In the early moderera the notion of imagination was made responsible for phenomena which were later explained in terms of embryology, genetics, psychology, bacteriology or other scientific disciplines. Images, often seated in the upper abdomen (hypochondriac region) or the womb (hysteria), were regarded as powerful influences on material reality. In the course of the seventeenth and eighteenth centuries the hypochondriac forms of imagination became mere whims and spleens, but they kept much of their original potency in respect of the uterus, accounting for monstrosities and the shaping of human offspring. The hysterical conversion of imagination into somatic phenomena has never been questioned. Since the two World Wars the realm of imagination has again expanded beyond the uterus and the older disease-concepts. In the last 10–20 years images seem to have regained some of their original creative force.

Nineteenth-century theories of human volition are discussed in relation to ideas on insanity and responsibility. Attention is drawn to the importance of accounts of volition for medical psychologists and to the implications of these accounts for medical and lay discussion of criminal responsibility.

Behavioural psychotherapy has long historical roots. Recently it has led to effective treatment for selected neuroses, including phobic, obsessive-compulsive and sexual disorders. Potent therapy has become a tool of experimental psychopathology which advances theory and practice. A pervasive principle is exposure of the patient to those stimuli which evoke his discomfort until this subsides. Level of arousal during exposure does not affect outcome. Theoretical issues are reviewed which decide when exposure will be sensitizing or habituating. Both psychoanalytical and conditioning models of neurosis are out of date, and models derived more directly from clinical experiment are becoming possible. The aetiology of phobias and rituals can be seen as failed extinction rather than enhanced acquisition. Relevant phylogenetic and biological factors are discussed. At the other extreme, well-documented faith-healing indicates huge gaps in our knowledge of psychotherapy.

A reappraisal is offered of the precise sense in which the introduction of moral treats ment at the end of the eighteenth century marked a decisive shift in our characteristic ways of responding to and coping with the mentally disturbed. Some connections are explored between this new approach to the ‘mad’ and larger social movements and processes.

Before 1600 the ‘hysteric affection’ was a paroxysmal ailment of women explained as primarily due to the condition or malposition of the womb. During the seventeenth century attention shifted from the womb to the brain. Then Thomas Sydenham's clinical method yielded a view of hysteria which comprehended a wide range of illness with a mental component, and which was related to the whole person. In the course of this paper the relation of the hysteric affection to witchcraft, demonic possession, St Vitus' dance (chorca), hypochondria and melancholy is also noted.

A sufficient number of studies of geographically defined populations have now been done so as to make it clear that when all the various kinds of mental illnesses are lumped together, the total prevalence rates commonly amount to 20% or more. These findings imply major challenges to theory, policy, administration, and practice in the health field, and point to a need for vastly more information and understanding. This article sketches some of the directions such investigations might take.

The concept of mental illness is explored by noting its connections with other concepts, such as deviance, social maladjustment and morbidity. It is fitted to patients from some of the chief categories into which they are classified by psychiatrists. The objection that the concept is one of value is examined and deflated.

An account is given of the role of an attached social worker to a primary care team in the management of chronic neurotic illness in the community. The medico-social implications of the findings are discussed.

Thirty patients were selected for a prospective study according to two criteria: (i) an irresistible urge to overeat (bulimia nervosa), followed by self-induced vomiting or purging; (ii) a morbid fear of becoming fat. The majority of the patients had a previous history of true or cryptic anorexia nervosa. Self-induced vomiting and purging are secondary devices used by the patients to counteract the effects of overeating and prevent a gain in weight. These devices are dangerous for they are habit-forming and lead to potassium loss and other physical complications. In common with true anorexia nervosa, the patients were determined to keep their weight below a self-imposed threshold. Its level was set below the patient‘s healthy weight, defined as the weight reached before the onset of the eating disorder. In contrast with true anorexia nervosa, the patients tended to be heavier, more active sexually, and more likely to menstruate regularly and remain fertile. Depressive symptoms were often severe and distressing and led to a high risk of suicide.

A theoretical model is described to emphasize the interdependence of the various symptoms and the role of self-perpetuating mechanisms in the maintenance of the disorder. The main aims of treatment are (i) to interrupt the vicious circle of overeating and self-induced vomiting (or purging), (ii) to persuade the patients to accept a higher weight. Prognosis appears less favourable than in uncomplicated anorexia nervosa.