[Giddens, Anthony (1990), Review of The Cement of Society, American Journal of Sociology 96 (1): 223-225]

[ Books | Articles | Reviews | Index | The main menu ]


Review of The Cement of Society

[start of page 223]

The Cement of Society: A study of social order. By Jon Elster. Cambridge: Cambridge University Press, 1989. Pp viii + 311. $44.50 (cloth); $15.95 (paper).

Anthony Giddens
Cambridge University

This book is much more interesting than its rather dull-sounding title might lead one to believe. For some years now, Jon Elster has been pursuing the quest of demonstrating the importance of rational choice theory for core problems of the social sciences outside the main area of its provenance - that is, neoclassical economics. In some respects, Elster's The Cement of Society is a direct continuation of his previous works, and it certainly contains much the same mixture as before of lightly worn erudition and incisive arguments. In other ways, however, the book marks a new departure. In trying to connect the ideas of his earlier writings to a set of real-world problems - the explanation of patterns of collective bargaining in Sweden-Elster came to see that rational choice theory has more fundamental limitations than he originally believed. He now acknowledges that rational choice theory needs to be complemented with an analysis of social norms; and that norms provide sources of motivation that are "irreducible to rationality."

The problem with which Elster's book opens is a very familiar one, the celebrated "Hobbesian problem of order." What is the "cement" that binds individuals to one another, preventing their lives from disintegrating into chronic conflict or chaos. Elster distinguishes two aspects of social order, predictability and cooperation. In fact, in this particular book at least, he says very little about the former. Most of his discussion concentrates upon the nature of cooperation (as well as the conditions under which it is not forthcoming). Elster distinguishes five basic forms of cooperation. One type is cooperation in respect of "externalities" brought about by individual activity. People "cooperate," in a broad sense of the term, by means of individual acts that produce communally valuable outcomes. Examples are voting, paying one's taxes ,or giving blood. A second category concerns "helping behavior" - as when someone helps a person fix the other's car, keeps a promise, or tells the truth. In each of these types, cooperation benefits others whether or not they reciprocate in kind. What Elster labels "collective action theory" deals with these cooperative forms, in which the "free-rider problem" presents the main obstacle to successful cooperative outcomes.

A third type of cooperation is represented by a "convention equilibrium." Many, although not all, forms of convention equilibrium,

[end of page 223, start of page 224]

such as the rule of driving on the right-hand side of the road, are "pure" conventions: it does not matter on which side of the road everyone drives so long as the rule is observed by all. The two final types are "joint ventures" and "private orderings." Joint ventures depend upon direct collaboration between individuals in order to achieve a cooperative outcome. The example is the division of labor as described by Adam Smith: producers working collectively have a much higher output than they would in isolation, given that all play their part in the overall productive enterprise. Private orderings have the same characteristic, save that physical collaboration is not necessary: a prime example is the capitalistic labour contract. According to Elster, the last three types should be addressed primarily by bargaining theory. The main obstacle to cooperation in these types is not the free-rider problem, but rather failure to agree upon an appropriate distribution of the benefits accruing from cooperation. Cooperation in both collective action and bargaining is heavily influenced by normative commitments. A theory of social norms, Elster asserts forcefully, does not imply (as Durkheim thought) a commitment to holism in social science. Elster sustains the methodological individualism advocated in his previous writings: a social norm has no content other than the propensity to anticipate sanctions on the part of others at the prospect of behaving in a particular way that is either approved or forbidden. A central element of Elster's discussion is the demonstration that social norms cannot be interpreted reductively as the outcome of rational or optimizing actions. Of course, on some occasions, norms may be followed instrumentally, as a cover for self-interest. But norms cannot be upheld in this way most of the time because such strategies only work given the existence of binding normative commitments.

Underlying the mixture of self-interest and normative commitment that guides human action are various distinguishable sorts of motivation: envy, opportunism, and what Elster terms "codes of honor." The concluding chapter of the book analyzes these motivations and considers their interrelations. Here Elster reverts in some degree to the first aspect of the problem of order mentioned above, trying to show how various forms of trust in and credibility for the behavior of others is sustained even in situations in which those others seem to be acting contrary to self-interest.

Space precludes a full critical engagement with the book, which, in spite of its many attractions, seems to me to have a number of distinct weaknesses. Elster effectively retraces the path that Parsons took when he set out to analyze the "Hobbsian problem of order" some half a century ago, yet Parsons is barely mentioned and The Structure of Social Action is not listed in the bibliography at all. Elster at one point remarks of a particular author that he is "illiterate in political philosophy," but Elster's grip of the relevant sociological literature seems less than secure. Authors like Homans, Blau, Gouldner, and many others who have ad dressed questions of norms and social reciprocity are not mentioned. No allusion is made even to Goffman, surely in some respects the theorist par

[end of page 224, start of page 225]

excellence of cooperation, bargaining, envy, and trust in social life. On a more analytical level, Elster's book makes virtually no mention of phenomena of power and domination, which surely must be central to any endeavor that seeks to illuminate the nature of social order.

[end of page 225]

[Giddens, Anthony (1990), Review of The Cement of Society, American Journal of Sociology 96 (1): 223-225]

[ Books | Articles | Reviews | Index | The main menu ]


1 1