We discussed two questions:
Why do we care if there are alternatives? Remember what Singer is trying to do. He’s trying to show that our reactions to the drowning child case and famine are inconsistent. His principle explains why this is so: we’re required to save the drowning child because not doing so would be bad and, for the same reason, we’re required to give aid to famine victims.
If there is an alternative principle that gets the right answer in the drowning child case but does not show that we are required to give famine aid, that would offer us a defense. We could say that our reactions to the cases are consistent because they conform to this alternative principle.
More to the point, Singer’s drowning child argument would not prove that we are inconsistent in treating the two cases differently.
If you look carefully at pp. 231 and 241, you’ll see that Singer has different versions of his principle. The difference concerns when our obligations to prevent bad things from happening are cancelled. According to the strong version, there is no obligation to prevent bad things from happening only when doing so would mean making a sacrifice of comparable moral importance to the bad thing. According to the moderate version, there is no obligation whenever acting would mean sacrificing something of moral significance.
Why bother with two versions? Well, he believes the strong one is true. But he wants his argument to persuade. And it is much easier to defend the more moderate one. So he proposed both, claiming that either one would lead to his conclusion that famine aid is morally required.
If we came to believe the strong version, I think he is right to say that we would have to make radical changes in our behavior. Perhaps we should. But I don’t think the drowning child example shows that we are committed to believing the strong version. The drowning child example involved only trivial sacrifices on the part of the rescuer but the strong version of the principle demands much more than that. To show that we’re committed to the strong version, the drowning child example would have to involve a very risky rescue, such that the rescuer would suffer serious injury or a high risk of death. Do we think it’s obvious that a rescuer would be required to suffer a serious injury or run a high risk of death to save a drowning child? Again, maybe we should. But it isn’t obvious and most of us think we are not required to do this sort of thing, though, of course, there is no prohibition on risky rescues. If so, the drowning child example doesn’t prove that we are committed to the strong version.
The more moderate version of the principle would require us to prevent bad things from happening unless doing so would involve any morally significant sacrifice. This formulation of the principle, however, is too weak. If we took it seriously, we would think that we aren’t required to save the drowning child if doing so would mean missing an appointment or if we had to take someone’s boat without permission in order to save the child. Speaking for myself, I think that’s ridiculous: of course you should break the lunch date or take the boat. The kid is going to drown if you don’t!
In my opinion, the correct principle lies somewhere between the strong and the more moderate versions. We are required to prevent suffering and death whenever we could do so without making a sacrifice that is close to the significance of suffering and death. We didn’t do the work to make a precise formulation. But I’m sure there is one.** Note that I took out “something bad” because I think it’s too broad. I’m not required to prevent your girlfriend from dumping you in a mean way, for instance. That’s her responsibility.
All I meant to say is that neither of Singer’s formulations struck me as satisfactory. The strong version isn’t proven by the drowning child case and the more moderate version is refuted by the drowning child case.