by David A. Bell
Eric Rauchway, in his response to my post, makes some excellent points. I agree with a good deal of what he says. But I do think we are talking about two different sorts of counterfactuals. Eric is of course right to say that counterfactuals belong to the social sciences. But they are counterfactuals of a special, well-controlled sort, where, ideally, you imagine changes in a small number of variables, and trace the effect on the larger system, in a manner akin to a scientific experiment. As Eric algebraically puts it, "without x, no y." But I'm not sure we can identify this sort of counterfactual speculation with the sort I was talking about, as exemplified by the scenario of the Confederacy winning the Civil War as the result of a set of orders not falling into Union hands before Antietam. In this case, unlike the ones Eric is talking about, the initial variable itself--the wayward orders--is essentially insignificant, and random. But as the consequences of changing it pile up, the number of subsequent variables that are altered quickly becomes unmanageably enormous--modeling the weather would be simple by comparison. The same is true for the example Eric himself gives: Niall Ferguson's speculation in The Pity of War about what might have happened had Great Britain not gone to war in 1914. In reality, this is not a question of a single variable, but of thousands: the loss of British manpower on the Western front; the effects on German strategy; the impact on British politics of "abandoning the Belgians"; the creation of tensions between Britain and France; the effects on British power of not being bled white by the war; etc. etc. etc. etc. etc. Trying to keep track of all these and imagining how they could all have played out together is not, in the end, a useful social scientific exercise--which is why Ferguson's book was in fact much better on what World War I actually did cost Britain, than on what might have been saved had the country acted differently. Ultimately, as I suggested, this sort of counterfactual is more properly an exercise for the imagination, which is why fictional treatments are often the most illuminating.
Incidentally, in my discussion of counterfactual fictions I neglected to mention two particularly prominent recent ones: Philip Roth's The Plot Against America, and Michael Chabon's fascinating new The Yiddish Policeman's Union, which imagines a Jewish national refuge in the Alaskan panhandle.