Counterfactuals and uptight social scientists

A few months ago I was intrigued by Michael Shanks's thoughts on counterfactuals (maybe it was his remark that 'history is all about things that didn’t happen' that drew some anomalous readings on our truth-o-meter). I was going to make some remarks, but time passed and more interesting things came along. Like enchanted islands full of hobbits and giant lizards and midget elephants. (Please, Santa, bring us a midget elephant for Christmas, please!) But then a few days ago some comments showed up and breathed a little new wind into the old post.

Dr Shanks thinks social science historians who are always nattering about counterfactuals like Walter Scheidel and Ian Morris need to loosen up: they're boring 'positivists' who use counterfactuals to subject historical interpretations to truth-tests because they're freaked out about 'fake' pasts that they want to keep from seeing the light of day. Dr Shanks suggests there's an entirely different use for counterfactuals; counterfactual history à la Niall Ferguson, he suggests, opens up a creative, enlightening potential inherent in alternate narratives. For Dr Shanks, counterfactuals are important because people are always thinking up what-if scenarios and this affects what they do. And so it changes history. Anyway I think that's what he's saying.

There are a couple of problematic issues in this post which I'll bypass: whether Prof Morris and Dr Scheidel are really as dismissive of culture as he suggests; whether positivism -- which is more an intellectual posture rather than an activity or agenda -- really sits all that well with people strongly committed to the kinds of big metanarratives that Dr Shanks ascribes to Scheidel and Morris; whether Niall Ferguson is really doing something that's different not just in degree but also in kind from what counterfactually obsessed social scientists are doing. My point is a narrow one: that there is nothing grand or mysterious or sinister or repressive about counterfactuals. They're just tools, like a lot of others, and their scope is fairly limited. They're not meant to prove interpretations right. What they do is help narrow the range of interpretive possibilities by enabling us to discard ideas that we know for sure can't be true. In the comments to the post Joe Manning mentioned causation, and that's really at the heart of the issue. Making causal connections is the meat and potatoes of a lot of historical work, cultural and social history included. If you say that something A happened that caused something else B but then have every reason to think that something B would still have happened even if A hadn't, then you should probably give up the idea that A, in that context, is all that important. Well, no one will hold a gun to your head to make you give it up, but if you keep going off on how important A is for B is then you're going to look pretty dumb. In the opposite scenario, where B is unlikely without A, it still doesn't 'prove' you're right. It just means that, at this point, there's no immediate reason to think you're wrong. In the absence of any statement or implication of causation, counterfactuals don't really do much at all.

Dr Shanks seems to think the social scientists invest counterfactuals with some extraordinary power. They don't. But I guess you could say that they're worried about 'fakes'. It comes down to stuffy old logic and its uncanny ability to get rid of nonsense. People can sit around cooking up millions of explanations for millions of different things. If some of them don't pass a counterfactual 'test', get rid of 'em. There will be plenty of interesting ones left standing.