Its a review about this product
Peter Hedström and Richard Swedberg are prominent sociologists who have put together an edited volume including several other prominent sociologists. The papers derive from a symposium held in Stockholm in 1996, and is dedicated to James Coleman and Robert Merton. The former is a representative of "grand theorizing" and the latter "middle-level theorizing," thus revealing a central ambiguity in the volume.
Hedström and Swedberg's introductory essay fall squarely in the middle-level theorizing camp, and describes a "social mechanism" as a process of beginning with a macro phenomenon with micro implications, followed by a causal process on the micro level that leads to a new micro level, followed finally by a macrosocial change that reflects the microsocial situation. The "enemy" in this endeavor are sociological models that simply work out statistical associations, such as the famous path analysis of O. D. Duncan that inspired generations of labor market and social stratification researchers.
Following Hedström and Swedberg, Thomas Schelling and Jon Elster make eloquent cases for using social mechanisms. This is not difficult for them, as they have been major contributors, including Schelling's model of neighborhood racial segregation and Elster's model of "sour grapes." Timur Kuran, another major contributor to the social mechanisms approach, also supplies a nice chapter.
But, is this really "an analytical approach to social theory?" These researchers have given us some great models with deep and surprising insights, but is there no general sociological theory?
My own take on this is that Talcott Parsons made an immensely insightful but flawed attempt at providing a general model for sociological theory, but this was rejected by the profession, and nothing was substituted for it. This is a tragedy that I struggle to understand. Why did Parsons not do a better job (he was very smart and accurately appraised the intellectual situation at the time), and why did other sociologists (such as his eight coauthors in Towards and General Theory of Action) not take up the baton rather than being either sycophants or enemies of the whole project?
For whatever reason, sociologists have not come forth with an alternative. Middle-level social theory is great as far as it goes, but middle-level is inherently ambiguous and shakey unless grounded in higher-level social theory. The less than exalted position of Sociology among the behavioral sciences, and especially its weakness when compared to economic or biological theory, is due precisely to its lack of a powerful general framework within which "middle-level" models operation synergistically to build up a general picture of society.
One quite marvelous essay in this volume directly addresses what passes for Grand Theory in contemporary sociology. Axel van den Berg analyzes Habermas, Bourdieu, Giddens, and Alexander, whom he considers the leading grand theorists of our age. Habermas, like many Continental philosophers/social theorists, seeks an interpretive model of society rather than an explanatory/predictive model. This is completely unacceptable from a scientific standpoint. I love Habermas, but he is about as much a scientist as a good cook is a chemist; i.e., not at all. That sociologists take Habermas as a social theorist is symptomatic of the weakness of sociological theory as a scientific discipline.
Van den Berg's treatment of Bourdieu is iconoclastic and devastating. Bourdieu figured out early in his career that the French public likes smoke and mirrors rather than real science, but also respected the power of science. So Bourdieu combined massive statistical data analysis with smoke and mirrors social theory. Everything Bourdieu says is either common sense, impenetrable nonsense, or common sense framed as impenetrable nonsense. Many intellectuals believe that if a theorist is comprehensible, he cannot be very insightful, and this sentiment is prevalent in France and Germany to a degree that theoretical hopefuls in other countries could only dream about. Academic success in these countries often takes the form of wrapping triviality in obtuse verbosity---Lacan and Latour are only the most egregious examples of this. Why American sociology could possibly place credence in Bourdieu vaporware is beyond me.
Van de Berg's treatment of Giddens is also quite cogent. Giddens is a lucid writer with fine sociological insights. But, despite his promotion of "structuration," he has nothing to offer as a Grand Theorist.
I think the most constructive and insightful paper in this book is Raymond Boudon's "Social Mechanisms without Black Boxes." Boudon is handicapped, as are virtually all sociologists, by the notion that the rational actor model implies that actors have self-regarding preferences. For instance, Boudon claims that the rational actor model implies that voting is irrational, since a single vote affects the outcome of an election with near-zero probability. Similarly, he claims that rejecting a positive offer in the ultimatum game is "irrational." This is simply not the case. If people like to vote, because they believe it is the proper behavior of a good citizen, then voting appears in the agent's preference function, and the individual will trade off between voting and accomplishing other personal goals can be made on irrational grounds. If people care about being treated fairly, then there is nothing "irrational" about rejecting unfair offers in the ultimatum game.
It is about time sociologists stopped perpetrating the myth that rational behavior precludes costly prosocial behavior. Much of what Boudon says in critiquing the rational actor model is just wrong. Valuing the welfare of others, or valuing the human character virtues (honesty, loyalty, etc.) for their own sake, is perfectly compatible with the rational actor model. However, Boudon proposes a new model, which he call the "cognitivist model," that is eminently reasonable, but is either the rational actor model, or a strengthening of the model in particular areas (e.g., imposition of conditions on beliefs so that rationality is stronger than Bayes' rule).
Admitting the centrality of the rational actor model in a theory of action would have monumental and salutary implications for sociological theory, and would reward skills (especially mathematical skills) that are currently in short supply among sociologists. This is the task of the current generation of sociologists.
