In a recent paper, it was argued forcefully by Thomas Szasz that it is crucial to the scientific credibility of psychiatry that it abandon talk of the behaviour of the mentally “ill” in terms of causes: such behaviour is not caused by their condition—it simply has reasons, which are discounted by the medical model. It is argued in this paper that Szasz's theory is incomplete for two reasons: first, in assuming that reasons are radically different from causes, it cannot account (...) for the possibility that “sane” behaviour might be just as much caused as “insane”; and second, it tacitly assumes that the origin of behaviour always lies with the agent—a view that arguably is an accident of grammar. Hence while there is no mental illness, this is because there is nothing that could be ill—and this means that there is no such thing as mental “health” either. (shrink)