This is the first in a series of posts by several of us at The Disorder Of Things on Patrick Thaddeus Jackson‘s The Conduct Of Inquiry in International Relations: Philosophy of Science and Its Implications for the Study of World Politics, released last year to considerable critical acclaim. The next few weeks will see further posts, followed by a reply by Jackson (or PTJ) himself. Taken together, we hope they go some way to meeting the challenge, to paraphrase PTJ himself, of emerging from splendid isolation to engage in some contentious conversations on inquiry.
UPDATE (24 Jan): Joe’s call to free the pluralist imagination is now up.
UPDATE (3 Feb): Nick’s speculative realist examination of inference, progress and materialist monism is now available.
UPDATE (17 Feb): Meera’s further untangling of ‘science’ and challenge to the stasis and status of reflexivism completes this round of responses.
UPDATE (14 March): PTJ has begun his reply.
Conclusion: neither science nor the methodology of research programmes provides arguments against anarchism. Neither Lakatos nor anybody else has shown that science is better than witchcraft and that science proceeds in a rational way. Taste, not argument, guides our choice of science; taste, not argument, makes us carry out certain moves within science (which does not mean that decisions on the basis of taste are not surrounded by and entirely covered by arguments, just as a tasty piece of meat may be surrounded and entirely covered by flies). There is no reason to be depressed by this result. Science, after all, is our creature, not our sovereign; ergo, it should be the slave of our whims, and not the tyrant of our wishes.
Paul Feyerabend, ‘Theses on Anarchism’ (1973), For & Against Method (with Imre Lakatos)
I: Conducts Of Inquiry
Paul Feyerabend’s Dada-ist approach to the philosophy of science was informed by a hostility to singular conceptions of the world, a rejection of rigid prescriptions for the correct character of knowledge and a healthy scepticism towards how people thought their forms of inquiry worked. Setting himself against method, he argued that ‘science’ has no common structure and that no general laws can explain its success or prescribe its methods. The growth of knowledge has not only historically been associated with a disrespect for prevailing methodological rules, it in fact requires such violation (whether deliberately or unwittingly).
Patrick Jackson’s The Conduct Of Inquiry attempts both to expand and to limit the knowledge practices counted as legible and legitimate in International Relations. Like Feyerabend (and Imre Lakatos and Thomas Kuhn and Karl Popper for that matter), Jackson evinces a healthy distrust of hard, fast or easy judgements about truth, reality and research procedure. And so he both avoids making the argument for a particular philosophical perspective with which to forge ‘better’ IR and eschews usual schemas that start with the putative ‘fundamentals’ of ontology and epistemology. Instead he offers a Weberian ideal-typification of methodologies (not methods), intended not as a menu of coherent and precise options but as a useful typology of functional categories (on which more in a moment).
The conception of methodology proffered is an expansive one, taken to designate ‘the logical structure and procedure of scientific inquiry’, but not the actual nature of reality and knowledge or the correct technique for a particular problem. Four commitments to philosophical ontology organise the options, generating four methodologies through their various combinations. The purpose is not somehow to test between them, but to reveal their assumptions and explain how they hang together as philosophical wagers. Hence the need to discuss the forms of faith and commitment that under-gird our styles of research.
So, in one sense, we could read PTJ as speaking with Feyerabend in demonstrating that “all methodologies, even the most obvious ones, have their limits…like an undercover agent who plays the game of Reason in order to undercut the authority of Reason (Truth, Honesty, Justice, and so on)“.
This intervention is much needed. How refreshing to see an acknowledgement that IR’s approach to science has been the opposite of that taken by science’s major philosophers. For the discipline, the concern has largely been to elaborate correct procedures and foundations in the expectation that this would consequently lead to success (by whatever metric), whereas for the likes of Lakatos or Kuhn the success of the sciences was hardly at stake (was in fact obvious and uncontentious) and the problem was a retrospective one: how to explain that success and reconstruct it in such a way that might provide a particular method. These were problems which they took to be vexed and complicated, even (or perhaps especially) when dealing with the progressive might of something like 20th century physics.
We are not experienced to encountering the language of science in IR at this level. We prefer the technocratic deployment of tools, a quick move from questions about how and what we can know to the bold neophilia of policy applicability. Behold: a mathematical model of counter-insurgency! An equilibrium proof of civilian victimization! A system for predicting political events available to the highest bidder! Proponents and self-conscious dissidents alike prefer a simplified schema, one which allows the enactment of disciplinary oppositions, like the one between those who naïvely believe in the fact/value distinction and those caught in a hideous relativist vortex, unable to differentiate the existence of a thing from its desirability.
So, in case there was any doubt, a dense and often thrilling book, well worth your time and effort. Thinking in Jackson’s terms can certainly alleviate those ills and begin the pluralist conversation. Whether it does or not will likely have less to do with the substantive content of what he proposes than with the institutional character and attendant incentives of a divided field. I suspect it will redouble (or retriple) the commitment of those of us already interested in the philosophy of social science but also confirm the anxieties and suspicions of those who see it as an over-complicated distraction from thinking about politics proper.
II: Against Methodologies
That is the expansion (and the O.B.N.). But what of the limits? First, a definitional point. Jackson half-discards and half-preserves the category ‘science’, a privileged designator we might well do without. I prefer ‘inquiry’ (and will generally use it to refer to what PTJ calls ‘science’) . In one way, this is an easy critique. Any work can be undermined by quibbling over its definitional architecture rather than by evaluating its substantive claims. As some critics might suggest, any postulation of science implies non-science and hence involves a boundary operation. But the same goes for inquiry, knowledge, ethics, or any other descriptor that we deign to self-apply. So the problem is not so much one of distinctions or even disciplinary manoeuvres, but rather of their content and the consequences (political or analytic) they entail.
So what of the content and consequences of speaking in terms of methodology? Despite Jackson’s claim to be setting out the kinds of analysis without seeking to particularly praise or denounce any one, there are many critical interventions in The Conduct Of Inquiry, whether on the methodological similarity (if not unity) of large-n quantitative and small-n qualitative research, the ‘red herring’ of detectable unobservables in critical realism or the results of confusing different conceptions of mind (as subjective consciousness, intersubjective consensus and/or a materially emergent sense of self).
This kind of critical argument is fairly inevitable, but it gestures towards a possible unravelling of the distinction between methodology and ontology/epistemology, the latter having been rejected because they tie us to specific claims and perhaps then to universal standards which cannot be said to exist ‘in any intellectual defensible way’. Put differently, the conceptual architecture risks both being too wide (because we might well be pushed into reasons for excluding some methodologies) and too narrow (because a four-way typology may not bear the weight of the multiple philosophical problems it is required to contain, even if only ideal-typically).
How so? PTJ is surely right to oppose the idea that we first figure out our view on ontology, then progress to epistemology, and only finally sort the methodological details of how we inquire. Jackson’s response is to include ontologies within the methodologies, but to adjure an account of which ontological position best captures the nature of the social world. But saying that ontology doesn’t come first is not the same as saying that ontology isn’t involved or that a resolution of the ontological contest isn’t necessary. A typology of methodologies may help us in certain ways, but it cannot banish ontological quarrels. Moreover, if we treat the four methodologies as simply appropriate to different questions and problems then we do reduce them to methods and tools, and not to more fundamental accounts of how it all ‘hangs together’. And this is certainly not what is intended.
The difficulty is that Jackson rejects grounds for deciding which methodology is best, or at least concludes that such grounds can only ever be provisional, since no one can really speak of the constitutive basis of reality with sufficient authority. At the same time, he is not suggesting that we can all simultaneously practice as neo-positivists and analyticists. In PTJ’s view the methodologies are sufficiently well contained to resist meaningful comparison at the level of philosophical wagers and faith. And incommensurable doesn’t mean interchangeable.
Take the distinction between philosophical and scientific ontology that PTJ adopts. Philosophical ontology deals with the prior, and more fundamental, claims about the ‘hook up’ between access to knowing (mind-world dualist or mind-world monist) and kinds of knowing/observing (phenomenalist or transfactual). Scientific ontology deals with the objects that matter in the social world (like structures and agents or cultures of anarchy), the ‘bestiary’ of a given worldview in Jackson’s attractive formulation. The relationship between philosophical and scientific ontology is such that being a mind-world monist at the philosophical level need not dictate certain forms of inquiry (participant observation or the like) at the level of scientific ontology.
The point is well-taken in so far as we frequently confound the registers and levels at which our disagreements take place, but leaving it at this bold disarticulation doesn’t seem quite sufficient. Surely it is true to say that while mind-world monism in the philosophical sense does not dictate the use of particular methods, it nevertheless restricts the range of possible options, and makes some techniques implausible as producers of knowledge. Most obviously, covering laws are unlikely to persuade a mind-world monist, just as mind-world dualists in the neo-positivist mode will always find analytical narratives (which for Jackson involve a certain singularity) ultimately unsatisfying and partial, relative to the kinds of explanations they take to be actually possible.
So, although ontology, in both senses, doesn’t ‘come first’ it is nevertheless implicated in a way that risks damage to the programme of a genuinely pluralist science of IR dependent on methodological diversity but not ontological or epistemological disputes. The relationship between methodology as an organising criteria and ontology, epistemology and methods blurs and flickers. PTJ stresses, for example, that mind-world monists can deploy any model of behaviour “without hesitation” since “philosophical ontology determines many things, but it does not generate any substantive expectations“. But mind-world monists of the mould that Jackson describes do seem constrained in this sense. A model of human behaviour which included an idea of mental processes as merely the preference orderers for objective threats and opportunities may or may not aid a neo-positivist in the construction of general covering laws about the incidence of war. But it would seem to violate basic wagers for analyticism – which, in Jackson’s words, “must terminate in a case-specific narrative” (surely a rather specific research method) – and for reflexivism – which is committed to an account of being that involves historically-located experience.
It might be claimed in reply that this hypothetical model of mind is not really a ‘model’ or a ‘tool’ at all but rather a fully-fledged ontological position masquerading as one. This ambiguity is the problem: where do techniques end and foundational assumptions begin? For Feyerabend, the defence of science in terms of consistency with given facts involved a similar dilemma. The consistency condition foundered in practice since:
the material that a scientist actually has at his disposal, his laws, his experimental results, his mathematical techniques, his epistemological prejudices, his attitude towards the absurd consequences of the theories which he accepts, is indeterminate in many ways, ambiguous, and never fully separated from the historical background. It is contaminated by principles which he does not know and which, if known, would be extremely hard to test. Questionable views…may invade the observation language itself, constituting the observational terms as well as the distinction between veridical and illusory appearance.
This kind of conclusion, which resonates at elemental philosophical levels, animates existent attempts to provide an account of what the social is and to designate appropriate and inappropriate conducts of inquiry accordingly . A methodology does not predetermine results but it does seem to invoke ontologies and epistemologies and to narrow the options for research tools. The anxiety produced by the threat of necessarily bounded knowledge (in the example of the analyticist and reflexivist focus on social practices) seeps from the realm of scientific ontology into philosophical ontology and outwards. And so research projects cease to be about the tools best fitted to social objects and engage the character of the social world itself.
What isn’t clear in all this is whether PTJ disagrees. As well as defending the separation between methodology and method, and between philosophical and scientific ontology, he also tells us that a position like monism or dualism “both affords certain strategies of knowledge production and rules out others”, but that this is an incomplete process. At the beginning of the chapter on reflexivism he seems to say that particular wagers generate particular problems which then generate particular “concrete practices of empirical research” and, by the end of The Conduct Of Inquiry, four methodologies have given rise to four different statuses of knowledge (evaluated by four different procedures); four different notions of cause (explained through four different procedures); and four different types of comparison (for four different purposes). These links seem much more solid and ideal-typically neat but again threaten the stability of an unbreachable distinction between ontologies, methodologies and methods/tools.
III: Political Demarcations
The problem is that arguments can only be judged by the standards of their methodology, but that the methodologies themselves can only be evaluated pragmatically, which on PTJ’s Weberian account means “whether, once applied, the ideal-type is efficacious in revealing intriguing and useful things about the objects to which it is applied”. The work follows Lakatos in the ‘rational reconstruction’ of methodologies rather than their flawless reflection. And so it is to be expected that readers will find ways to challenge the demarcations and to dispute the placement of thinkers. Is Waltz really an analyticist if he claims that structural realism cannot be used to explain particular foreign policy decisions (which surely count as objects of singular causal analysis)? Moreover, at times the methodologies (or their wagers) don’t seem so incommensurable, partly because while the broad strokes of the ideal types are clear, there are many shared problematics and varied responses which cross their boundaries.
We might take from both reflexivist and analytical traditions if we thought transfactualism somewhat more controversial in feminist, post-colonial and critical work than PTJ implies, or considered the dispute over the meaning of ‘social facts’ or ‘structures’ between reflexivist Marxists and Critical Realist Marxists illusory, or if we found the procedure adopted by many self-identified critical thinkers in terms of facts/values not so far from the Weberian model claimed for analyticism. Such admixtures indicate the productivity of focusing on a series of philosophical and conceptual disputes and a continuum, or wide menu, of responses to them, combinable in a number of complex forms. Might we then propose a parallel account of inquiry, one which parcels the philosophy of social science into a series of sometimes connected, sometimes disparate problems: problems of knowledge and verification and ontological status, often familiar, but which can be articulated or forced together variously, and in overlapping ways? The threat of multiplicity enters.
In concluding on the relationship between science and politics, PTJ claims argues “the operation of scientific inquiry means taking value-orientations and setting them to work in explaining things in the world…mak[ing] possible a mediated form of contestation over value commitments and their implications”. The value-orientation animating The Conduct Of Inquiry seems born of the experience of encouraging graduate students to think seriously about the status of knowledge, and to encourage the IR community, particular in the US, to think outside of a neo-positivist framework.
The pluralist agenda offers a way to broaden those intellectual horizons and reshape those disputes. One of the things really brought into focus by this project is the confusion in IR over the level at which disagreements about the conduct of inquiry occur. Often, depth understood in terms of the philosophical foundations of a position (usually ontology) is taken, metaphor-like, to correspond to depth understood in terms of the seriousness or importance of the disagreement. Yet rejections of a particular way of understanding intervention, or war, or state formation, are seldom substantially fought in terms of absolute claims about what kind of object the social is at root, still less about some universally-applicable way of interpreting and measuring all known phenomena. They are claims about that particular phenomena, that there creature from the bestiary, and usually claims that are seriously tangled up with other commitments not reducible to philosophical bases or empirical disagreements; commitments such as the idea that a particular research style is intended mainly to make a social phenomena manageable, and thus controllable in political terms by certain privileged actors .
In that last example, the submerged issue might even be a fear that the opposed approach is correct, that is to say that, unless stopped, it will in fact deliver advantages to certain actors in a way likely to result in undesirable outcomes for the people concerned. Much squabbling could be avoided were we not so readily to translate very real disputes over all kinds of questions onto some final ontological level. Since questions about the nature of knowledge and the conduct of inquiry are also difficult conceptual questions, this has the further advantage of providing some wide thinking space. No one should be too ready to rush to an easy identification in the belief that it will facilitate a particular research project or outcome. Stressing the relative autonomy of domains means that wishing to deal in the same substantive subjects or even categories as Marx (classes, alienation and the like) in no way requires that one adopt the same philosophical ontology or methodology as Marx. This is what PTJ suggests. But we can do this as well as see this autonomy as a relative one, and so pay attention to the kinds of analytical and political options that such basic commitments shape and limit (what we might call the politics of meta-theory).
The Conduct Of Inquiry thus seems caught between politics and methodology. In elaborating a methodological pluralism, it has sought to move the discipline away from the ‘holy war’ of science and the hegemony of a particular brand of inquiry. This is in many ways a political project, and one which is progressive in opening up institutional and intellectual space. But its coherence rests on a limiting move. Although the categories are announced as functional, they are also somewhat final. By adopting a frame in which we cannot settle the ontological status of the world itself (for example by showing that mind-world monism is a more sensible and justifiable lens that dualism), the undecidable stage for methodologies is set. The pragmatic warrant for the initial divisions hardens: “the only philosophical defensible response to methodological diversity is methodological pluralism”.
But methodology will not stay in its box. Ontological and epistemological arguments will not be exiled, and it is not clear that they should be. We have our tastes and we carry out our moves, seldom aware of their foundations, and the academic insurgencies and faux-paradigms we encounter today are the product of disciplinary and political history, not logical argument. In our attempts to say something useful, why not explore the places between methodologies or challenge the wagers of others? Whether for philosophical foundations or more concrete research programmes, a methodological eclecticism at least allows that we may combine or settle or renounce our confusions and commitments.
Paul Feyerabend also argued that progress in knowledge is not as driven by carefully delineated propositional content as we think. Instead, there is a considerable, if not overwhelming, role for non-argumentative growth, based not on semantic content or exposition, but on experiment, repetition or combination in generating new perceptions and new ways of knowing and being. This is partly why he thought no methodology could properly account for the “maze of interactions” that made the world so rich. That richness poses a special challenge for a field as expansive in its ambitions, as multiple in its concerns and as diverse in its inspirations as IR. The Conduct Of Inquiry has pushed, hard, at a locked door. The spirit of contentious engagement demands that we give it a hand, even if we forge new distinctions and disagreements amongst the splinters once we have broken through.
 In PTJ’s terms, IR can be science insofar as it is: 1) systematic (consistent in reasoning from presuppositions to conclusions, which makes understanding and evaluation possible); 2) open to criticism designed to improve knowledge; and 3) orientated towards the production of worldly knowledge (where worldly knowledge means the realm of arguable systematically demonstrable facts but allowing for different conceptions of what is included within the remit of ‘the world’). These definitional issues deserve further attention, not least because the formation of an adequate conceptual vocabulary was a primary motivation for The Conduct Of Inquiry in the first place. Able co-bloggers are likely to take up this theme in more depth, but I will say that, although I agree with PTJ that ‘politics’ and ‘normative theory’ count as distinct from some of the practices he identifies, I would put the divisions in somewhat different places and would consider two core elements of his definition of science (the requirements that we are systematic and that our work is open to criticism) as applying to both normative theory and politics (if in different ways).
 I’ve already gone on too long. For a classic statement see Peter Winch on The Idea Of A Social Science. For a more contemporary, and bold, attempt at something similar, I got a lot from Jason Glynos and David Howarth’s Logics Of Critical Explanation In Social & Political Theory.
 A further problem is that sometimes scholarship is just bad, even within the terms of its particular form of inquiry. Work which claims legitimacy from the discourse of Critical Realism or reflexivism or neo-positivism or whatever is hardly guaranteed to amount to a good example of it just because it claims that honour. Just as much disagreement could be avoided by a more careful understanding of the philosophical issues, so too could relations run smoother if we didn’t take the worst examples of other methodologies to be representative of their respective wholes.
Paul Feyerabend, Against Method (London: Verso, 2010 )
Jason Glynos and David Howarth, Logics of Critical Explanation in Social and Political Theory (London: Routledge, 2007)
Ian Hacking, The Social Construction of What? (Princeton, NJ: Harvard University Press, 1999)
Ian Hacking, ‘Screw You, I’m Going Home’, London Review Of Books, 22(12), 22 June 2000: 28-29.
Milja Kurki, ‘The Politics of the Philosophy of Science’, International Theory: A Journal of International Politics, Law and Philosophy, 1(3), 2009: 440-454.
Imre Lakatos, ‘Criticism and the Methodology of Scientific Research Programmes’, Proceedings of the Aristotelian Society, 69(1968-1969): 149-186.
Imre Lakatos and Paul Feyerabend, For and Against Method (Chicago, IL: University of Chicago Press, 1999)
Karl R. Popper, ‘The Logic of the Social Sciences’, in Theodor W. Adorno, Hans Albert, Ralf Dahrendorf, Jurgen Habermas, Harald Pilot and Karl R. Popper, The Positivist Dispute in German Sociology (New York, NY: Harper Torchbooks, 1976 )
Colin Wight, ‘Metacampbell: The Epistemological Problematics of Perspectivism’, Review of International Studies, 25(2), 1999: 311-16.
Peter Winch, The Idea of a Social Science and its Relation to Philosophy (London: Routledge, 2007 )
UPDATE (18 Jan): Two new paragraphs added for clarity in conclusion.
UPDATE (3 Feb): Following the example of colleagues, I’ve added a bibliography.