Official Statistics and Family Sociology

Abstract

In generating the data required for their purposes, sociologists of the family have often worked with official statistics. Instances need not be offered because they are so numerous, but reference to any textbook on the family, or to issues of this journal, indicates extensive use of census returns, registration data, judicial statistics and similar material. For one who has worked in this way in connection with divorce, it is therefore discomforting that there has developed from within sociology a radical attack on the use of data of these kinds. Radical, in this context, means an attack not on any particular series of statistics, nor on any particular use made of them, but rather a principled rejection of all such statistics for any valid sociological purpose, except possibly for studying the processes of what Douglas (1971a) calls the bureaucracies of official morality. This attack comes, of course, from within recently-emerging schools of phenomenological sociology, particularly ethnomethodology, and the assault on official statistics is but part of a thorough-going critique of sociology which amounts to a denial of its credibility as anything other than a folk discipline. Initially this approach was marked by esoteric practices and indifference to dialogue which enabled most sociologists to ignore it, or at least to treat it as no more than gossip. In view of the literature now developing, however, such a response would be in bad faith, and conventional sociologists must attend to the claims being made. These amount to saying that ethnomethodologists have effected, or are effecting, a paradigm shift in sociology, that is, a fundamental reconstruction of its charter. To be sure, the ph nomenological thrust is not monolithic, c ntains unresolved difficulties and ambiguities, and is still somewhat inchoate, but the central drift is clear. This goes far beyond suggesting an extension of the field of sociological inquiry, and seems to consist of an antirationalist critique which dismisses the kinds of evidence and forms of argument on which most sociologists have based their work. By implication this dismissal must apply also to other kinds of social scientists worki g in a similar style, and if such work is to continue in good faith then a response is needed. Critical response to the generality of the phenomenological position is indeed being provoked, and this paper has only a restricted purpose. By means of exposition and an extended illustration relating to divorce, it seeks to justify the continued use of official statistics in family sociology, and to do this it draws heavily upon the critical writings of Douglas (1967; 1971a; 1971b). These are selected as examples of phenomenological argument because they provide particularly elaborated criticism of the use of official statistics, although such elaborations can also be found elsewhere (Cicourel, 1968).