In Not even if they were really bad headaches? Jonathan Ichikawa asks whether there is some number of headaches which if they could be prevented by killing someone would justify that killing. The reasoning (according to a paper by Norcross which he finds persuasive) goes something like: Suppose you had a bad headache and no pain reliever in the house; suppose further there was a one in a million chance that if you drove to the store to buy some, you'd get in a fatal traffic accident on the way; it's not irrational to go to the store for pain reliever instead of suffering; ergo headaches can be traded off against the chance of death, and thus there must be some number of headaches such that they outweigh the certainty of a single death.
Note, however, that this chain of reasoning involves a classic bad move that utilitarians (and most other consequentialists) make all the time: the unjustified substitution of one person for another. I sometimes wonder if they even notice that they're doing this.
While it may be rational for you to risk your own life to some small degree to cure your own headache, that intuition says nothing about the rationality of imposing that risk on others. For instance, would anyone argue that it's intuitively obvious that it was rational to coerce someone to go get you the pain killer so they underwent the risk? The utility is the same. If you think that it's not the same, because being coerced causes distress, suppose instead that you send someone out who is ignorant of the risk. You might further suppose that you were such a thoroughgoing utilitarian that neither the deception nor putting someone at risk causes you any greater distress than contemplating your own risk. Folk morality holds no terrors for you.
You're right that I'm skipping an important step. Yes, I do believe that all lives are equal -- that is, I don't believe in agent-centered restrictions, and I didn't argue for that.
Still, this thought experiment carries some force -- the anti-utilitarian is robbed of the option of claiming that *life* is what makes murder fundamentally different from headaches... it forces him to defend his intuition on difference-of-person grounds, as you do here.
There's more to be said... Norcross has some nice examples about poison gas that you might want to check out.
Posted by: Jonathan at September 30, 2003 09:16 PMI read the Norcross pieces before posting, and they simply don't establish what he wants to establish.
First, "it is not irrational" <> "it is better". He makes no argument that one ought to prefer risking death to having a headache, even though later on in the argument he assumes that the world in which you risked death is simply better than the one in which you endured (you can't get from Headache to Many Headaches without such a move). The world where you risked and won, so are alive without headache, is clearly better than the world where you endured the headache, but the world where you died is clearly worse not just from your own pov, but from the utility maximizing one.
But even if you accept the Many Headaches proposition that it is better for one million people to choose a one-in-a-million risk of death than endure headaches, you can't get to Worse where you inflict a premature death on someone to prevent the headaches without supposing that people are fungible and it makes no difference whether you (and a million other people) accept a disutility yourself(selves) or foist it on someone else. And yet Norcross makes that transition without comment. Rather than "making explicit what most people already believe," I think most people would read Norcross's argument as a reductio of the consequentialist position.