"Torture Doesn't Work"

The following post continues the conversation begun here, and carried on here and here, between myself and an anonymous commenter whom for convenience I address as "Arnie".

I think, before tackling the complex question of implementation of the Golden Rule, I’d like to put to bed a very common argument: “torture doesn’t work anyway.”

Now, just on an a priori ground, this doesn’t make much sense. We’re talking interrogations in which tens of thousands of people may be potentially at risk if the interrogation fails. If an interrogation tactic has a one-in-a-hundred chance of extracting genuinely useful information, that chance – given the stakes – would seem worth taking a shot at. Besides, if it were really useless, and if its uselessness is the established fact that Arnie and others seem convinced that it is, then you would think the interrogation experts would know it. I mean, it’s not like Arnie’s going to be up-to-date on the latest information about the various methods’ efficiency ratings while the guys actually in charge of anti-terrorist intelligence have somehow had that memo slip past them unnoticed.

I’m not sure Arnie realizes this, but his position appears implicitly to be that the American interrogators who wish to use these tactics, are sadists who are just looking for an excuse to inflict pain. Think about it: Arnie wants us to believe that the uselessness of interrogation by “torture” is a fact established pretty much beyond reasonable doubt. Yet American interrogators still want to use it, or else nobody would be arguing about doing it in the first place. Now, I see only three explanations for why these interrogators would be pushing the nation into such an ethical dilemma. One would be because they reasonably believe that such tactics have a reasonable chance of extracting information that turns out to be life-savingly, and potentially even city-savingly, useful – but if I understand Arnie properly, he would have us accept that it is not possible that such a view could reasonably be held by informed persons. A second explanation would be that Arnie believes that he knows more about professional interrogators' jobs than they do; but I think Arnie is humble enough to admit that anything he knows about the effectiveness of various interrogation tactics, our professionals know as well. But with the first and second possibilities ruled out by Arnie’s position, we are left only with the third, which is that our guys know perfectly well that torture is useless – but they are lying and pretending to think it’s useful, because they get a sick kick out of it somehow. Arnie’s position on the uselessness of torture, in fact, entails the corollary that Americans who “torture” terrorists under the pretense of getting information from them, are doing exactly the same thing that the jerks at Abu Ghraib were doing – they are giving themselves some sort of pleasure at their prisoners’ expense.

Being rather unready to condemn our interrogators so sweepingly, and rather more skeptical of conventional wisdom than Arnie seems to be, I instead think it reasonable to say that if our interrogators want to use the tactics, it’s because they think those tactics have a decent shot at giving us potentially critical information – and they would know better than anybody else. With all due respect to Arnie, they would certainly know better than he does, and better than any Congressman does, and far better than any Democratic Party talking head does. In particular, they would have a much better idea of what sort of modern techniques are effective in the hands of Americans dealing with Islamofascist terrorists, than would John McCain – whose expertise comes from being an American on the receiving end of thirty- or forty-year-old tactics wielded by the Vietnamese.

Furthermore, consider that our interrogators know perfectly well that people under torture will sometimes make stuff up in order to make the torture stop, and are perfectly capable of taking the information and cross-checking it with other information. It seems obvious to me that if somebody gives you bogus information, then you simply go back and say mournfully, “Ah, Ahmed, you lied to me. How unfortunate for you. It appears we will have to return to our little chats.” If, as Arnie assures us, a person under torture will say anything to make the torture stop, then I need merely make sure my subjects realize that in the end, only the real truth will make the torture stop for good – and wouldn’t at least a certain percentage of subjects, upon the dawning of that comprehension, give you the real truth? Remember, with the stakes that we’re talking about (e.g. a dirty bomb in New York Harbor), we really just need a decent chance that some of the subjects will break and give us the information we need. The “it doesn’t work” objection, it seems to me, is an objection that can have no compelling force unless you can be confident enough to say, “The tactics under question are guaranteed never to give us useful information that we couldn’t get just as quickly any other way.” And it seems to me to be a priori absurd to think a torture-skeptic could possibly have any rational justification for being that certain of such methods’ unvarying failure.

Now, if you wanted to argue that such methods should be methods of last resort, then you might have a good argument on your hand. You could say that more reliable but slower methods should be in general preferred; you could say that quality interrogations would tend to avoid torture in favor of methods that gave better results; and you could argue that therefore torture should be confined to situations in which there’s special reason to think that nothing else will work or else that speed is of the essence. But you know what? Everybody already agrees with that. None of the “pro-torture” people are running around saying that anytime we can lay our hands on somebody with an Arab name and a funny accent, we should start right in on ripping his fingernails out and shoving cattle prods up his butt. None of us like the idea of “torture,” and none of us want to use it unless there’s a good chance it will prove useful, and none of us want to use it when it isn’t necessary. The issue is precisely: what about when there’s good reason to think nothing else will work and that innocent lives are at stake? That torture is rarely necessary, and that it shouldn’t be used unnecessarily, we all agree on already. So unless Arnie can show that “torture” never works, he’s not going to get far with the “it doesn’t work” argument.

All of that seems to me to be what a reasonable person would think if he didn’t know all that much about the subject. We can actually do even better with Arnie’s argument, though, if we start looking into the evidence more closely.

Having reached this point in my musings, I decided I might as well do some research about the efficacy of “torture” tactics. And then I stumbled across a reference to this interview on Bill O’Reilly – which, being somebody who never watches Bill O’Reilly if I can help it, I hadn’t seen. O’Reilly was interviewing Brian Ross, who I think is the ABC News guy who broke the news about the secret prisons and is certainly no member of the Bush Propaganda Team. Now, according to Ross, the CIA broke fourteen high-value al Qaeda prisoners using a range of techniques of which the worst was waterboarding – though they didn’t use waterboarding on all of them because it wasn’t necessary. “They start with a slap, and then a slap on the chest, the cold room, uh, sleep deprivation, which seems to be the most effective, but for some, the waterboarding is what it took.” In other words, the most effective method generally speaking wasn’t torture (which I think everybody agrees with as a general statement), but for some people the only thing that worked was waterboarding (which contradicts the Sullivan/McCain/Arnie position that, with exceptions so rare as to be insignificant, either waterboarding doesn’t work at all or else something else will work just as well instead). What struck me were these two points in particular:

1. Some of the information the subjects gave was bogus, which is what McCain and company would have us believe almost always happens. Yet not all of it was; useful information was extracted by waterboarding even though we’re only talking about a sample size of fourteen people. “Some” would appear to be at a minimum two and probably more – which is to say, you’re talking at least 10% of these guys would not respond to anything but waterboarding, making the “it doesn’t work except in exceptions too rare to be worth it” seem pretty thoroughly refuted by actual empirical experience.

2. The information we got from Khalid not only was information that clearly we could not have gotten without waterboarding, but that also was clearly valuable and accurate once he did finally break – and that, according to ABC sources who are hostile to the idea of waterboarding, was instrumental in breaking up more than a dozen plots including the plot to pull a second 9/11-style attempt in L.A.

So can we all agree that the “torture doesn’t work” line has been well and truly and thoroughly discredited by actual events? Arnie has expressed in his comments a preference for actual events that have already taken place rather than hypothetical events that might take place. Okay, no problem. Let’s get rid of the hypothetical, and look at those dozen actual plots, in which I think it is reasonable to assume that at least ten or fifteen entirely innocent Americans would have died. Now, given the choice between breaking Khalid, and saving those lives, by waterboarding, or else protecting our moral purity and allowing the plots to proceed as planned, would those of you who oppose "torture" say that you wish we hadn’t waterboarded Khalid?

That’s the most basic question. A "yes" answer places you in the realm of moral absolutism in which no ends can justify this particular means; a "no" answer means all our disagreements are pragmatic questions about where the line is prudently drawn. But the complaint that "torture doesn't work" is really just not one I'm likely to have any further patience with. Because, clearly, sometimes what you guys call “torture” does work, and it can save innocent lives – can, does and has.