about what evidence is), and indeed the existence of causation is essential to every scientific explanation. But it is tricky to define, and trickier to pin down evidence of causation. When it comes to explanations of anything involving human opinions and decisions, it becomes even trickier, but is equally essential. I think that some of the more vitriolic and long-lasting debates in the study of human behaviour – including the nature/nurture debates about IQ and about mental illness – are caused either by entrenched, rival conceptions of causality or by confused or inadequate conceptions.

Let me give two simple examples and then ask a question.

Let's define a cause as a factor with the property that if it had been different, the effect in question would not have happened (or, perhaps, would have been less likely). I think this is the common core of all definitions of causation. You mentioned the Holocaust. There are many levels at which one could address the issue of what caused it. According to my definition above, Hitler caused the Holocaust by ordering it: had he given different orders, it would not have happened. However, by the same definition, many other factors also caused it: the propensity of the German people to condone such orders is one of them. So is the propensity of the German political system a few years earlier to bring a tyrant to power.

That all makes sense, but unfortunately, according to the above definition, it is just as true that the Holocaust was 'caused' by the attributes of the victims – particularly by the fact that they were Jews, Gypsies, etc., for if any of them had lacked those attributes, they would almost certainly have survived. If a reputable historian were to insist on using that definition, and to publish studies of the 'causes' of historical events in that sense, you can imagine what legions of bad people, and bad journalists, would immediately and forever afterwards seize on the fact that "studies have shown" that the Jews themselves caused the Holocaust. So that definition of causation is inadequate – and highly misleading as it stands – for use in an explanation of the cause of the Holocaust.

For the same reason, if we use that definition of causation in the study of the genetic origin of any other human behaviour, we shall make equally massive mistakes. For example, we would easily conduct a scientific study and find overwhelming evidence that lynchings of black people were caused by the black people's own genes.

Now I come to my question: when you say that there is abundant evidence of brain disease causing what is defined as "mental illness", what do you mean by "causing"? Do you mean that there is evidence that if certain brain lesions detected in the victims of, say, schizophrenia, had been absent, then the victims would not have displayed schizophrenic behaviour? (I.e. the same level of evidence as that which indicates that black people's genes were a cause of lynchings or Jewish genes were a cause of the Holocaust.) Or do you mean something more?