After a long hiatus, I am ready to take up a number of the other points raised in the four Disorder of Things posts on my book The Conduct of Inquiry in International Relations.
II. ethics and politics, which are good things but not science
At the end of the first part of my reply I suggested that the “gang of four” interlocutors (Paul, Joe, Nick, and Meera) who commented on the book are quite correct to point out the importance of my professed but not expansively justified position on the irresolvable nature of value controversies to the overall argument, because many of the points that they raise — the absence of a unitary and uniform metric for scientific progress, my refusal to defend any of the four parts of my typology as uniquely scientific or to disqualify any of them as somehow not worthy of that title, my studied inattentiveness to the politics of methodology or to the implicitly value-laden character of seemingly instrumental definitions — relate quite strongly to my position that fundamental differences on points of philosophical ontology dealing with mind-world relations are not resolvable. If it were possible to actually resolve these controversies, then the engaged pluralism for which I call would not only be unnecessary, but would actually be a detrimental impediment to scientific progress. What, after all, would be the point of continuing to engage discredited positions? Although Paul Feyerabend suggests that we always need alternative research traditions to generate anomalies with which others have to grapple, this presumes a certain fundamental commensurability between claims which might be meaningful within a given philosophical ontology, but could hardly apply to those philosophical ontologies themselves. If we’re all neopositivists or critical realists or analyticists or reflexive scholars, then we have a common methodological basis on which to compare claims, and “discredited” research traditions can always be utilized as a source of claims and findings that provide a spur for continued innovation in the dominant approaches. But if we’re not all working in any one methodology, and if (as I have argued) methodologies are in effect different ways of worlding, this argument for diversity and pluralism falls short. So either there is a way to resolve methodological controversies, and we should all simply adopt the correct answer to the mind-world conundrum, or there is no such way, and we are fated to irresolvable and irreducible pluralism and plurality.
None of my interlocutors are entirely comfortable with this dichotomy.
Paul suggests that my four philosophically ontologies implicitly privilege particular kinds of substantive claims (true, and part of the exercise of my book is to make that privileging less implicit and more explicit) and as such tacitly support universal or at least global claims to resolving methodological controversies by demonstrating that their particular category in the typology is really the superior one (again, true, but what impels research in each methodology is not, I would say, a self-sufficient justification for the superiority of that methodology. I intended the book to deprive every methodology of a claim to superiority over all others on philosophical grounds, and to my mind the best way to deal with multiple claims to epistemic privilege is to adopt the same attitude as one does to ecumenical religious dialogue: yes, you think that God gave you a unique and special revelation, but so do we, and so do they, which makes everyone’s claim to uniqueness and exclusive divine sanction somewhat suspect). Nick draws on Robert Brandom and his impressive beard to suggest a novel approach to philosophical ontology (which, as I said earlier, sounds pretty darn analyticist to me … although it probably sounds pretty darn realist to others) which claims to resolve the controversies between monism and dualism, and to refute phenomenalism in favor of a limited kind of transfactualism (which is a neat exercise, but I am not yet sure what Brandom-esque research on world politics would look like, and in particular I am not sure that it would not look like one or another subtle variant on the existing cels of my 2×2 matrix — as I said, to me it looks like Brandom is wielding transfactual realism as an analytic, which is intriguing but is not likely to be accepted by critical realists). It remains unclear to me how would would even *recognize* a solution to the problem of methodological diversity, as any such solution would, in effect, simply collapse into a defense of and advocacy for one part of the conceptual landscape over and against the others, and thus collapse simultaneously into an expression or reiteration of a particular methodological position. The only way that something else might happen, I would say, is if there is sufficient drift and mutation in the broader cultural milieu that one or another methodological perspective (or one or the other of the dichotomies that I use in producing the matrix in the first place) simply ceases to sound plausible or compelling. This kind of shifting of the stream-bed (to use one of the later Wittgenstein’s metaphors for the process by which notions of common sense change) is not a rational or scientific process; we can observe and describe it, and maybe analyze it, but we don’t get far if we presume that this shifting reveals some march of progressive rationality through History (or maybe the problem is that we get too far too quickly, and end up in Hegel-Land contemplating the Absolute Idea in all its glory, standing next to Plato who is gazing straight into the sunlight of the Good). Ideas about methodology change, yes. But is that “progress”? Who can say? (And on what basis can they say it?) Better to focus on what we can actually ask and answer, which concerns the practical productivity of different methodological commitments, and leave the cosmic questions aside.
I want to proceed carefully here, because I want to be very clear on the differences — even though there are obvious similarities and parallels too! — between my pluralist stance on methodological diversity, and the analyticist perspective on methodology from which I self-admittedly do my empirical scholarship. Both are derived, in my account, from a blending of Weber, Wittgenstein, and American pragmatism, but there is no need that one be an analytical monist to accept the skeptical defense of diversity that frames the book as a whole. Joe is precisely correct to call attention to the Weberian sensibility of my overall approach, however, so I need to spend some time elaborating that and especially tackling the issue of some sort of critical ethical reasoning could provide solutions to the diversity of methodologies. This flows naturally into a consideration of whether a political solution of the sort advocated by Meera — political in the sense that methodologies are instruments for doing things in an unstable world rather than for producing knowledge about the world — is a better way to go. It should be no surprise to any reader that my answer on both counts is “no,” but clarifying why will take some discussion.
The procedure of ideal-typification that I describe in Chapter 5 is, of course, a process designed to generate substantive claims about the world, the epistemic status of which depends both on their artificially systematic (idealized and abstracted) nature, and on their necessary grounding in the sphere of cultural values (a grounding that differentiates ideal-types from merely logical “pure types,” which would affect a “view from nowhere” rather than acknowledging their dependence on value-orientations. Reproduced here is the table from Chapter 5 in which I delineate the complexity of the procedure, but also its goal, which is to connect values to facts in a way that leaves the logical and philosophical distinction between the two intact.
The application of ideal-types to worldly conundrums is what produces factual knowledge, but these facts are only completely acceptable to someone who both appreciates the internal consistency of the process *and* shares the relevant value-orientation; Weber is quite clear that a “Chinese” (Weber’s rather orientalist term for a contemporary as far removed from Weber’s own Western European cultural context as one could possibly be) could appreciate the technical correctness of a demonstrative argument based on ideal-types, but that’s as far as he goes, since the “Chinese” would lack an inner feeling for the correctness of the value-orientations on which the relevant ideal-types were based. And this, for Weber, is precisely the importance of the procedure of ideal-typification and the resulting social science: instead of an irresolvable clash about values, we have a somewhat calmer and more pragmatic discussion about facts.
So what of the ideal-typical 2×2 matrix that provides the basic scaffolding for the book? Matters are a little different here inasmuch as I am doing something that Weber would never have countenanced, which is to ideal-typify not social objects and processes, but methodological orientations. For all of his insights on methodology, Weber did not consider himself a methodologist or a philosopher of social science, and (like Charles Tilly, whom I would argue and have argued was one of the great Weberian social scientists of our time) would have pressed for arguments about stuff rather than arguments about arguments. A proper Weberian ideal-type is equipment for producing facts about the world, and the matrix in my book is not itself a fact about the world. Rather, as Dan Nexon and I suggested in a recent article, it, like the ideal-typical “IR diamond” we generated in that article, is a way of mapping or organizing scholarly arguments. It could be part of an empirical analysis if combined, configurationally, with other ideal-types about scholarship, and used as part of an account of how and why notions of knowledge change or how they came to be the way that they are presently institutionalized, but on its own my matrix of methodologies is not intended to be a contribution to worldly knowledge. But like Weberian ideal-types properly understood, my map of methodologies (like Dan and my map of IR theories) is rooted in a value-orientation and cannot be fully grasped apart from that value-orientation. That orientation is, broadly speaking, pluralist, and pluralist of a specifically humble (better: anti-hubristic) variety: the most basic things that animate our talk about are not up for discussion, and when we appear to be discussing them we are either simply expressing our positions or are engaged in what Wittgenstein called a special form of persuasion that aims to give someone our worldview — which is not, strictly speaking, “persuasion” but is more like *conversion* because of the unavailability of a shared common standard to which worldviews could be subjected (which is a logical point: if we had such a shared common standard, then we’d already share a worldview. QED.). This is not to say that these basic commitments are somehow transhistorically constant, either for individual people or groups of people or for the species as a whole, but that change in these basic commitments does not come about because of reasoned argument; instead, some other form of social process is implicated, and even if arguments are part of the story it is their functional and causal impact rather than their formal and logical character that is relevant. (I’d further suggest that basic commitments, or what I have elsewhere called commonplaces, are themselves intrinsically ambiguous, so it is simply not possible to exhaustively spell out the implications of a given commitment in advance. The process of specifying just what commonplaces in combination imply in terms of concrete social action is something that we can study empirically, but in any event no purely rational resolution of such issues is possible — nor should we naively expect the Habermasian “unforced force of the better argument” to prevail in any actual situation, even as a transcendent ideal subtly working to bring about the possibility of a rational consensus.)
I have no way to persuade you of any of this. The best I can do — and by “best” here I am ruling out the coercive techniques of trying to force you to adopt this way of worlding, this value-orientation towards an ever-present awareness of the partiality and the limitations of knowledge, since to do so would be performatively contradictory in the extreme — is some combination of sketching what a world founded on that value-orientation looks like in practice, and demonstrating the social-scientific utility of ideal-types formed out of that value-commitment for making sense of pressing and puzzling worldly situations. If I could persuade you, I could only do so on the basis of appeals to already-shared standards, and if I could do so in this instance all it would reveal is that in some way you *already* agreed with me, or at least already shared (in a rough sense) this value-orientation. Now, my account might circulate and become part of the mental furniture that we use to make sense of things, if people find it useful — there’s the pragmatic part of my stance, since the “cash value” of an account is what it does in practice — but this wouldn’t be the same as rationally persuading you. Indeed, here’s where I think that sociological studies of scientific knowledge can be at their most insightful, since they set out to explain how things become accepted and taken-for-granted without having any recourse to the (undemonstrable, and indeed, pretty much purely utopian) notion that statements are believed because they are true. From which, yes, it follows that if you don’t buy the value-orientation out of which I have built my ideal-type mapping of methodologies, then you are not likely to be convinced of my repeated claim that methodological questions dealing with the mind-world relation are not rationally resolvable, since this claim is more or less a derivation from the initial value-commitment, and hence not really an empirical claim in the first place. It’s important to properly parse statements into their appropriate logical register; ideal-types and the value-commitments affording them are expressed in statements that are expressions of principle, not of fact, and they are means to the end of social-scientific explanation, not conclusions in themselves.
Let me briefly contrast this with Joe’s call for a form of social science that “responds to pressing social concerns and is explicitly oriented toward social reconstruction that enlarges our experience, in all its protean diversity.” On one hand this formulation is quite similar to mine, and I find a lot to like in it, especially its Deweyian innovation of experience as the ground and goal of knowledge. In fact I am unsure, in the end, whether this formulation is actually any different from the value-commitment animating analyticist monism, so although Joe’s personal position in the pantheon of social science might be (on his own account!) unclear, his sensibility is certainly continuous with one of the boxes of my typology. I would go further, actually: I am not sure that what Joe forwards as a form of ethical reasoning or ethical scholarship is, in my terms, concerned with ethics, and thus his reaction to my denial that the category “ethical knowledge” makes much sense is, perhaps, more of a semantic issue. Joe protests that ethicists are systematic, produce claims that can be publicly criticized, and are concerned with generating knowledge of the world; I’ll easily grant him the first two, but the third strikes me as problematic because in order to for a claim to actually have ethical force it can’t simply be rooted in the world (regardless of whether we think that world is made up of phenomenal objects or whether we think it also includes undetectable causal powers and emergent structures of activity) but has to come from, in some sense, outside of the world. This is especially the case for norms of obligation (“thou shalt not kill”), as their obligatory character has to be in some sense transcendental, but I think it is also the case for norms of shall we say lower resolution (“don’t cut in line”) because the normative force of such admonitions is reducible to a core imperative principle (some notion of fairness, in this instance, even if we mask the appeal to fairness behind putatively empirical notions of overall efficiency and average waiting-time) and for non-obligatory norms of courtesy and accepted social practices (“you should give up your seat on the train for that elderly person”) which, again, derive their moral force from some broader notion about the good of the existing social order. Any normative claim rests on some transcendent (“otherworldly,” or perhaps “extra-worldly”) ethical principle that can impel action, from which it follows that establishing such principles is either:
1) a philosophical or theological exercise, in which case it bears more in common with the creative construction of an ideal-type out of existing raw cultural materials, and the depiction of the relevant principle in sufficiently compelling detail that it strikes a responsive chord with the reader. Important, interesting, but not science, because it’s not worldly knowledge.
2) an exercise in application, in which a taken-for-granted norm or principle is used to animate an investigation of how best to achieve that principle in practice. This subtle shifting of the question leads us to focus not on the principle itself, but on the various forms of practice and organized social action that might encourage that principle. Now we’re in the realm of social science, but we’re no longer in the realm of ethical inquiry strictly speaking. “Applied ethics,” perhaps, and that would be just fine as a science on my account.
3) an attempt to derive principles and standards from already-existing common practice, i.e. the Walzerian project, or the “virtue ethics” variant from Macintyre. This is intriguing stuff animated by one very serious conceptual transformation: to observe that practice has in-built standards that have normative force for the proper conduct of that practice is to suspend any notion that the relevant practice could be criticized on ethical grounds (unless, of course, one situates the practice in a bigger and broader practice, and then uses standards from the larger sphere to critique the smaller sphere, but this just pushes the issue back a level and does not solve it). And relying on the reconstruction of a practice also runs straight into Wittgenstein’s insight that the formal rules are not self-sufficient, and as a result any reconstruction of the standards implicit in practice will be contestable, which contestation becomes its own game with its own tacit rules that can’t be formally summarized … In any event we’re now outside of worldly knowledge, and into a realm that can only be described as intervening in the world to change it. And no amount of argumentative demonstration that a given practice is animated by principle or standard X will ever suffice to prove that standard X is in any transcendent or other-worldly sense ethically justified, so arguments about what one should do fall on deaf ears if one is speaking to people outside of a given practice (or “form of life,” to use Wittgenstein’s terminology).
Joe’s commitment to the expansion of experience, like my commitment to the preservation of philosophical diversity, is not an argumentatively demonstrable one. I rather like Joe’s commitment, but the fact remains that it — like my attachment to diversity — inhabits a realm of philosophical/theological/ethical commitments where argumentative reason falls silent. In the strict sense, then, it is not possible to *know* anything about it, at least not in the same sense that we know things about alliances and colonialism and arms races and the like. But it does not follow that all such commitments are “subjective” — far from it, since if commitments were subjective they’d fall prey to the Wittgenstenian argument against private languages. The conceptual slippage between “arbitrary” (in the sense of not being grounded on anything outside of itself, precisely the same way that Heidegger talks about sentient beings being their own basis, and Derrida talks about language as incapable of being reduced to anything nonlinguistic) and “subjective” (in the sense of being dependent on the individual subject, a formulation that has close conceptual affinities to “relativism” and sounds a lot like “making shit up”) gets us into trouble here, since it excludes precisely the notion that we need to make it through the morass: “intersubjective.” A robust conception of intersubjectivity dissolves the “problems” of subjectivism and relativism, by disclosing thinking as an intrinsically social process; hence the idea that a single individual could, even in principle, take a completely distinctive position is grossly flawed, as all such positions depend intrinsically on a broader set of cultural and discursive resources on which people draw in articulating their positions in the first place. Value-commitments are not just something I invent in the presumed privacy of my own presumedly autonomous mind; instead, they are public performances of a value-laden identity, a kind of stand-taking that only makes sense in relation to others. Yes, it derives its power to animate my actions from the fact that I have committed to it, but that doesn’t make it “subjective” — especially if one is a part of a scholarly community where one’s work is continually, at least in principle, open for discussion. We need not always agree on our values and value-commitments, but we do have to articulate those commitments in a structured present within which not all possible resources and points of view are in fact in evidence; this alone, I would argue, is more than enough to keep the spectre of subjectivist relativism at bay. But it does not allow the further step that Joe tacitly calls for, which is some kind of argumentative resolution of value questions. We need not be monists to accept this limitation, and we do not need to embrace social differentiation as a product of some uniquely modern or capitalist arrangement in order to accept value diversity.
Thus, when Meera points out that my articulation of a broad notion of science “is not an argument grounded in the special epistemological status of inquiry but something like an appeal to the intersubjective agreement about what it feels like to conduct Weberian-style inquiry — more or less a type of orientation to one’s work,” I can’t help but agree. A defense of the special epistemic status of science would, I think, necessarily tip over into a sharp delineation of what science is and should be, which is both precisely what neopositivists and critical realists have been doing in the field for years, and what Meera herself wishes that I would do in the name of an appreciation of ontological instability engendered by reflexive self-awareness. I won’t, both for the philosophical reasons I have been spelling out, and because I am not convinced that a field dominated by a reflexive notion of science would be much better than a field dominated by a neopositivist one. I do not think that any of the four ideal-typical methodologies I sketch in the book are any kind of indispensable component of “being a science,” and even if it privileged me I would not want an IR field in which analyticists were dominant and others were forced to survive as best they could. In this sense, Meera slightly misreads my reference to the Ackerly et. al. volume on feminist methodologies; I intended to praise that book as a first stab at something I want to see more of, namely explicit IR feminist wrestling with the question of what it means to produce valid knowledge of world politics. There are of course a lot of different versions of what “validity” means here, but that’s precisely the point: neopositivist validity and reflexive validity aren’t the same animal, but we have a lot of work trying to define the former and not a lot trying to define the latter. I hope that this situation changes.
But this hope, as Meera is right to point out, is itself grounded in an embrace of the value of scientific inquiry, albeit in my broadened sense. I understand Meera’s critique to, in effect, press towards a more skeptical stance on the value of science, and to forward a conception within which science and politics are more closely linked — in this instance, underpinned by the ontological instability of the world. But this only follows if one adheres to the old, narrow definition of science that I am hoping to broaden. Meera at one point comments that “The ‘fact-value’ distinction within an analyticist framework requires that this is sustained by a valid and distinctive procedure for checking these ideal types against concrete actuality, although as I have suggested this is logically useless without the supposition of an empirical world that is independent of the mind,” but this is not actually the analyticist argument since one does not check ideal-types against concrete actuality as much as one uses those ideal-types to make sense of concrete actuality, and never does so by pattern-matching or falsification. If science necessarily meant knowledge of a mind-independent external world, then notions of monistic science would not make sense, but since I don’t accept the premise, I don’t accept the conclusion. Instead, I would continue to argue that there is both a logical and a practical distinction between knowing about things in the world and affecting change in the world, precisely since the orientation towards affecting change as the primary goal of the exercise gives up the first of my three criteria for science — internal consistency — even though it retains the other two (publicity and worldliness). What matters in politics is what works, regardless of logic. But in science, what works for an explanation can only work if it is first and foremost internally consistent, which is why social theory and social activism are separate exercises.
Now, since I have denied that one can warrant the value of science on the simple grounds of the production of results (and for IR this is a good thing, since we have no results to speak of!), how should it be warranted? We should not overlook the cultural prestige of “science,” and we should not ignore the proud tradition of science, especially social science, as introducing “uncomfortable facts” into political debates and the educational experiences of generations of students. But in the end I find that the most compelling justification for keeping science separated out is an ethical argument: the use of coercive force ought not be justified by the appeal to some kind of non-theistic certainty, and forcing those who would advance such justifications to confront the arbitrariness (in the technical sense I introduced above) of the positions that they take up. The positive vocation of science might be to articulate and introduce findings and facts and the conceptual apparatuses that produce them, in the hope that they might inform subsequent discussions, but the negative vocation may be even more pressing in our secular age: depriving political idealists of the claimed basis of their recommendations, and refusing to them a sanction as absolute in its way as divine sanction, but more in tune with an age in which direct appeals to the divine will (American presidents to the contrary) don’t seem to carry as much weight. You cannot make politics pure by purporting to ground it on scientific truth, and an important part of our job is to try, as much as possible, to prevent politicians from doing so. The condition of possibility for that, in turn, is a separation between science and politics.
Next up: part the third, “don’t cross the streams, but do talk about doing so.” Realistically, I won’t get to post this for another couple of weeks.