Apropos Matt’s recent post: Peter Singer argues we have stringent duties to give to charity. He relies upon a thought experiment:

One Drowning Child
You come across a child drowning in a pool. You can save the child at some personal expense. Imagine you’ll have to jump in right away and thus ruin your iPhone.

Most people judge that they *must* save the child, even though this costs them, say, $500. If so, Singer asks, why not conclude that rather than buying the iPhone for $500 in the first place, you should just send $500 to save a child’s life?

There are some practical objections to this: Perhaps it’s not actually possible to save a child’s life for $500, or there are various problems and moral hazards created by charity. But let’s put these aside.

The central problem with Singer’s argument is that he thinks that once you’re committed to saving one child’s life, you’re committed to the following principle:

If you can save a life without sacrificing anything of moral significance, you ought to do so.

Singer claims this is a demanding moral principle. It would forbid you from buying most luxury goods, extra clothes, video games, tickets to opera, etc., and instead require you to donate most of your money to saving lives.

But the central problem with Singer’s thought experiment is that it is *not* analogous to the situation we find ourselves in. In Singer’s drowning child thought experiment, I save one life at some personal expense, and then move on with my life. I don’t remain in perpetual service to others.

What Singer needs, for his thought experiment to be an actual analog of our current situation, is something like this:

Many Drowning Children

You’re walking alone one day, when you come across millions of drowning children. The children you save will for the most part remain saved, though some might fall back in. However, no matter how many you save, there will always be more about to drown. You can spend your entire waking life pulling children out of pools.

Singer’s entire argument rests upon people’s moral intuitions in the One Drowning Child. But One Drowning Child doesn’t do the work he needs it to do, because One Drowning Child isn’t analogous to the situation Singer thinks we actually find ourselves in. Instead, what Singer needs to do is determine what people’s moral intuitions are in Many Drowning Children. Even if you judge you must save the one child in One Drowning Child, you might not judge that you must dedicate your life to, or even spend a huge amount of time on, saving children in Many Drowning Children.

Note that I am not claiming that Singer’s conclusions are wrong, just that his argument for those conclusions doesn’t succeed.

As an aside, it’s interesting that Singer spends much of his time trying to debunk the use of intuitions as data in moral philosophy, when one of his two most famous and important arguments relies heavily upon intuitions and casuistry.