What explains this state of affairs? The decline of methodological talk may have reflected a general exhaustion or a dampening of enthusiasm for positivism and the strong behavioralist program of the 1960s. Many political scientists have been content in the last couple of decades to settle down quietly under labels such as postpositivism, postempiricism, and pragmatism, without much sustained analysis as to what such labels actually amount to. And political theorists, with few exceptions, have seemed uninterested in probing these questions.

Perhaps such quiescence is to be expected when there exists widespread consensus about the nature of political inquiry. But with the emergence of Perestroika, no one can pretend that this is the case. Perestroikans charge that certain methodological approaches have established a “hegemonic” position. If that is true, then what becomes most striking about the current state of political science is the stark contrast between the force of this expansionist strategy and the paucity of sustained methodological reflection bearing on its intellectual respectability.

What might we expect from a reinvigoration of methodological talk? One thing we shouldn't expect is any quick fix. Instant therapies were already making their appearance at the 2001 meeting. But such proposals are likely to be perceived either as new power grabs or as too amorphous to have much real purchase on our problems. An example of the former was David Laitin's conference paper that asserted categorically that we already have “impressive intellectual coherence” in our discipline; the real problem is institutional “anarchy” and “indiscipline” that just needs to be cleaned up (“The Political Science Discipline,” p. 3). An example of the latter was Elinor Ostrom's well-intended proposal to gather all political inquiry under a big-tent description: the systematic study of “all human activities guided by norms” (Roundtable on “Shaking Things Up? Future Directions in Political Science”). Upon hearing this proposal, a person in the audience behind me mumbled, “I'll study bass fishing.”

Given this situation, Bent Flyvbjerg's Making Social Science Matter is a welcome intervention. Flyvbjerg is a Danish social scientist who recently published an excellent case study of the city of Aalborg—essentially what used to be called a “community power” analysis—only this one persuasively incorporated a Foucaultian dimension to its understanding of power. The author's new book projects a general vision of how social science should reimagine itself. A key issue in this regard remains, of course, what one thinks about the unity of science thesis. My guess is that there are relatively few political scientists today who believe that natural and social science are exactly alike; but many still think that some kind of progress on a naturalistic model is possible, and thus the profession should adhere to a clear-cut research program like rational choice. But, despite the many real insights and limited-range generalizations generated by that program, the fact remains that there is no general consensus that political science over the last quarter-century looks like an emerging science on the naturalistic model. Critics of such a research program increasingly find its adherents to be like die-hard Marxists: absolutely convinced the revolution will come and full of an unending supply of reasons why it has failed so far.

Flyvbjerg shares this familiar skepticism about the naturalistic model. A useful way to engage his arguments is to compare and contrast them with ones that were developed in the earlier period of ferment in political science. From this angle, one can see better both what is distinctive and persuasive in Flyvbjerg and what is more problematic. Like earlier critics, he contends that the naturalistic model does not adequately account for the irreducible centrality of “context and judgment” in human action (p. 4). But Flyvbjerg differs from many critics in his construal of the grounds for such centrality. The more familiar approach runs something like this: Because judgment and behavior in political life reflect actors' interpretations of common meanings embedded in particular contexts, and because these meanings and contexts cannot be grasped by abstract, universalistic models of behavior, we cannot expect such models to do very well in predicting political action. This line of argument is perhaps most closely associated today with Charles Taylor. Flyvbjerg finds this kind of hermeneutic account to be inadequate. The problem with this approach is that it amounts only to “a more complex form of determinism”; in effect, if you know the rules governing common meanings, you can determine which actions will occur. But, Flyvbjerg contends, this is surely wrong: “The rules for chess are not chess,” and the rules governing common meanings are not a determinate map of expected action (p. 43). Flyvbjerg is, of course, right about this point. The problem is that he never indicates who actually ever believed that rules strictly determine the playing of a game. Wittgenstein certainly didn't; and Taylor doesn't either. Strangely, Flyvbjerg does not include any reference whatsoever to Taylor.

Flyvbjerg contends that we must turn to the work of Hubert Dreyfus as our main resource for showing why action is generally too open-ended to be successfully modeled. Dreyfus's research has shown that the knowledge and skill of a competent actor—whether a paramedic or a politician—is too intuitive, holistic, and context-dependent to be “externalized into rules and explanations” (p. 21). Flyvbjerg's discussion of Dreyfus is generally quite good in drawing out the latter's significance for social scientists. But I do have some reservations. It strikes me as unwarranted to make the case for a nonnaturalistic model depend primarily on Dreyfus's particular theory of human skill and judgment. Given Flyvbjerg's denigration of the hermeneutic account, he forces us to place all our bets on Dreyfus. I would rather reverse the weighting, with the heaviest bets put on a (correctly understood) Taylor-like account that can be combined (easily, I think) with Dreyfus-like insights. Moreover, we would do well to remember that Dreyfus's work actually fits into an existing tradition of reflection on the character of political inquiry. In his much discussed essay of 1969, Sheldon Wolin characterized political insight as involving “tacit political knowledge” (“The Vocation of Political Theory,” American Political Science Review 63:4 [December 1969]: 1070–4).

Flyvbjerg's heavy reliance on Dreyfus is related to the Aristotelian orientation he envisions for social science. Aristotle's phronesis, with its focus on the quality of judgment and practical wisdom exhibited by political actors, is the key concept. Flyvbjerg aims to “restore social science to its classical position as a practical, intellectual activity aimed at clarifying the problems, risks and possibilities” that are faced by “the communities in which we live” (pp. 4, 166). To his credit, Flyvbjerg is quite concrete in spelling out methodologically what this vision implies. Thus, for example, he includes a chapter on the importance of case studies (which tend to be denigrated somewhat within the naturalistic model) and a chapter elucidating his own research on Aalborg. Flyvbjerg's “phronetic social science” does make one sharp departure from Aristotle, in that it self-consciously attends to how conflict and power are constitutive of political life and inquiry (pp. 3, 129). Although much of what Flyvbjerg says about Foucault on this theme is persuasive, some familiar questions about the latter's perspective on power in general, as well as its relation to domination, remain unanswered. For example, Flyvbjerg unpacks Foucault's notion of power as “a multiplicity of force relations” and “the strategies in which the force relations obtain effects” (pp. 116, 120). But what, then, are “force relations”? Similarly, one thing a contemporary phronetic science is supposed to do is unmask relations of domination; but the normative criteria for picking out instances of domination remain unclear (p. 104). Flyvbjerg, like Foucault, wants to avoid using universalistic normative frames for addressing this issue. He contends that we need only appeal to the normative frames of “the communities in which we live.” Without further explanation, however, this would seem to move us toward a position like Richard Rorty's, wherein political inquiries aim at clarifying the implications of authoritative traditions (p. 140). Flyvbjerg seems unaware of the potential tension here: Rorty castigates inquiry that busies itself with just the sort of systematic unmasking operations a phronetic social science undertakes.

Although I have difficulties with some of Flyvbjerg's claims, the book does an excellent job of bringing a whole range of timely methodological issues onto the table and of projecting a thoughtful vision of social science that runs counter to the hegemonic one. Proponents of the latter owe the discipline careful responses to proposals such as Flyvbjerg's. Failure to do so merely feeds the growing suspicion that what is at issue is not a justifiable hegemony but, rather, simply a coercive occupation of intellectual terrain.