One of the most memorable sets of interactions I had at this year’s ISA conference came directly after the close of our impossibly-well-attended graveyard-shift panel on “The Relational Turn in the Study of World Politics.” (I say “impossibly-well-attended graveyard-shift” because it was in fact scheduled in the very last session of the entire conference, 4:15pm Wednesday afternoon; I and the other participants, including Dan, expected to outnumber the audience-members and were quite prepared to adjourn the panel to the bar. Instead, there were 64 people in the audience.) The panel was a deeply odd occasion for me in a number of ways, especially since I had the unenviable task of channeling our absent and very famous discussant — Alex Wendt — by reading aloud his written comments on the papers and the project. If you’ve never had the experience of reading aloud someone else’s sympathetic-yet-critical comments on a project that you’ve been involved with for years, and reading them aloud before a room full of your peers and colleagues . . . well, let’s just say that it’s tremendously odd to find yourself mouthing someone else’s words and having to forcibly restrain yourself from responding to the very comments “you” are making.

But enough about that for now. What I want to talk about instead is the interesting question asked of us by an Italian graduate student studying at the Sorbonne (if you’re reading this, do get in touch so I can add in your name — we were both out of business cards by that point in the conference so I cannot recall his name). The topic of conversation had by that point shifted to the problem of whether there were causal factors of interest to IR scholars that could not be captured by a focus on networks and processes of social relation/transaction — in other words, whether we needed to make a space for non-social factors in our analyses of the social world. His example was a meteor crashing into the earth, which he posited was clearly a non-relational factor: a meteor is clearly exogenous to all possible social networks, and its impact with the earth cannot be explained in terms of processes of social transaction. I replied that even a meteor’s impact would have to be socially mediated in order for it to assert causal effects on the social world; the physical impact needed to be made meaningful in order for the meteor to have any social impact, and that process of making-meaningful couldn’t be reduced to the physical act of the meteoroid hitting the planet. Hence, relational analysis could deal with even a meteor impact.

The questioner was not fully satisfied, and we started chatting after the panel. He tendered the following claim: the 1908 meteor impact in Siberia caused the next several winters to be harsher; those harsh winters contributed both to the demoralization of the Russian military; and hence the meteor caused a social outcome (the Russian Revolution, and hence the outcome of the First World War) without the mediation of social relations — much like, he claimed, the AIDS virus and global warming, both of which exerted effects whether anyone noticed them or not. If true, this claim would present a bit of a challenge to relational social theory, since it would appear to affirm the importance of essential physical dispositions in the face of social arrangements, and even to discount those social arrangements to a large extent.

Fortunately for relational theorists, I think that the questioner’s challenge is relatively easily met. I offer the following three points — basically the same three points I made in conversation with the questioner — in reply, in the hopes that some of our readers might find them of interest.

1) the purely physical causal claim is rather tenuous to begin with. Scientists have been investigating the 1908 event — usually called the Tunguska event — for almost a century, and the results remain somewhat inconclusive. The basic problem is that there are no pieces of the meteorite that one would normally expect to find at the site of such a collision between an extraterrestrial object and the earth. Nor is there any obvious impact crater to be found at the site of the blast, and debate continues as to whether Lake Cheko could in fact be that crater. So although scientists generally agree that the explosion was caused by some kind of collision with an extraterrestrial object, it remains unknown whether that object was an asteroid, a comet, a small black hole, or whatever (although most contemporary scientists say asteroid or comet; gone are the days when anyone reputably claimed that this event involved an alien spacecraft, except in various fictional works). So we still don’t quite know what the event itself was.

Regardless, scientists are also still debating precisely what the effects of the Tunguska event were. The questioner’s claim about harsh winters, though sounding plausible at first, does not seem to be echoed by any of the published research on the event; there are some claims that global warming might have been kick-started by the Tunguska event, and also some evidence that the event produced some cooling (about 0.3 degrees C) in the Northern Hemisphere relative to the Southern Hemisphere — but “volcanic activity during this period also contributed to the cooling.” So the physical effects of the meteor collision, if meteor it was, remain ambiguous.

2) but grant the meteor’s posited effects for a moment. Even if the meteor impact made things a bit colder, it does not necessarily follow that the meteor impact is somehow responsible for everything that came after that minor decrease in temperature. For one thing, the estimated decrease in temperature is so small — 0.3 degrees C! — that it probably wouldn’t even be noticed in the middle of a Russian winter. So in order to argue that this drop in temperature actually caused anything, one would have to somehow establish that dropping 0.3 degrees C crossed some kind of threshold from the ordinary “really fracking cold” series of Russian winters to an extraordinary “so fracking cold that we feel like having a social revolution” series of Russian winters. For another thing, we have to keep in mind that extremely cold and harsh winters are not exactly unknown in that part of the world; there is, after all, a famous old saying about how Russia is protected by General Winter and General Snow, and there are numerous other anecdotes that reflect that basic fact that it’s always harsh and bitter during the winters in Russia. Nonetheless, people have adapted to living there, which suggests that extreme cold is survivable given the proper preparations — in other words, that it’s impossible to reduce “the Russian winter” to a set of physical constraints without also focusing on the survival strategies that people have adopted in order to get through them.

Indeed, we can push this point a bit further. What makes the Russian winter so devastating for invading armies from the west is, I would posit, the lack of preparation for and familiarity with the local conditions that those armies bring along with them. Similarly, what makes the Russian winter manageable for those who live there are their coping strategies. Put together, these observations strongly suggest the inseparability of “the Russian winter” from a whole panoply of social arrangements and practices surrounding and reacting to a set of environmental conditions. Hence, as social studies of science have been telling us for years, it makes little sense to try to isolate the “social” versus the “non-social” aspects of a phenomenon, since what we have in practice is a complex tangle — Andrew Pickering calls it a “mangle” — of various aspects. Sorting through these aspects is both pointless and practically impossible.

3) but further, grant even that we could determine that the 1908 meteor impact has certain environmental effects that could be traced definitively in purely physical terms. The problem that we run into at this point is that to say that these temperature fluctuations caused the Russian Revolution is to imply that absent these temperature fluctuations the Russian Revolution would not have occurred. Such a counterfactual is, however, rather implausible, since the other things that scholars conventionally think of as having produced the Russian Revolution would operate completely irrespective of whether the temperature was a bit colder. That other combination of factors — virtually all of which are clearly more social than not, such as governmental oppression and economic dislocation — is, as Max Weber might put it, “adequate” to cause the outcome. In fact, Weber’s procedure is instructive here: if we can imagine a scenario in which some causal factor changes but the outcome does not, then that factor probably isn’t all that important to the overall causal account. Conversely, if we can’t imagine things going the way that they did without that factor, then it’s probably important. We can easily imagine the Russian Revolution with a series of colder or warmer winters, so that’s probably not all that important; we can’t imagine the Russian Revolution without, say, the economic problems that led to food shortages, or the leadership of someone like Lenin, so those are probably important.

Note the importance of “imagination” here; Weber is very clear that the only way that we determine whether something mattered is to test the limits of our ability to conceptualize a plausible alternative without it, and that is simultaneously a comment on the event and on our present-day cultural resources that we use to grasp the event. But this is not a problem for Weber, since social science for him is not about grasping how things “really are” (whatever that means) but instead about the analytical ordering of experience — both our experience and the experience of the people that we are studying. Imagining a counterfactual scenario is a good way to do this.

Note also that the argument I’ve sketched here stands up even if the 1908 meteor impact did alter the temperature. Because of the mediation of physical facts by social practices — because of the ways that physical facts have to be made meaningful in order to exert social causation — whether the meteor impact changed the temperature is, strictly speaking, irrelevant. That fluctuation would have to be made meaningful in order for it to matter, and social relations would remain implicated in any such process. And while it’s even possible that the temperature fluctuation did exist and was made socially meaningful (people commented on it, etc.), the fact that we can easily imagine the outcome without it is a good sign that it probably wasn’t all that important.

The only way around this would be to say that all social relations are reducible to non-social factors: that even the making-meaningful of temperature fluctuations has to do with biological facts about our species, and that we’re basically assembled such that we find certain kinds of temperature fluctuations of interest. By the same token, the only way to make those fluctuations that (in such an account) we cannot help but notice cause a change in the way that a society is governed is to link governance practices to non-social factors, such that our noticing the temperature and being physically affected by it would have a necessary spill-over effect into how we think about the government. It’s all or nothing: either social factors (including the making-meaningful of natural events) cause social outcomes, or the social is reducible to the natural. And you can’t have it both ways. On those grounds, even if a meteor does crash into the earth and alter the climate, I will continue to insist on the analytical priority of social relations, as I did on the panel.