Editor’s Note: This is a guest post by Janice Bially Mattern. It is the 15th installment in our “End of IR Theory” companion symposium for the special issue of the European Journal of International Relations. SAGE has temporarily ungated all of the articles in that issue. This post refers to Patrick Thaddeus Jackson‘s and Daniel Nexon‘s article (PDF). Their post appeared earlier today.
Other entries in the symposium–when available–may be reached via the “EJIR Special Issue Symposium” tag.
Jackson and Nexon’s map of IR theory presents a different perspective on the field. The authors are optimistic that from this different vantage point can follow different kinds of conversations among International Relations (IR) scholars; ones that improve the quality of IR theorizing. I am less optimistic. Different though their map may appear, I worry that it will ultimately perpetuate more of the same kinds of conversations—including and especially those that legitimate the kind of “brittle” IR theorizing that Jackson and Nexon hope their map will challenge.
My skepticism ‘trickles up’ from the practical orientation through which Jackson and Nexon (J&N) construct their topography. More precisely, J&N approach IR theory as a practice—an activity, the logic of which is encoded in the routines by which it is done. The authors’ discussion of paradigms, great debates, and middle range theories is not just any standard overview of the literature but a practical analysis. Through it, J&N discern the dominant kinds of routines via which IR theorizing is done and then decode from those routines the implicit common sense logic of IR theory. Their conclusion is that IR theory is an explanatory activity that is done through routines that “relat[e] conceptual tools to empirical observations.”
This practical conception of IR theory moves J&N’s argument at multiple levels. First, it grounds their view that IR theory is relatively healthy. There is, after all, a “great deal of” explanatory work on world politics that is done by “relating conceptual tools to empirical observations”. Second, it gives shape to their topography. It enables J&N to delimit the universe of IR theory around explanatory function, and so, to recognize the variation in its routines. Those variations form the axes of contention and clusters of theory that make up the map. Finally, the authors’ practical conception of IR theory underwrites their optimism that this topography could help improve IR theorizing. It makes IR theories intelligible as scientific ontologies, or catalogues of conceptual terms that are analytically separable from those conversation-stopping categorical philosophical commitments that have long dogged the improvement of IR theory. In this way, J&N’s optimism about their topography echoes Adler and Pouliot’s about practical analysis more generally. Both imagine that the practice analytic provides a gluon that “cuts across” theoretical differences in the field.
But practices do not work like gluons. As I have argued here and here, practices discipline differences rather than cut across them. Practices are not just ways-of-doing things. They produce exclusive social spaces that are organized internally through normative hierarchies that establish and regulate deviance. As J&N are aware, the practice of IR theory is no exception. For instance, the “fairly predictable” routines used for middle-range theorizing have become a script, or norm of practical ‘competency’, for publishing in top-ranked American journals (6). Against this benchmark other routines for doing IR theorizing appear as inferior as do their authors and the journals that publish them. The overall effect is a hierarchical disciplinary system that establishes and regulates deviance through professional incentives (cf. Zarakol, Towns, Adler-Nissen, Rumellili).
J&N recognize this. Their hope, however, is that their practical topography can create the space for conversations that challenge the existing norms. Above all they hope it can upend the norms that legitimate “brittle” IR theorizing—that is, ‘quick and dirty’ theorizing in which scholars grab “off-the-shelf” theoretical mechanisms” for connecting input X to output Y, without concern for whether the assumptions and logic of theory is “bogus”. This habit, J&N argue, developed as middle-range theorizing became conjoined to neopositivist methods. But since the practical logic of IR theory pertains to scientific ontology and not to philosophical ontology, J&N’s the latter does not surface in J&N’s topography. The gluon-like hope is that this will enable debate about theorizing, that proceeds unencumbered by the analytically separable issue of methodological difference.
It is here that I become skeptical. Let’s grant, for argument’s sake, that J&N’s topography could launch a conversation in which analytically irrelevant methodological differences do not ‘pollute’ debates about theory. In such a scenario, the conversation—both its form and content—would be structured by J&N’s topography: that is, by the social space that is generated by the practical logic of IR theory it encodes. So what does that social space looks like? Where are its boundaries and by what logics are its organizational hierarchies structured? The answers to these questions are important for they indicate the kinds of conversations J&N’s topography can countenance. But what the answers suggest are a very familiar social space of IR theory and so a familiar set of conversations about it.
Recall that the common sense logic of IR theory encoded in J&N’s topography is (a) that it is an explanatory activity; and (b) that is done through routines that relate conceptual tools to empirics. From (a) follows the boundary of IR theory as a field. It is a rather familiar boundary: to count as normal IR theorizing must articulate an explanatory in function. In fact, J&N’s topography is arguably even more exclusive than the familiar status quo in that it suggests that theorizing which fails to articulate such a purpose is not merely deviant. It is flatly exiled as beyond the border of IR theory proper.
From (b) follows the logic of at least one key normative hierarchy. This is that since IR theorizing pertains to scientific ontology, the greatest value is attached to those theories that explain empirical observations by drawing only on conceptual tools that are scientific ontological in character. J&N endorse this explicitly in arguing for the exclusion of philosophical ontological issues from the conversation. The latter introduces analytically irrelevant differences that might ‘pollute’ the discussion and perpetuate the logic that justifies brittle theorizing. The same, then, must also apply to all other non-scientific ontological issues. To introduce non-scientific ontological conceptual tools into the process of explaining empirical observations is be theoretically abnormal; it is to enact deviance.
And therein lies the problem I foresee with J&N’s typology. Besides philosophical ontology I can think of one rather important set of theoretical routines that falls outside the scope of scientific ontology: conceptual theorizing. Understood as the work of signifying the substances and processes that make up the world, conceptual theorizing is what gives our theories integrity. It is what enables us to construct theories with defensible assumptions and terms; theories that are not bogus or brittle. But conceptual theorizing is analytically distinct from scientific ontology. It is a prior necessity for the latter but it is concerned with distinct theoretical activities.
Conceptual theorizing is already devalued, even deviant, within the practice of IR theorizing. Indeed, in only one year as an IR journal editor it has become painfully clear how many scholars—and not just those doing middle range theorizing—see conceptual theorizing as an irritating, unnecessary distraction. Given J&N’s aspiration is to reverse the trend of brittle theorizing, it is striking that their topography offers not ground upon which to challenge this attitude. On the contrary, it produces a social space in which it ought logically be just as deviant and ‘polluting’ to debate the conceptual integrity of an IR theory’s assumptions and logic as it is to debate methodological differences. J&N’s topography does not facilitate resistance to the habit of brittle IR theorizing. It colludes with it.
If such explains my skepticism about J&N’s optimism, there is a broader point I mean to make about the practice turn in IR. For some reason, practices are often represented in IR as formations in flux and as sites of change. This puzzles me. Practices are not just ways-of-doing something. They are self-reproducing social hierarchies. Just as the practice of IR theorizing appropriates Jackson and Nexon’s theoretical intervention to its own ends, so do practices in world politics appropriate the differences that resist them. Better IR theorizing, among practice theorists anyway, needs to consider this seriously.
I can see the logic in much of Janice’s critique, but I do take issue with one thing. The fact that we think social-scientific theory always–but not *exclusively*–serves an explanatory role does not preempt non-applied theoretical work, such as “conceptual theorizing.”
Thanks for posting, Dan. I agree that if we are talking abstractly about social scientific theory, in general, your point holds. But your piece is talking in concrete practical terms about IR theory *in particular* as you and PTJ have reconstructed it. And you do define the practice of IR theorizing around an explanatory function (which is why it can be separated analytically from philosophical ontology). I’m not entirely sure what you mean by non-applied theoretical work. But given your practical/explanatory/concepts-to-empirics characterization of IR theorizing, the activity of formulating the stuff that goes into our scientific ontology (i,e, what I am calling conceptual theorizing) seems to be just as analytically separable as philosophical ontology.
Maybe. Conceptual theorizing, as I read it, would take place at the level of scientific ontologies and at the “hook up” between scientific and philosophical ontologies, all of which is part of the practice of international theorizing in our account (see p. 551, for example).
You’re right that we are, in some respects, making a strong claim about “explanatory function.” I, at least, think that meta-debates that don’t hook into explanation *in some way* aren’t the core of what IR is about. But I’m not sure what kind of stuff we are excluding that we shouldn’t be. It can’t be “conceptual theorizing” per se.
You seem to be saying two things in this when you say “Conceptual theorizing, as I read it, would take place at the level of scientific ontologies and at the “hook up” between scientific and philosophical ontologies, all of which is part of the practice of international theorizing in our account.”
1. conceptual theorising is part of scientific ontology
2. conceptual theorising is the activity that hooks PO up to SO. It ‘fills up’ the SO
If #1, then my objection stands corrected. But this is certainly not clearly developed in your piece; and I can’t really find traces.
If #2, then this is my point. It indicates that conceptual theorising is a distinct register from SO or PO. As you write (p551) explanation involves both theory and methodology (and so now we might add conceptual theorising, or whatever we cant to call it) but, as you continue these are different registers that “are, or ought to be treated as, logically and philosophically distinct.”
The promise of your map to generate conversations that don’t repeat the same-old, same-old depends upon keeping those registers are that logically and philosophically distinct from IR theorising (relating concepts to empirical observations) out of the discussion. Hence…
So what kind of stuff gets excluded? This is the part that is so confusing to me because it seems to suggest something that I simply cannot see as something you would suggest. It seems to suggest exclude the work of theorising the theoretical terms that make up our scientific ontology (anarchy, state, war etc.) After all, we can’t do that–as Colin notes in a reply to me below– without implicating philosophical ontology.
But to bring it all down to earth, the scenario that came to my mind was of a job talk– on say, how democracy affects trade or how identity affects bargaining. I pictured asking the candidate to defend their conception of any of those concepts and being met with that kind of perfunctory, dismissive response that involves recounting a definition and then closes off debate about its acceptability/coherence on the grounds that lingering on such things is not the core of what IR is about.
It looks like conceptual analysis is getting over-privileged here. The tools for building (or “filling up”) a scientific ontology can also include statistical reasoning, experiments, etc. Concept-chopping has its limitations.
Keep in mind that I’m not sure exactly what you mean by ‘conceptual theorizing.’ This would be easier if you specified such that we could agree on the activities that you say are excluded.
That being said, the relevant point here is “and.” We say that these are logically distinct, but that doesn’t mean when formulating or arguing about scientific ontology philosophical ontology wouldn’t ever come into the picture. For example, I’ve seen plenty of debates that center around whether a particular scientific ontology is more or less comsistent with a shared set of philosophical commitments. YMMV.
What we are proposing with the typology is not an exhaustive catalog of debates, but emerging ones that seem to us very important, to cut across a lot of work, etc. .
I think Janice looked at the map and couldn’t find her own hometown. That’s likely to be a common way of evaluating the map. I think she forgot that it is not a map of “styles of theorizing” (conceptual, explanatory etc), but a map of ontologies. Now, the question is, can I find my own ontological hometown on the map …?
I’ll admit that I’m confused about what ‘conceptual theorising’ means and why it is excluded from this typology. If we understand our project as, ultimately, a practice of explanation, that doesn’t exclude conversations about metaphysics or, say, the internal coherence of certain definitions of power. It simply implies that these conversations are impelled by our interest in the practice of explaining. Nor is it necessarily ‘analytically prior’ to the formation of scientific ontology, if we understand the need to generate and revise conceptual categories as driven by our confrontation with those experiences we seek to explain; we instead end up with a reciprocal relationship between experiencing and organising experience.
Hi Simon, thanks for engaging. Your point is a good one–especially about the relationship between conceptual categories and our experiences. But if we go with it, then IR theorizing expands to include building our scientific ontologies rather than just as PTJ and Dan characterize it, as relating conceptual tools to empirical observations. And once we go there we are on the slippery slope back to defining the practice of IR theorizing in a way that lets philosophical ontology back in. I do want to emphasize here that *I* don’t like the idea of keeping philosophical ontology out; nor do I really think it is possible; just as I don;t think it is possible to keep conceptual theorizing out. My point is that PTJ and Dan’s piece gets bootstrapped into a strikingly narrow vision of IR theory by its own pragmatism.
Hi Janice, I liked your response, and before reading it I made similar points (but not as eloquently) on Dan and Patrick’s post. Glad (but also sad) that your one year of editing confirms our experience. I’d also just like to push home the point, despite the fact it was, I think, Heikki and I that first introduced the PO/SO distinction in IR, that the idea that we can leave behind (or bracket out) PO is mistaken (do I have to keep saying IMO, if it’s not me who else is saying it…:)?) It’ll play a role in how concepts are developed, how we think about the hook up to ’empirical observations’ (sounds like an argument for empiricism to me; I’m sure they didn’t mean it that way), whether we want to go beyond ’empirical observations’ and so on. It’ll also always play a role when we turn to the judgement form of why one typology is better than others etc. Then, just to complicate things further, as I’ve mentioned in Dan/PTJ thread, disciplinary dynamics come into play here, and how our maps of the field function as identities, or identity markers. Breaking out of this way of thinking might be harder than we think (or would wish). Btw, I can’t believe that Dan and PTJ actually think it’s possible to leave PO out of it, so I’m suspecting it’s a call to stop the debate on those issues and move on to linking our conceptual tools to empirical observations. In that respect, they might be closer to David’s (Lake) position and the ‘analytical eclecticism’ idea than I’d thought. Could be wrong about that though.
Thanks, Colin. I did recognise the similarity of our points and appreciated it.
The interesting difference is that we arrive here from different starting points. You arrive here through your monism/dualism debate with PTJ, I think (correct me if I am wrong). I arrive here through the more prosaic process of internal theoretical critique of J&N’s argument. That’s important to me because a) I can’t speak intelligently on your debate with PTJ! and b) it underscores the limits of a practical orientation as a platform for change. It highlights that although J&N use it to help shepherd change in how our conversations and theorising goes, it ends up logically disciplining out forces of change.
In other words, like you, I have no doubt that J&N did not mean to imply empiricism or to bracket what I’ve been calling conceptual theorising (perhaps this is a confusing term?… Stefano Guzzini, what should I be calling it???) But it was an effect of their methodology. Perhaps then they end up as a case in point of precisely what you suggest: that in any case we wouldn’t have an easy time of doing what they ask.
Hi Janice, you’re partly right about me, as I’m certainly keen to show how the attempt to ‘bracket out’ PO, is dependent on PO and particular versions of it, but also I agree with your internal critique, of the idea that a practical orientation specified on empirical observation is the way forward. I just think most of the important stuff we study isn’t empirical.
It’s weird though, both Patrick and Dan would probably still hold onto their commitment to relationalism, but in privileging empirical observation they don’t have a particularly strong account of relations, because they don’t think these relations exist, only the empirical observations. And since I think structures are essentially relational, and I think structures of inequality, structures patriarchy and so on, really exist, and only the outcomes of them are empirically observable. Then in order to change those empirical outcomes we are going to have to change the structures; i.e. effect material change on the relations in the world (although I’m using a broad Leninist notion of material there), not just change the way we talk about them. I just can’t see why anyone would try and change structural inequalities unless you think they really exist. Or to put it another way, empirical observations are extremely important in our field, I wouldn’t deny that for a second, but in many of the most important instances, they still require an explanation; why is this pattern/behaviour/process/outcome happening? And that’s not to suggest that ‘explanation’ (and explanation can’t get off the ground without conceptual analysis) is the only thing we should be doing, but I wouldn’t want to see it neglected either.
Mind you, I’m drawing a very broad brush across both Dan and PTJ here and maybe there are differences between them on these issues. Certainly Tilly considers himself to be a relational realist.
Agreed that most important stuff we study isn’t empirical. Or more exactly, I think that the empirical stuff we study is important because of the frames through which we give it meaning–philosophical, conceptual, theoretical, sensual/experiential, etc. But that is about as far as I can go for now on this line of thinking. On the rest, I shall pass the baton back to Dan and PTJ, whom have thought far more deeply about this than I. Nice conversation. Thanks.