Are mental disorders harmful as a matter of definition, or are they simply conditions that quite often cause problems? Should someone who has “symptoms” but who suffers no harm be diagnosed with a mental disorder? In this chapter I shall show that these are crucial questions but
have not been given the attention they deserve. The idea that disorders are linked to “distress or impairment” was first explicitly introduced into the DSM in the DSM-III. This criterion originally came to be introduced following the removal of homosexuality from the DSM in 1973. In the DSM-IV both the general definition of mental disorder contained in the introduction and many of the individual diagnostic criteria sets required that symptoms cause (or at least increase the risk of) distress or impairment before a disorder could be diagnosed. In DSM-5 the requirement that disorders cause harm, or an increased risk of harm, has been dropped from the definition of mental disorder. This matters because many people have the “symptoms” of mental disorder but suffer no distress or impairment, and are not plausibly at an increased risk of suffering distress or impairment in the future. Under DSM-IV such people could not be diagnosed, now they may be. This chapter offers a philosophical history. I show how the conceptualisation of the link between harm and disorder has shifted through the different editions of the DSM, and make it clear why this matters.