This is a label for a battlefield on which philosophers and sociologists fight each other using their most sophisticated conceptual devices.37 Philosophers appear entrenched in one of the fields, their arms made of elements of a traditional artillery which, it is supposed, will one day be the definitive weapon for defending the distinctive feature of science: a theory of rationality. Sociologists at the beginning seemed to be on the defensive in the face of something that did seem more like an accusation than a mere description.38 Later, relativism became something to embrace rather than to avoid. Yet "relativism" is a term whose meaning has no level of precision higher than that of a "straw man" against or in favor of which a large part of the controversy is directed. Collins's work can be considered the turning point because he uses the term "relativism" in the title with which he presents his own SSK subprogram.39 But there is something interesting with respect to his use of the term. For Collins presents relativism in the first place not as a thesis about the nature of science, but as a strategy for the study of science.40 The interesting point is that the aim of this strategy in Collins's work is the same as that of other authors, with their other strategies. It is this that I am interested in highlighting first.

As we have seen, for Collins, science, as (or better, more than) any other collective activity, relies on an established order, whose principles as is always the case with unchallenged, stable orders are taken for granted and so become invisible. For this reason, one needs to make an extra effort to design a special strategy to uncover them. Collins calls his strategy to uncover the taken-for-granted principles which scientific culture relies on, precisely, relativism. His relativistic strategy, he says, "rests on the prescription: treat descriptive language as though it were about imaginary objects." But why precisely this prescription? Because if we want to know scientific culture, we should not take for granted what it does; this is what relativism intends, it "demands that the analysis of the way knowledge is established is not shackled at the outset by common sense judgments about what is and is not true." And the way scientists establish the truth is precisely by referring to descriptive sentences, i.e., sentences that are supposed to describe what the world is really like. If what we want is to study scientific culture itself, then we need the relativistic prescription, because "if cultures differ in their perceptions of the world, then their perceptions cannot be fully explained by reference to what the world is really like."41 So taking a relativistic stance (his prescription) would lead us to begin by doubting what is generally taken for granted, namely, the nature and decisive role with which descriptive sentences are empowered.

Thus relativism, such as Collins presents it, is first a strategy for knowing knowledge making rather than a thesis about knowledge itself. This recommendation of distancing oneself from what everyone especially scientists thinks about science, as a fruitful strategy for uncovering taken-for-granted assumptions, is a topic in the SSK literature.

Bloor's principle of impartiality also demands that a naturalistic explanation of knowledge be independent of how scientists evaluate it; it has to explain both what they take to be true and what they take to be false, as well as what he takes to be rational and what he takes to be irrational.

This is also the starting point of Latour and Woolgar's (1979) book. Here they recommend "maintaining analytic distance upon explanations of activity prevalent within the culture being observed." "In the case of scientific culture in particular there is a strong tendency for the objects of that culture [facts] to provide their own explanation. Rather than produce an account which explained scientists' activities in terms of the facts which they discovered, our interest was to determine how a fact came to acquire its character in the first place."43 And when Shapin and Schaffer (1985) present their impressive work on the experimentalist tradition in modern science, they claim: "Ordinarily, our scientific culture's beliefs and practices are referred to the unambiguous facts of nature or to universal and impersonal criteria of just how people do things (or do them when behaving rationally')." "To be a member of a culture," they add, "is to act considering some things as if they were self-evident, and this marks decisively the version of that culture that can give native members." Inversely, "The member who poses awkward questions about what everybody knows' in the shared culture runs a real risk of being dealt with as a troublemaker or an idiot. Indeed, there are few more reliable ways of being expelled from a culture than continuing to seriously query its taken-for-granted intellectual framework. Playing the stranger is therefore a difficult business; yet this is precisely what we need to do with respect to [scientific] culture. . . We wish to adopt a . . . . suspension of our taken-for-granted perception of experimental practice and its products. By playing the stranger we hope to move away from self-evidence." The advantage of the perspective of the stranger, the authors say finally, is that the stranger "knows that there are alternatives to [native] beliefs and practices."

In this way, relativism, as Collins explicitly presents it as the concrete way to approach science, follows the pattern generally recommended by SSK studies; namely, choose a path that will permit the uncovering of taken-for-granted assumptions. Why do they make this strategy a necessary starting point in the study of science? Perhaps the best answer to this question would be another question: why not? After all, our philosophical culture takes for granted that science is something very different (demarcation), that it operates according to exclusive rules of its own (internalism), rules, in turn, which are empowered by an epistemological virtue (rationality). Although the technical terms come from recent philosophical developments, the assumptions behind them come from the conventional wisdom about science that we all share lay person, scientist, and philosopher alike. For this reason, these assumptions represent too committed a starting point in an approach to science, especially if one intends to be a naturalist, one for whom what really happens must take priority over what is supposed to be the case.

When Gooding, Pinch, and Schaffer (1989) comment on the work of Franklin a paradigmatic member of the philosophical rationality field they say that for Franklin what counts is that there do exist rational strategies to evaluate scientific results. For SSK, instead, what counts is that those strategies are "culturally accepted practice"; but this does not mean that science is either an irrational enterprise or one in which anything goes. What the two sides disagree about is not what they say about science, but about how to approach the study of science. Contrasting Franklin's way and the SSK way, the former takes as explanans what the latter takes as explanandum.45 There is thus a difference between philosophers (rationalistic) and sociologists (relativistic) and it is that the former use epistemological norms to explain knowledge reliability this is all they think they have to explain while the latter extend the field of what must be explained to include the epistemological norms.

It is important to point out that Franklin is working within a recently modified framework regarding methodological matters. For philosophy of science began with a foundationalist/normativist view, making demarcation a radical dividing line and cognitive phenomena part of the Platonic-Popperian "third world." Then came naturalization which treats cognitive phenomena as part of the natural world and bases its normative claims on the empirical investigation of real scientific practice past and present. It is as a result of this kind of work that Franklin claims that there do exist epistemological criteria, and they do explain knowledge reliability. Yet he does not ask for an explanation of the criteria themselves.

Now enters sociological naturalization, which asks for such an explanation, and it gives one: its status is that of "culturally accepted practices." With this move, cognitive phenomena appear as part of the realm of social convention. And if this were true, demarcation would collapse. Here is where relativism stops being a mere strategy for knowing about knowledge, and becomes a substantive thesis about knowledge itself. It stops being a metamethodological strategy and becomes an epistemological thesis. And here, of course, is where the controversy about philosophical rationalism versus sociological relativism gets its focus. Yet the opposite of rational is neither relativistic nor social terms the debate has turned around but irrational.46 Only if sociologists had called their own work or science itself irrational would philosophers be justified in using rationality as a weapon to fight against them.

Inversely, the opposite of relativism is not rationality but absolutism; and the absolute be it norms, principles, or truth may properly have a place in the Popperian third world, but it is hard to see how a naturalist could find it in the messy and ever-changing reality of which both the world we want to know and knowledge itself are a part. This difficulty could be the reason why relativism is not a problem that SSK imposed on philosophy from the outside. For, from Quine's ontological relativity to Putnam's conceptual relativity, and on through Kuhn's incommensurability, philosophy in the second half of this century, to a great extent, has been an attempt to banish the phantom of relativism. And perhaps these philosophers are the only ones who have remembered that one of the major tasks of philosophy has always been that of uncovering taken-for-granted assumptions.