Is mental illness an object of knowledge? The history of psychiatry teaches
us to doubt it, by emphasizing the infinitely variable and fluctuating cha
racter of psychiatric entities. Mental illness is not simply 'out there', w
aiting to be described and theorized by psychiatrists; it interacts with ps
ychiatric theories, clinical entities waxing and waning in accordance with
diagnostic fashions, institutional practices and methods of treatment. This
should be a warning to psychiatrists and therapists: their intervention is
part of the 'etiological equation' of the syndromes that they claim to obs
erve from the outside. But this should also be a warning to historians of p
sychiatry, who can no longer be content with writing the history of ready-m
ade syndromes and psychiatric theories. They must, if they want to remain f
aithful to their improbable 'object', study the complex interactions from w
hich those syndromes and those theories emerge, somewhere between the docto
rs, the patients and the society that surrounds them. In short, they must s
tudy the making of psychiatric history, and understand that they participat
e in it.