The Toolbox Does Not Shrink
For the past forty years, Jon Elster has attempted to explain things ranging from the emotions to technological change. The result is dozens of books (and even more papers) in three languages across four universities. And throughout, his work has not just been exemplary social science, but has always struggled with the question of what social science should be — what kinds of explanations are legitimate, which techniques should be used, and so on.
As he reaches his late sixties, it is understandable if he begins to think of his legacy. That certainly would help explain his latest book, Explaining Social Behavior: More Nuts and Bolts for the Social Sciences (Cambridge University Press, 2007), a 500-page masterpiece that I expect will be seen as the summation of a brilliant career.
It’s a book unlike any other and, as a result, unless read from start to finish can seem bizarre, if only because one has little sense of what the book is trying to do. It is not a guidebook, or a textbook, or a piece of social science in itself. In short, it is nothing less than an attempt to summarize an idealized vision of the whole of social science in simple language.
The book’s foundational assumption (as implied by its title) is that the goal of social science is to discover explanations for social phenomena. It begins by describing what explanations are and discussing their different forms. But the bulk of the book consists of tools that can be used in explanations: emotions, norms, time discounting, weakness of will, magical thinking, cognitive dissonance, heuristics and biases, rationality, irrationality, neuroscience, evolution, externalities, game theory, pluralistic ignorance, informational cascades, collective action, cyclical preferences, institutions, etc. — in short, the entire toolkit of the social sciences.
Just as amazing as the breadth topics is the way in which they’re covered. Elster explains each phenomenon clearly and concisely, so that any educated reader can understand them with little effort, without ever sacrificing intellectual depth. His explanations are peppered with examples from an amazing variety of sources: ancient history, recent history, personal experience, the classics of social science (e.g. Tocqueville), the great philosophers (Montaigne, Pascal, Mill), and classic novelists (e.g. Proust). The result is a book which not just introduces readers to the discoveries of the social sciences but to the intellectual world as a whole. Bibliographical notes following each chapter as well as the conclusion provide a rich guide for further exploration.
And yet it’s not simply a compendium of interesting results in the social sciences, but attempts to defend a particular conception of what the social sciences should be. In the conclusion, Elster defends his notion of social science as the attempt to discover particular explanations for particular phenomena against the “soft obscurantism” of the literary theorists and the “hard obscurantism” of the economists. As part of this, he turns his back on the notion of rational-choice models being an explanation in themselves, noting that their many assumptions are in desperate need of empirical defense.
In response to an earlier draft of this review, Elster wrote “I’m glad you appreciate the details in my book, but you’re missing the big picture, which is that there isn’t any.” Instead of trying to build a Grand Theory which explains all of social life, we should try to build explanations of particular phenomena from the nuts and bolts we have lying around. And “even if a dominant explanation of a given event or episode is discarded and then resurrected, the building blocks or mechanisms at work in the discarding and resurrection remain. The repertory, or the size of the toolbox, does not shrink.”
For anyone who cares about social science, Elster has done an amazing service in clearly describing the toolbox’s contents and defending its importance.
This is the first post of Elster week.
You should follow me on twitter here.
May 11, 2008
You can also send comments by email.