Abstract

This chapter reviews the problem of finding the 'right level' of causal explanation in psychiatry. This is not a purely philosophical problem, but one that frequently arises in practice for psychiatrists. For most scientists, experiment is the crucial test of a causal hypothesis: for X to cause Y is for intervention on X to be reflected in a change in the value of Y. But this kind of approach cannot tell us the right "level" at which to specify the causes of a particular outcome. The natural idea is the right level is one that specifies the 'mechanism' by which Y is produced. But the notion of a "mechanism" in psychiatry is obviously problematic. This chapter attempts to locate the sources of the difficulty here, looking at both a priori and empirical views as to when a "mechanism" has been correctly specified.