Wednesday, April 09, 2008

Torture and Walzer's Sliding Scale

So here's a thought that occurred to me recently. Chapter sixteen of Michael Walzer's excellent Just and Unjust Wars offers the argument that violations of just conduct in war (jus in bello) have been justified in some historical circumstances where the moral emergency of losing was sufficiently great.

His first example of this is the bombing of Dresden. In this case, because the threat of Nazism's victory constituted a supreme moral catastrophe, and because the British seemed genuinely in danger of losing the war at that moment, the British bomber command was justified in destroying Dresden in spite of the fact that the attack's intent was to terrorize the civilian population and restore British war morale. Walzer suggests that this justification is not absolute, but that there is a "sliding scale" that can justify normally prohibited actions (bombing civilians) when the moral emergency represented by losing the war is sufficiently high.

Walzer's contrast case is Hiroshima. At that point in the war, Japan's navy was essentially non-existent and her ground forces had been beaten back to their home islands. Walzer argues that dropping an atomic bomb on Hiroshima was not justified in the way that bombing Dresden was because America was in no danger of losing the war at that moment. (I realize that there are lots of reasons why we dropped atomic weapons on Hiroshima, and I think the most compelling ones are utilitarian arguments based on casualties to the Japanese had we invaded their mainland. However, I wish to bracket these other reasons, and for the purposes of this essay assume that civilian non-combatant immunity is sufficient to overrule them.) Walzer claims that we owed the Japanese an experiment in diplomacy given our position in the war in 1945, and that dropping the bomb was not the right thing to do.

Now we come to an interesting question. Suppose that Walzer's right. If the test case is not bombing civilians in wartime, but torturing terrorists who know the location of a weapon of mass destruction, can Walzer's sliding scale be used to justify such actions? Both are thought to be violations of jus in bello, and I can imagine cases in which there might be imminent threat and supreme moral emergency. Suppose that we knew there was a nuclear device disguised as a vending machine somewhere in Washington D.C., and that we had a man in custody who knew its location. If the weapon goes off, then all at once we lose the White House, the Congress, the Supreme Court, the Library of Congress, and the Smithsonian Institute, effectively decapitating our country. (This is not an exhaustive list of what would be lost, and we might have the upside of losing a few special interest lobbies, but let this case stand as a sufficient example of moral emergency.) In this case, can we make the claim from this Walzerian argument that torturing the prisoner is morally justified?

I can think of at least two objections. The first is that this is a veiled utilitarian argument, and so just losing the lives of a few people would be enough for Walzer's argument, but not enough for our intuitions about torture. Against this charge, I suggest that the aggregate value of life is problematic to policy-making mainly because to the extent that it's an absolute value, it can stall any policy making that has that value on each side of the equation, and simply aggregating that value can justify policies of majority tyranny that we historically reject. In other words, one need not be utilitarian to take the view that it may be the right thing to do to perform a little evil to prevent a lot of evil.

The second objection is that we might reject the reasoning in the last objection and take a strong Kantian line: we say that the value of life is absolute, and that moral responsibility for the chaos that ensues following the destruction of our government in the above example attaches exclusively to those who perpetrated the act. We should not engage in further wrongdoing. Against this objection I suggest that the objection seems less intuitively plausible the greater the degree of moral emergency one stacks on the other side. One's imagination can run wild here, but I won't turn this post into a Stephen King story; I simply reject the intuition as an intuition when the moral stakes are high enough.

So what say you all? Shall we torture to prevent moral catastrophe?


  1. I agree that Walzer's point seems mostly utilitarian. And of course I'm comfortable torturing under certain conditions for these reasons. See my post:

    One thing that needs to be clear, and I'm sure you and Walzer realize this, is that we shouldn't adopt a principle that says winning is always a justified goal. This principle would allow any old evil regime to win at all costs. The second thing that I'd add is that even non-evil regimes might sometimes be obligated to lose a war rather than engage in civilian slaughter. This kind of situation would arise in cases where the costs of losing are not atrocious.

    So right away I'll take a stand and say that I can imagine circustances in which a country should lose a war rather than commit some horrible act of mass murder. This comes from a utilitarian perspective, but many ethical views would likely agree here. The more deontological you are the more you'll be inclined to agree with this, I suppose.

    What needs to be added to the account you describe is something that distinguishes it from a basic "the ends justify the means" doctrine. Not even utilitarians believe that, for the ends only justify the means NECESSARY to achieving those ends. If you know the end is justified, you are only permitted to do the minimal amount of harm required to reach that end.

    This is where I often see breaches of just war theory. There is sometimes not enough attention paid to keeping the damage as minimal as possible. Certainly in WWII this got out of control (see the allied bombing of Rohan, France as one of countless examples).

    One peculiar thing, while I'm rambling, is how one views the civilians of either side of a war. It would seem to me that non-combattants on either side of a war are equally innocent and of equal moral standing. I doubt they are treated so by warring armies, however. Even with their belief in double-effect (comes in handy in wartime) I doubt the US army would bomb a building to kill 5 terrorists if they knew US civilians were also inside. Yet they would certianly do so if they were Iraqi civilians, right? That could be cynical speculation on my part, but consider this unlikely scenario: Suppose in WWII we could have ensured an and to our war with Japan if we killed 200,000 American Civilians. It would be unthinkable, right?

    Basically, despite the bad rap that utilitarianism gets regarding the death of innocents, I think that principled utilitarian thinking (instead of a vague "ends justifies the means" nationalism) leads to war strategies that result in fewer civilian casualties than what the US has employed in the last century.

  2. I agree with Arthur on this.

    I am a little less cyincal regarding the bombing of buildings containing both terrorists and civilians, but perhaps I'm not as likely to characterize "enemy non-combatants" as non-combatants. These seem much more difficult to parse out when the "enemy" is a guerrilla force or if the State coerces folk into the military.

    Probably the decision to bomb-or-not-bomb would be based on the importance of getting rid of those particular terrorists versus the likelihood and number of civilians in the building.

  3. Arthur, I think you're right that Walzer and I do not think "moral emergency" just means "whatever happens if we lose." I also think that much depends on how we limit the conception of moral emergency, but I think Walzer's sensitivity to this is present in the sliding scale analysis. Presumably the sliding scale would limit violations according to the degree of moral emergency that follows the loss of a war. I suppose one kind of disanalogy that I'm interested in thinking about is whether we think this analysis can be properly used in discussions of terrorism. Some see terrorism as a kind of global crime, endemic to modern global society and without the fixed aims of war. Others see "the war on terror" as a genuine war against radical Islam, even if it isn't two standing armies slugging it out on the field. Walzer's analysis seems to make sense on the latter view, but can we think that moral emergencies are present in the former kinds of cases? Do police have the same kind of moral analysis to make when confronted with moral catastrophe? To generate "catastrophe" that means more than merely the loss of life, I've suggested the thought experiment of a nuclear device in Washington D.C. because the loss entailed by its going off would include much of our national identity (Library of Congress, Smithsonian, etc.) in addition to our national leadership and military institutions. If that were the case, I think Walzer's analysis might cover it.

  4. I'm not sure I fully understand your last point. Are you wondering if the possible damage of a terrorist attack is sufficiently "an emergency" that traditionally prohibited war tactics are legitimized? I would think the answer to that is "yes." If I understand your second point correctly, you're wondering if police officers should be authorized to use these tactics to defend against a terrorist attack since they are sort of "the soldiers" we've got to defend us against terrorists in a major US city. I think the answer to this second question would be "no."

    This is because authorizing police to use such tactics would introduce too much confusion into their role. Let's see if I can make that more clear. It's hard enough to train a soldier to kill the enemy (and kill civilians when the mission calls for it) and then expect him to be able to turn off that "killer instinct." Even Plato was on to this in the Republic when he speaks of soldiers as analogous wild dogs who cannot always control their passions. This very issue has been part of our difficulty in our occupation of Iraq. People trained as killers make lousy neighborhood liaisons.

    Police officers are barely holding onto their roles as neighborhood liaisons as it is. Most cops have not fired their weapon at a person, and would have great difficulty doing so. If we gave cops increased training in combat you can bet there would be a lot more violent "incidents" because some training is hard to turn off. For this reason, I think it would be a mistake to train police to be able to do something like torture a man to get information, or blow up two civilians to get the terrorist. This would screw with their psychology too much. They would lose the ability to do the rest of their job. These are just claims I'm making, but I'd be prepared to defend this view of human psychology if it came to that.

    With the current risk of an immanent domestic terrorist attack pretty low (we assume) I think the costs of authorizing police to use traditionally "unjust" war tactics has too great a cost. They'd lose the ability to do the more delicate police work that is currently an important part of their job.

    If things changed, however, and we were getting bombed fairly regularly, obviously the cost/benefit analysis would shift. Then I might agree that we should allow the police to jettison their more delicate sensibilities and train them to do horrible things in the name of avoiding a nuclear explosion, etc.

    Was that at all related to what you're getting at?

  5. I'm not sure what all the fuss is about, given the example you raised, Ben. You said we know the guy who knows where the vending machine is is a terrorist. If we know this, then what's the problem with torture?

    It might be wrong if the terrorist is not personally involved in this particular act of terrorism. But if he is, then torture away. If he isn't, then let me think about it some more.

    Opposing torture (and the death penalty) is more a matter of practical obstacles rather than in-principle objections. We don't know that Jones is guilty, we only think he is with some pretty high probability.

    And we don't know that we can trust the government to only use the death penalty or torture on all-and-only guilty people (rather than political opponents or whatever), or all-and-only terrorists (rather than, same thing as above).

    And we also don't know what will happen to the legitimacy and stability of the institution of punishment, or to the international perception of the U.S. government should, in the former case, the U.S. use the death penalty or, in the latter case, should the U.S. use torture. It matters a great deal in the latter case, since we might see more torture in general, given that the U.S. has set a precedent.

    At any rate, these are all practical problems of a lack of knowledge. If we assume these practicalities away, what's so difficult about agreeing to torture terrorists who are participants in a plot that will kill even one innocent? Some Kantian objection? What is this, the 19th century? Commit the Kant to the flames already, good people.

    Hugs for everybody!

  6. I'm afraid I wasn't as clear as I ought to have been. I think you're right about the distinction between the police and the military, but the original thought was in regards to the analysis of just conduct in war and how portable it was. My initial question was whether Walzer's sliding scale analysis could be used in cases unrelated to the conduct of soldiers in wartime, and the way I put the thought then was in regards to the particular kind of warfare fought in regards to terrorism. By contrast, police (among whom we include the FBI) seem to have a similar analysis to make, even if it is not in the context of war. So I agree that we don't want police trained as soldiers, and I'm thinking a little about whether they have similar kinds of moral analysis to make in regards to narrow cases of domestic terrorism that meet the "moral catastrophe" requirement.

  7. Interesting and thoughtful discussion. However, one major misrepresentation of Walzer here. He vehemently disagrees with the use of the sliding scale, or more just more rights, doctrine. If you read him more carefully you will find that offers an alternative to this. A kind of tipping point rather than a slider. Walzer believes we must obey the war convention and refrain from stepping on inviolable rights as long as we can. Only when we reach the point of "supreme emergency" (imminent threat and tremendously morally unacceptable outcome of losing) can we then knowingly override the rights of innocents or violate the war convention. Even when doing this the "human" rights of the victims are not diminished or somehow move from one side to the other as in a sliding scale. We violate them, accept that we are guilty of such, and can only live with the guilt because we "had" (really really really had) to do it.

    Walzer would not agree with the torture of a surrendered soldier or civilian in your scenario. Even the horrible outcome you propose would not endanger the existence of the US or its political sovereignty, for Walzer, requirements of a true "supreme emergency".

    The only leeway one can find in Walzer would possibly be if the prisoner with the intel was not afforded war rights or innocent civilian status, more that of unrecognized non-combatant criminal, like his examples of assassins. Even then I doubt Walzer would encourage torture.