The Third Way? The specter of communism? Philip Kitcher's imagery reminds us that there are political stakes in philosophy, but I fear that the lenses provided by the rhetoric of Tony Blair and Karl Marx and Friedrich Engels cloud his analytic response to The Fate of Knowledge. Bipolar representations of multipolar situations work no better in politics than they do in epistemology.

The views Kitcher and I defend in our new books are closer to one another than either of us might have anticipated on the basis of our previously published work. We both support some kind of democratic inclusiveness in science, we accept some forms of theoretical pluralism, we agree on the context-dependence of scientific significance, we agree that a complete account of scientific rules or norms must include rules applicable to social interaction, we even agree that scientific judgment should be held to evidential standards. Why then do we express such reservations over the other's formulations? Perhaps one answer lies in the political histories conveyed by Kitcher's images. But part is accounted for by the broadness of the principles on which we partially agree, a broadness that leaves room for a range of specific articulations. In translating my views into his preferred locutions, Kitcher tries to place them in the sensible middle (which he and other “centrists” occupy) of a left-right alignment in science studies. The “isms” he marshals for this task either omit central elements or are still broad enough to encompass quite different positions. As a result, Kitcher represents some of my views as more closely aligned with his than I recognize. Others are dismissed as incoherent, unmotivated, or just plain wrong.

There are a number of crucial points at which the positions which I advanced diverge from Kitcher's representation of them. Taken together these constitute ingredients for a more coherent, but possibly more provocative, stance than Kitcher credits me with. Just as real socialism was expunged in New Labor's Third Way, the epistemological socialism for which I argue just doesn't fit Kitcher's niches. At risk of re-exile to the margins, I must decline the apple and reclaim my own position in the garden.

Individualism. Kitcher understands individualism partly as a thesis about the aims of scientists (his (a) and (b)) and partly as the constitutive thesis (c) “community acceptance is to be understood in terms of acceptance by the individuals who comprise the community.” He sees nothing in my book that constitutes rejection of individualism thus understood. But there's more to individualism, or at least to the individualism I am concerned with. Individualism is also about the self-sufficiency of individuals. The individualist thesis against which I argue is something like the following: “empirical knowledge can be fully understood in terms of processes undergone by epistemically self-sufficient individuals.” I offer a number of reasons for rejecting this thesis, but my rejection of it means that I advocate a stronger form of the “Interactionist Socialism” to which Kitcher says he, Alvin Goldman, Susan Haack, and I equally subscribe. Instead of Kitcher's “scientists accept statements as the result of interactions with others (debate, collaboration, instruction, and so forth),” which suggests a contingent relationship between acceptance and social interaction, I argue for a normative social element as part of the meaning of “knowledge”, i.e., that epistemic acceptability of content (or epistemically justified acceptance of content) requires the satisfactory performance of certain kinds of social interactions. Thus, my version of “Social Methodism” is that the normative notions central to common understandings of “knowledge,” such as epistemic acceptability and conformation (in my preferred locutions), involve both traditional evidential norms and norms of effective critical interaction, indeed that evidential norms, properly understood, include those social norms. This is a bold claim and The Fate of Knowledge devotes considerable space to developing, defending, and clarifying it.

Kitcher advances a view that might be called Socialism Lite, on the other hand, which accepts a contingent version of Interactionist Socialism, i.e., that as a matter of fact scientists are located in communities and accept content as a result of their interactions with each other (or, in an alternative formulation, that some knowledge is produced by scientists working in groups or in institutions). Given this de facto sociality, it is possible to channel the exercise of individual rationality within these structures and institutions to affect their outcomes. The mathematical models of Kitcher and Goldman are designed to show that the social institutions of inquiry can be organized to maximize the chances that individual (cognitive and non-cognitive) decisions will result in community acceptance of the correct theory. Socialism Lite might acknowledge that the intricate and powerful knowledge produced by scientific inquiry as we know it is made possible by the community of scientists, but hold that such knowledge is the cumulative product of multiple instances of individual knowledge, which, as knowledge, can be understood without reference to social interactions. Kitcher is quite clear on the dependence of science at any particular stage on prior stages, but strangely suspicious of concurrent epistemic interdependence. Although allowing for the cooperation of the social and the rational, Socialism Lite maintains their inherent distinction.Footnote ¹

Pluralism. The philosophical point I am at pains to make is that a theory of knowledge should not presume either pluralism or monism. Thus I think such a theory should be as open to a strong form of pluralism as to monism. Kitcher agrees that it is a mistake to presume monism and also that such a presumption has vitiated a number of previous attempts to find an alternative to excessive rationalism and excessive skepticism. But he thinks we need only accept a “modest pluralism” according to which “the bits of nature we choose to represent accurately are a function of us, our capacities, and our interests.” These representations will per force be incomplete, and different interests will lead us to classify the objects of nature in different ways. No sense, however, can be made by Kitcher of my suggestion that equally successful representations may be irreconcilable or non-congruent in any non-redundant way. Hmm. I countenance pluralism because I countenance the possibility of different equally defensible background assumptions facilitating inferences to quite different and irreconcilable, even non-mutually-consistent, representations of what is pre-theoretically identified as the same phenomenon. One kind of assumption, use of which has such results, is the sort distinguishing the stuff of a universe into entities and environments. Different equally defensible parsings of an ontological space can result in approach B treating the causally significant factors of approach A as undifferentiated portions of the environment of causally significant factors for B and vice versa.Footnote ² Something less innocuous than mere incompleteness is involved here, deeper than the commonplace points about reference that disarm Kuhnian incommensurability. Not only do I think this deeper form of plurality may characterize our knowledge of some complex natural systems, I think the context-dependence about kinds and categories to which Kitcher does subscribe also lends itself to this less modest pluralism, so I'm surprised that he wishes to draw the line here.

Democratism. I think that underdetermination and plurality mandate democratization of inquiry as well as of the policy directing inquiry. Kitcher wants to democratize science policy, not science. Kitcher thinks I am confused about the problems for which democratization is a solution. I think he thinks I (perhaps as a member of a group whose interests have been insufficiently attended to) should be concerned about the interest problem and not about the Millian problem, and that if I were clear on this point I would see that democratization of science policy is the appropriate goal. I think Kitcher has been seduced by his clever juxtaposition of significant truth (the interest problem) with significant truth (the Millian problem). But once we've decided what we want to know (significant truth), there is still a question whether our investigations have resulted in knowledge (significant truths or conformation Footnote ³) or how to insure they do. On my view, (tempered) democratism is required to secure this further issue. (There's also a question about the extent to which decisions about what we want to know shape the subject matter and hence ultimate outcome of our investigation.) Of course, it may matter to us, may affect our interests, which representations get accepted so the Millian and the interest problem are not unconnected. I admire Kitcher's call for (effective) democratic deliberation about the research agenda for and applications of scientific inquiry. But I would argue that the democratic agenda setters of Kitcher's well-ordered science should insist that the epistemic practices of the experts on whose testimony they must rely conform with the social norms of inquiry I advocate.

Exactly what arrangements will thus conform is difficult to spell out, and I am prepared to agree that the form of democratism that must characterize agenda setting for inquiry may in some ways differ from that that characterizes inquiry itself, or at least that we may wish to distinguish what functions inclusiveness serves in these two contexts. But I think attempts to democratize science without fully acknowledging the kind of strong epistemic interdependence for which I argue will fail to produce the intended results. I also think that the vast diversity of situation and outlook that characterizes human cultures and subcultures poses a challenge that current mainstream European and North American political thinking insufficiently appreciates. The Third Estate, whose solidarity in demand for “one deputy, one vote” is depicted on the jacket cover of Science, Truth, and Democracy, included no women and no deputies from the French colonies. Even more heterogeneous constituencies are excluded from but subject to contemporary regimes of power. It's much easier to countenance equality when all candidates for such status resemble each other and have relatively similar stakes in the political order. If we are going to do better in the twenty-first century than our forebears did in the eighteeenth, whether in politics or in philosophy, different thinking and different action are required. Scientific-technological development, broadly understood, has the capacity to improve global welfare. But if it is to do so, all those highly varied constituencies must come to have a stake in it. In my view the fate of knowledge rests in how the current monopoly of expertise can be broken and its production redistributed. That seems the challenge of really democratizing science.