MOST work in the social sciences, like most work in history and in the natural sciences, is not directly concerned with the furnishing of explanations. It may well be true, as is so often said, that the tasks of any empirical science are to explain, to predict, and to apply. But in any such field these tasks are embedded in a workload that may be referred to as identification, classification, description; and measuring and reporting as well. These activities intersect, of course. We may be able to identify the practice of cross cousin marriage by the description that we have given of it, though identifying the practice is different from describing it. Nor is it true, as is sometimes said, that ‘classification and description are really the same process’. 1 The view supporting this belief has been put as follows: ‘To describe a given animal as carnivorous is to classify it as a carnivore; to classify it as a reptile is to describe it as reptilian. To describe any object as having a certain property is to classify it as a member of the class of objects having that property. 2