the causal impotence objection (2)

May 27, 2008 in Ethics, Philosophy

Alastair Norcross has a cool (but ultimately, I think, wrongheaded) response to the causal impotence objection (see here) in a 2004 Philosophical Perspectives paper (pdf here). (Edit — Mike Almeida writes to tell me that he and Mark Bernstein published a version of this argument in 2000 in the Journal of Applied Philosophy. See here.) The presentation here is slightly more general than Norcross’s, in that avoids some inessential empirical assumptions.

Let n = the number of individuals such that omni > veggie, where omni = the number of animals killed if n individuals do not go vegetarian after t, veggie = the number of animals killed if n individuals do go vegetarian after t. Let saved = omni – veggie. Your subjective probability for the truth of “if I go vegetarian, I will be the n-th person to do so after t” should plausibly be set at 1/n. As long as saved > n (as it surely is) the expected value of you going vegetarian (in terms of number of animal lives saved) is greater than 1.

It’s important to note that this is not an argument that your going vegetarian is causally efficacious: n-1 times out of n, it isn’t. Norcross’s idea is that it is wrong to ignore the chance that your going vegetarian will be causally efficacious, as “We commonly accept that even small risks of great harms are unacceptable.”

But something’s amiss. Let E be the event such that n individuals go vegetarian after t in E, and suppose that your going vegetarian will bring E about. Even if you don’t go vegetarian, E will still occur, and really in very little time at all — all it takes is one more person (call this person “Steve”) to go vegetarian. So it’s false that the expected value of you going vegetarian (in terms of number of animal lives saved that otherwise would not be) is greater than 1. (It is true that the expected value of you going vegetarian — in terms of number of animal lives saved that otherwise would not be saved by you, but rather by someone else — is greater than 1, but, morally, that’s neither here nor there.) Continuing to suppose that if you go vegetarian, you will bring about E, the relevant issue here is how many surplus deaths there will be if you don’t go vegetarian, and it’s Steve that brings about E. It’s hard to say with any confidence, but this number is probably comparatively quite small — surely several orders of magnitude smaller than saved. And this affects the expected value (disvalue) of you going vegetarian (staying omnivore) proportionally.

You can (and Norcross, in a way, does) try arguing, whether or not your going vegetarian brings E about, your not going vegetarian is practically certain to delay E (and therefore practically certain to result in excess animal suffering). This is a strong argument, but it’s distinct from the one we’ve been considering here.