The Epistemic Problem

I just read a good paper by Tyler Cowen on "the epistemic problem" for consequentialism. The problem is that it's hard to predict the consequences of actions, esp. over a long time horizon, and we are meant to infer that this implies consequences can't be a good basis for deciding how to act.

If that sounds like a stupid argument, it is.

Unfortunately, philosophers have constructed sophisticated examples to help pull the wool over their own eyes. Take this example, discussed in Cowen's paper. You are a general deciding where to invade France to ensure victory over the Nazi's. You can land at the Pas-de-Calais or in Normandy and have no good reason to prefer one over the other. You do know, however, that if you land in Normandy you will break a dog's leg during the landing. (Ignore how implausible it is that you could know something about the dog but nothing about the beaches.)

A utilitarian would land in the Pas-de-Calais so as not to harm the dog. But that is stupid, we're supposed to say. It's patently so stupid the utilitarian must be bonkers and we shouldn't be utilitarian's or we risk being thrown in Arkham. And you don't want to be thrown in the asylum, do you? I didn't think so.

But here's the catch that got swept under the rug. If we didn't know about the dog we wouldn't have a way for deciding on Calais or Normandy. We'd just pick one of the beaches randomly, probably by flipping a coin. Is getting in touch with our inner Two-Face a better way to decide? I guess some Very Serious Philosophers think so.