Perverse Effects and Social Contradictions: Analytical Vindication of Dialectics?

Raymond Boudon's recent textbook La Logique du Social, constitutes a systematic attempt to provide sociology with solid foundations, by putting at its core the assumption that man acts rationally, or at least intentionally. Interactionist models, which rest on this assumption, Boudon argues, should replace determinist models, which assume that man is passively determined by social structures, roles or norms.l In particular, he argues that the notion of perverse effect, whose use presupposes the adoption of an interactionist perspective, should play a central role in the attempt to account for social change. In this review article, I shall focus on this notion and on the use Boudon makes of it in his theory of social change. I shall point out a number of obscurities and other defects, and make some suggestions as to how they can be remedied. In so doing, I shall draw heavily on two other books. One is an earlier collection of essays by Boudon, Effets Pervers et Ordre Social, which deals with such topics as educational reform and Rawlsian justice, the 'logic of relative frustration' and sociological determinism, and which provides a less systematic formulation of the ideas expressed in the textbook. As most of the essays make explicit use of the notion of 'perverse effect', their relevance to the concerns of this article are obvious enough. The other book I shall heavily draw on is a somewhat more technical one by Jon Elster, Logic and Society. The purpose of this book is to use various logical tools in order to work out a number of sociologically useful concepts. One of these is the concept of social contradiction, on which we shall concentrate in subsequent sections. Not only does this concept lie at the core of Elster's theory of social change-by itself a sufficient reason for wanting to undertake a comparison with the key notion in Boudon's treatment of change. As we shall see, it is also analytically related to Boudon's notion of perverse effect and, just like the latter, it only makes sense on the background of an intentional model of man.2 Bringing Boudon's and Elster's analyses of these issues together, will, I hope, help make the former more rigorous and the latter more