The Duck of Minerva

The Conduct of Inquiry: part the third of my contribution to a symposium

2 August 2011

 Ah, life — you get in the way of important things, like finishing a reply to an online symposium on my book that was started way back in January of this year at The Disorder Of Things. Part One of my reply is here; Part Two is here. At long last, Part Three is below the fold.

III. don’t cross the streams, but do talk about doing so

In a 2006 article entitled “Public Knowledge and the Difficulties of Democracy,” Philip Kitcher suggests that we need to pay less attention to the epistemological problems generated by a focus on the situation of the putatively isolated individual knower (the problem of true belief) and more attention to the critique of public reasoning. Specifically, he directs attention to what he calls the IIS, the Inquiry-and-Information-System, which he further subdivides into subsystems for inquiry (figuring out which issues are worth investigating), certification (determining which “claims” should be “deemed worthy to inscribe ‘in the books'”), and dissemination (communication of those certified claims to people to whom they would or should make a difference). As a contribution to a broader discussion about the role of science in society, Kitcher’s paper is particularly noteworthy for his insightful observation that it is not enough to concentrate on issues of the transparency of information and on potential distortions of public knowledge produced by interested parties shaping the field of inquiry instrumentally; rather, we also need to think long and hard about what democracy means in the absence of “an IIS whose standards of certification are widely endorsed as reliable.” Indeed, what he calls “hybrid epistemologies” — the idea that people might accept scientific findings only inasmuch as they are not in conflict with some revealed religious truth common only to members of their particular tradition — emerges as one of the significant problems facing democratic deliberation, as it directly undermines the idea of a single common set of certification standards.

It should come as little surprise to any reader of my book or of the various pieces of this discussion that I am in broad agreement with Kitcher’s suggestion to ditch the (Cartesian) isolated knower and his (it’s always a guy, isn’t it?) peculiar hang-ups and challenges related to developing true beliefs about a mind-independent external world. Knowledge isn’t a personal, subjective possession, even if one is a neopositivist; Karl Popper was pretty adamant that knowledge was common to the community, which was part of his solution to Cartesian anxiety, and basically no one after him would disagree in any significant way. And this in turn does mean that we have to focus on the social conditions of knowledge-production, albeit not to the exclusion of the properly philosophical-ontological aspects of the problem — so here again I agree with Kitcher that the certification subsystem is worthy of some serious attention. But then I start to diverge from Kitcher, since his entire position depends on the idea that public knowledge should be based on consensus, both a consensus about particular claims and a consensus about what makes those claims good ones. The entire problem with “hybrid epistemologies” (his term, remember, and a revealing one — he didn’t say “different epistemologies,” after all, which strongly suggests that he places the blame on the non-scientific part of the package) is that they don’t permit diverse citizens to come to overarching consensus about factual issues, and therefore (I can only imagine the shock and outrage in his tone) leaves these issues to be decided by political struggle.

There is something quite interesting going on here, and it has important parallels in the disquiet that my four interlocutors feel, in their different ways, with my deliberate decision to produce an account of contemporary IR scholarship that is content with a diversity of methodological traditions but that insists on the distinctiveness of the scientific endeavor. In Kitcher’s terms, I would definitely say that we need a certification system capable of distinguishing between scientific inquiry and other modes of human expression like art or religion, but I would blanch at the suggestion of inscribing a scientific finding “on the books” if that means that it is now irrevocably fixed or certain or immune to critique — even if, like anthropogenic climate change, there is broad scientific consensus about the finding. That could only be the role of science and of Kitcher’s certification subsystem if there was one unique set of standards that would somehow guarantee the scientific validity of results, and it is the major burden of my book to illustrate that this is simply not the case. As such, a term like “scientific consensus” needs to be used with extreme caution; at most it means “the consensus of (virtually) all of the practicing scientists,” and the decision to go with that consensus can’t be somehow portrayed as anything other than a practical one.

Indeed, it would be fair to say that my entire position on this issue is a deflationary position, and to further say that the main possibility that my argument is arrayed against is the possibility that the claim to scientific validity will be used to insulate an actor from the necessity to take responsibility for her or his action. (Note that I am concerned with the claim to scientific validity, not the claim that something is established in experience; one need not invoke “science” to explain why it’s a bad idea to jump out of an airplane in flight without a parachute if one wants to go on living, and in such a circumstance I fail to see how a discussion of the theory of gravity would contribute much to the deliberation.) “Science made me do it” is just not a valid excuse, since that’s a self-deceptive encoding of “I went with the scientists on this one” — and the less self-deceptive statement opens the possibility of actually deliberating the reasons why to go one way rather than another. Deciding to inscribe a scientific finding “in the books” is a political act, not a scientific one. As I’ve said before and will undoubtedly say again, nothing frightens me more than idealists with weapons, because they think that their use of those weapons is simply and unquestionably *right* in the pursuit of their ideals; this misuse of science as a crutch for an idealism that doesn’t self-identify as idealist but as somehow stemming from a de-transcendentalized truth simply equal to the way things are, and as such looks more potent than an honest statement of principle would, is what I am most concerned about. Scientific inquiry doesn’t resolve the highest and most profound issues we face, since those issues are political and aesthetic and ethical and perhaps even theological, even if actors cite scientific inquiry and scientific findings as they express themselves on these issues. I am reminded of Wittgenstein’s comment that it is striking how few problems are actually resolved when we learn to see philosophical and logical claims for what they are. Limiting science in this way — withdrawing the kind of blind faith that all too often accompanies the unmodified use of the term “science” — makes room for other things, room that we desperately need in a world characterized by immense diversity.

So if science isn’t for achieving global consensus, what is it for? And why am I so insistent that IR-as-social-science — remember that the whole book is set up as an answer to the question “supposing we wanted to study world politics scientifically, how would we do that?” and not as an answer to the question “why should I study world politics scientifically?” — be characterized by multiple incompatible methodologies, and not by a grab-bag from which a scholar may select different elements at will? Some of this is because the book is an ideal-typification of present debates, which both means that a) the only positions I discuss in detail in the book are those that are present in contemporary IR scholarship and b) because the exercise is fundamentally ideal-typical, the positions are mutually exclusive and logically incommensurate in ways that actual pieces of research tend not to be. And some of this is because I am betting that the best way to puncture the veil of mysticism that surrounds the vague use of the term “science” without abandoning the notion altogether is to illustrate the internal plurality of ways of being scientific. In addition, some of this is because of the sociological fact that the contemporary IR field is dominated by the discipline of Political Science, in no small degree because of the institutional location of most American IR programs and scholars within departments of Political Science, and as such is slanted in the direction of neopositivism; hence any call to mix or combine methodological positions is more likely than not going to tacitly privilege neopositivism when it comes to the fundamental design and epistemic status of the research project (the bulk of the “qualitative methods” movement in Political Science serves as a case in point here). The vocabulary of the philosophy of science, as long as it is vocabulary that celebrates pluralism in ways of being scientific, can be a great asset in maintaining the autonomy and independence of non-neopositivist ways of producing knowledge about world politics; “mixing” would dilute this position, and make adherence to neopositivist strictures the default ante for playing the game of IR-as-social-science.

But there’s another reason for insisting on multiple methodologies that can’t be easily combined, and it has to do with the notion of “scientific consensus” — and therefore with Kitcher’s certification subsystem. Because any argument about whether to side with the consensus of practicing scientists can’t be a scientific one but has to be articulated on other grounds (grounds which might be philosophical, as in Fred Chernoff’s defense of a modified Duhemian position, but need not be), we can legitimately ask what qualifies a consensus of practicing scientists as noteworthy enough to affect action. If there is rough methodological homogeneity among a group of scientists, then any consensus that they come to is quite possibly, even quite likely, the result, at least in part, of their methodological consensus — a consensus that precedes, at least logically and most probably also temporally, the substantive consensus in question. The fact that a group of people who already agree on a lot have now agreed on another thing in addition to their prior consensus does not strike me as all that significant on its own. But if a group of people who disagree on other fundamental philosophical issues of ontology and methodology reach a consensus about something: now that strikes me as profoundly significant. Without probing too deeply into the reasons for this consensus or trying to explain it — dualists would say that it reflects the intransigent pressure of the real world on our conceptions, monists would say that it graphically illustrates the power of certain assumptions to order our experience, transfactualists would claim evidence of deep generative structures, phenomenalists would celebrate the revelation of a discernible universal — the bare fact of the consensus itself is noteworthy. Anthropogenic climate change, which I mentioned earlier, might qualify as such a improbable consensus across divisions of philosophical ontology; the near-universal opposition of IR scholars of all stripes to the US-led invasion of Iraq in 2003 might also qualify. With multiple reasons not to emerge in the first place, a consensus has emerged, and that should be taken seriously.

Of course, a consensus like this is only unlikely to emerge if everyday scientific practice and ordinary science is dedicated to refining distinctive methodological traditions in their own directions. Let me be clear here: I am not suggesting that we adopt something like the pop-Kuhnian notion of “normal science” and urge that scholars work on minor technical puzzles within their own hermetically sealed paradigms, waiting for a scientific revolution to change the game entirely when their paradigms become brittle and worn enough. I am instead suggesting that scholars be cognizant of the logical incompatibilities between different methodologies as they set about doing their research on a myriad of problems and issues and even puzzles, and develop their accounts with sustained reference to the logical requirements of particular procedures of knowing. Scientific inquiry is supposed to be hard, so that a fact scientifically established has some claim to epistemic privilege; one of the key ways in which it is hard is the stringent insistence on internal consistency and logical coherence, particularly in the linkages between philosophical-ontological presuppositions about the mind-world “hook-up” and the eventual facts produced. Differences in such presuppositions mean differences in methodological orientation, unless one is either a) willing to sacrifice logical consistency or b) articulate an orthogonal solution that somehow transcends or sublimates those differences. Option a) places us outside of even broadly defined science, and into the realm of politics or engineering or other sorts of practical performances, in which results matter more than validity; refusing to do that means, in practice, that the scientific enterprise is quite frangible — we get a mosaic of diverse knowledge-claims rather than the seamless web dreamt of in more naive notions of progressive cumulation. And that in turn makes the rare moments of consensus that much more precious, and worthy of being heeded.

But what about option b), an orthogonal solution to methodological diversity? The challenge with such a solution is that it lies beyond the limits of philosophical reflection on methodology, since methodological reflection is an inherently reconstructive endeavor: our accounts of what makes a claim valid necessarily run behind the actual practice of making and validating claims. In Wittgensteinian terms, methodological reflection of the sort I pursue in my book is an effort to specify the rules of a game that is already being played, and no such set of rules can exhaustively define or specify the game — the best they can do is to give a sense of how the game is played that is useful for guiding the players. A rule is thus a resource to be used in the playing of the game, irrespective of whether that resource is being used in a constitutive or regulative or strategic or tactical manner. As such, a rule change can only function properly if it somehow captures a sense of how the game is supposed to be played; otherwise the community of players rejects is as somehow violating the “spirit” of the game (the ongoing furor over the “designated hitter” rule in Major League Baseball strikes me as a good example here). And the surest way to get rejected is for a rule to run way out ahead of how things are played in practice, which parenthetically is perhaps why so many philosophers are concerned to establish that their proposed rules for just societies or defensible truths are in some sense already implicit and immanent in everyday life. So the rules that I or any other scientific methodologist proposes are in that very important sense subordinate to ongoing practice, constituting both a reconstruction of that practice and — of necessity — an intervention into the ongoing flow of that practice itself. What I have done in the book is to take existing divisions within the scholarly study of world politics and artificially and ideal-typically sharpen them, in the hopes of making more space for different approaches to the production of scientific knowledge. Either the scientific practitioners — IR scholars who self-identify as scientists — find my version of the rules of their game to be compelling and useful, or they do not, and if they do, those rules make orthogonal logical solutions deeply problematic since those proposed solutions would be torn apart by the distinctions I have explicated. But if scientific practice itself runs out beyond my formalization, and if knowledge is produced by researchers deep in dialogue with the issues I have sought to foreground but equally deeply reluctant to accept my categorization, then two conclusions follow:

1) we need new rules.

2) my ladder has been climbed and kicked away, and I will be content.

I do not think that any of us are served by wishy-washy, vague, relaxed standards of scientific validity; I think that precision and intellectual rigor and sharp disagreements when called for are essential components of productive conversations, as they help us maintain the tenuous balancing-act between treating every claim as valid and treating only claims tendered in accord with a single restricted grammar as even potentially valid. I can’t defend that proposition scientifically, and I won’t try to do so. I will instead only say that a science wracked by methodological disagreement is a surer defense against possible abuses of epistemic authority; that emergent consensuses from such a science are inherently more worthy of attention; and that thinking space for contemplating the pressing problems of our time is better preserved by a set of rules that begin with diversity than by a set of rules that begin and end with conformity. Scientific practice may — and probably, hopefully, will — prove me wrong or at least outmoded in the future, but I hope that it does so at least in part as a reaction to the prodding I have provided with my methodological lexicon. But in the meantime, accentuating methodological diversity leaves more room for non-scientific endeavors (art, ethics, theology, and yes, even politics) to inform our actions, interacting with a chastened science in ways that humanizes the whole process. Feyerabend was, I think, quite right when he repeatedly pointed out that science and the philosophy of science should be our servants, not our masters; insisting on rigorous engaged pluralism is, I think, a most efficacious way of helping to ensure that this remains the case.

And lest you think I am exaggerating for rhetorical effect, remember the stakes. As scholars of world politics, we are or should be concerned with the biggest “we” imaginable: the whole world, insofar as it is marked by human social action. Someone has to witness that grand spectacle, and prevent too-hastily-established “truths” from foreclosing its future potential. Someone has to unseat fallacious claims advanced in defense of narrowly partisan interests. Someone has to remember the whole panoply as a human endeavor, as a creative endeavor, as a messy social and political process wherein we produce the knowledge that can help us address pressing global challenges. Others will do it if we don’t, but the unique and special potency of scientific research is that it transmutes values and philosophical commitments into facts through disciplined procedures of inquiry. At the end of the day, that’s our vocation: to make sure that this potency is not abused, to channel it into productive forms, to hold open the space for innovation by insisting on the boundaries and limits of logic and reason. We place ourselves between science and the rest of human endeavor, ensuring with our lives that science doesn’t tell people how to live — or that people don’t look to science for instructions on that score. And we focus on the facts — as Weber says, we “serve only the matter at hand” — in order to let the rest of the world learn to take care of itself. Here we stand; we can do no other. Who else are you gonna call?