Trying to Reason about torture is dangerous as a ticking bomb
Matthew Yglesias is reasonable and brilliant as always. He concludes "Even without resorting to any implausibly absolutist arguments, then I don't think we're likely to find any situations where tortuting people is a good idea." Taht is, while it is possible to imagine a hypothetical in which torture is justified it is unlikely that such a situation will ever occur.
I think that for pure practical purposes, there is little difference between "will probably never happen before the humankind dies off" and "happens very rarely". Let's say that the cost of an absolute taboo against torture is that people are sometimes but very rarely killed by a time bomb which could be diffused. That is a small cost compared to the cost of the increased torture which occurs if people with a darker side (that is people) are liberated from the taboo.
If we are deciding whether we should do what we can to support the taboo on torture, very rare cases in which torture is good doen't tip the practical moral balance any more than a purely hopothetical case. I also think that, if we are trying to support the taboo, conceding that torture would be moral in a purely hypothetical case does about as much damage as conceding that it is moral in very rare cases. I'd say the best thing to do if someone tries to defend torture is to shout that it is always morally unacceptable. I admit, here, that I don't really believe that, because I assume that none of the few readers of this blog will ever be seriously tempted to torture anyone.
I have some other quibbles with Yglesias' argument.
Yglesias' argues that with torture "You can get information (as seen in, say, The Battle of Algiers) like what the name of the guy you reported to was, but not stuff that's incredibly time-sensitive ("where's the bomb!")." Who said that all terrorists have short fuses ? I mean couldn't the bomb be set to explode in a year or two ?
I think Yglesias uses a false dichotomy to argue that "Thus, the goal of torture is not to shift the incentives facing the victim (cooperate and we won't hurt you) but rather to destroy the personality of the victim, thus creating a shattered wreck of a human being who, unlike his predecessor, lacks strong ideological commitment to the cause." This may be the way torture generally achieves its goals (aside from the goal of gratifying sadists).
"If your proposed victim is disposed to respond self-interestedly to inventives, it's easy enough to get him to cooperate without resorting to extreme measures. As with flipping a perp in a criminal case you offer him a nicer cell, more cigarettes, a shorter sentence, whatever, to be provided if his information pans out. Torture is for precisely those cases where the victim is sufficiently committed to the cause that he won't respond to incentives in an instrumentally rational manner." I think there is an equivocation her. Two different statements are equated -- to "respond self-interestedly to inventives," is not the same as to " respond to incentives in an instrumentally rational manner." Instrumental rationality does not imply complete selfishness. The victim may be partly committed to the cause and thus not does respond self interestedly but not completely and absoltutely committed and thus responds in an instrumentally ratioanal manner. Yglesias' argument relies on asserting that there is nothing between absolute selfishness and absolute commitment, that is, on asserting that a mild incentive has the same effect as a major incentive.
There are many things which I would do to avoid torture which I would not do to get a nicer cell or more cigarettes, even assuming that the mere threat of possible future torture does not reduce me to a "shattered wreck of a human being".