Abstract
Philosophical discussions about health and disease often refer to a 'medical model' of bodily disease, in which diseases are regarded as causes of illness; diagnosis consists in identifying the disease affecting the patient, and this determines the appropriate treatment. This view is plausible only for diseases whose cause is known, though even in such instances the disease is the effect on the affected person, and must not be confused with its own cause. But in fact the medical diagnostic process which progresses from recognition of patterns of symptoms and signs, through search for abnormalities of structure and/or function, towards knowledge of causation often stops short of this desirable end-point; and at whatever point it comes to a halt, its result is expressed in terms of 'diseases'. Thus in medical discourse the names of diseases are a convenient device by which the current conclusion of the diagnostic process can be stated briefly; and they have widely varying factual implications. This nominalist analysis of the medical usages of the names of diseases has consequences for definitions of health and disease, and for some problems in medical ethics.