Philosophy Discussion

Torture — The Case for Dirty Harry and against Alan Dershowitz Author(s): UWE STEINHOFF Source: Journal of Applied Philosophy, Vol. 23, No. 3, Themed Issue: The Ethics of War (2006), pp. 337-353 Published by: Wiley Stable URL: https://www.jstor.org/stable/24355181 Accessed: 10-10-2018 18:29 UTC

REFERENCES Linked references are available on JSTOR for this article: https://www.jstor.org/stable/24355181?seq=1&cid=pdf-reference#references_tab_contents You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide

range of content in a trusted digital archive. We use information technology and tools to increase productivity and

facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at

https://about.jstor.org/terms

Wiley is collaborating with JSTOR to digitize, preserve and extend access to Journal of Applied Philosophy

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

Journal of Applied Philosophy, Vol. 23, No. 3, 2006

Torture — The Case for Dirty Harry and against Alan Dershowitz

UWE STEINHOFF

abstract Can torture be morally justified? I shall criticise arguments that have been ad duced against torture and demonstrate that torture can be justified more easily than most philosophers dealing with the question are prepared to admit. It can be justified not only in ticking nuclear bomb cases but also in less spectacular ticking bomb cases and even in the so called Dirty Harry cases. There is no morally relevant difference between self-defensive killing

of a culpable aggressor and torturing someone who is culpable of a deadly threat that can be averted only by torturing him. Nevertheless, I shall argue that torture should not be institu tionalised, for example by torture warrants.

What is so bad about torturing people, anyway? People also kill people. Soldiers kill people, policemen kill people, doctors kill people, executioners kill people and ordinary people kill people. Some of these killings are justified. So why shouldn't it be justified in some cases to torture people? After all, being killed seems to be worse than being tortured. Even most of the tortured people seem to see it this way; otherwise they would kill themselves in order to escape further torture (yes, some do, but they are only few). So, if killing is sometimes justified, torture too must sometimes be justified. Henry Shue thinks that this is a weak argument. Referring in particular to the

legitimate killing of people in war, he states:

Even if one grants that killing someone in combat is doing him or her a greater harm than torturing him or her . . . , it by no means follows that there could not be a justification for the greater harm that was not applicable to the lesser harm. Specifically, it would matter if some killing could satisfy other moral constraints (besides the constraint of minimizing harm) which no tor ture could satisfy.1

The moral constraint he has in mind is the prohibition of assaults upon the defence less. However, it can be doubted whether this is a valid constraint at all. Shue's idea

that it is valid is based on the fact that one 'of the most basic principles for the conduct of war (jus in bello) rests on the distinction between combatants and noncombatants and requires that insofar as possible, violence not be directed at noncombatants'.2 One fundamental function of this distinction, according to Shue, is to try to make a terrible combat fair. No doubt, the opportunities may not have been anywhere near equal — it would be impossible to restrict wars to equally matched opponents. But at least none of the parties to the combat were defenceless.3

© Society for Applied Philosophy, 2006, Blackwell Publishing, 9600 Garsington Road, Oxford, OX4 2DQ, UK and 350 Main Street, Maiden, MA 02148, USA.

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

338 Uwe Steinhoff

To be sure, Shue admits that this 'invokes a simplified, if not romanticized, portrait of warfare', but he explains that

. . . the point now is not to attack or defend the efficacy of the principle of warfare that combat is more acceptable morally if restricted to official combat ants, but to notice one of its moral bases, which, I am suggesting, is that it allows for a 'fair fight' by means of protecting the utterly defenceless from assault.4

It is true that the fact that the prohibition of assaults upon the defenceless is not always obeyed in war does not pose any problem for Shue's argument. What does pose a pro blem, however, is that this principle is no jus in hello principle at all, and it does not, pace Shue, underlie the principled distinction between combatants and non-combatants.5 This can be seen quite easily. Suppose an army that possesses superior artillery advances upon the enemy trenches. It has two options. It can shell the enemy trenches from a safe distance with artillery fire. The enemy would be entirely defenceless against such an attack; his weapons do not have the range for fighting back.6 (This was pretty much the situation in the first and the second Iraq wars.) Or it can advance without shelling the enemy first. Here the enemy would be able to fire upon the advancing troops and to inflict heavy casualties. The point now is that there is simply no article in the laws of war (and no constraint in just war theory, for that matter) that would rule out the first option as such. Of course, if the objectives of the war could be achieved without shelling the enemy in its death-trap trenches and without inflicting higher casualties on one's own forces, this alternative course of action would have to be adopted. But the reason for this is the principle of proportionality.7 An alleged prin ciple of the immunity of the defenceless plays no role here.

To be sure, I do not deny that there is something fishy about attacking the defence less. What is fishy about it might be captured very well in this passage:

The pilot of a fighter-bomber or the crew of a man-of-war from which the Tomahawk rockets are launched are beyond the reach of the enemy's weap ons. War has lost all features of the classical duel situation here and has

approached, to put it cynically, certain forms of pest control.8

Judged from a traditional warrior's code of honour, a code that emphasises, among other things, courage, there is nothing honourable in killing off defenceless enemies (whether it is therefore already dishonourable is yet another question). But honour and morality are not the same, and honour and the laws of war are not either. In short, the prohibition of assaults upon the defenceless is neither an explicit nor an implicit principle of the laws of war or of just war theory.

Moreover, it is not even clear what should be unfair about killing the defenceless in war in the first place. If the rules of war allowed both sides to kill, for example, defenceless prisoners of war, where is the unfairness? Both sides are in the same situation. And if the laws of war allowed the torture of POWs, the same would be true. Invoking principles of fairness does not deliver an argument against either of the two practices.

David Sussman has recently attempted to give another explanation of why torture, allegedly, 'bears an especially high burden of justification, greater in degree and differ ent in kind from even that of killing'.9 However, he also cautions that this 'is not to say that when torture is wrong, it is worse that [sic] wrongful acts of maiming and killing'

Society for Applied Philosophy, 2006

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

Torture — The Case for Dirty Harry and against Alan Dershozvitz 339

or that 'it is more important for us to prevent torture than it is to prevent murder, or that torturers should be condemned or punished more severely than murderers. The moral differences between torture, killing, and maiming may not always make a difference in such third-personal contexts'.10 But why, then, should torture bear an 'especially high burden of justification'? Does it bear such a higher burden only in second-person or first-person contexts? But what is that supposed to mean? That whereas one does not need a better justification for allowing others to torture or for not condemning torturers than for allowing others to kill or for not condemning killers, one does need a better justification for torturing (and perhaps also for risking being tortured) than for killing? Sussman argues that:

. . . torture forces its victim into the position of colluding against himself through

his own affects and emotions, so that he experiences himself as simultaneously powerless and yet actively complicit in his own violation. So construed, tor ture turns out to be not just an extreme form of cruelty, but the pre-eminent instance of a kind of forced self-betrayal, more akin to rape than other kinds of violence characteristic of warfare or police action."

He elaborates:

What the torturer does is to take his victim's pain, and through it his victim's body, and make it begin to express the torturer's will.12 . . . Torture does not merely insult or damage its victim's agency, but rather turns such agency against itself, forcing the victim to experience herself as helpless yet complicit in her own violation. This is not just an assault on or violation of the victim's autonomy but also a perversion of it, a kind of systematic mockery of the basic moral relations that an individual bears both to others and to herself. Perhaps this is why torture seems qualitatively worse than other forms of brutality or cruelty.13

But all this can hardly explain the alleged difference between 'third-personal' and other contexts. In fact, it would seem rather natural to assume that, if torture isn't worse than killing from a third-person perspective, it isn't so from a second- or first-person perspective either. If Sussman sees this differently, the high burden of justification, it appears, would be on him.

Conversely, assume a dictator confronts prisoner A with the following choice — to either kill one of ten prisoners or to torture one of them for two hours (all these prisoners are innocent and have no special relation to the first prisoner).14 If A refuses to choose and to act on his choice, all ten prisoners will be killed. If he decided to choose (and it seems plain to me that he is justified in choosing even though he may not be obliged to do so), should the prisoner facing the choice rather kill one prisoner than torture one of them for two hours? This is to be strongly doubted. On the other hand, there can hardly be any doubt that the ten prisoners would prefer him to opt for torture if he asked them what to do — to refuse to choose, to choose to kill or to choose to torture. He is not permitted to ask them (if he did, all ten prisoners would be killed). However, in light of the fact that it is reasonable to assume that they would prefer two hours of torture to being killed for good, he should honour that preference if he decides to choose. It seems, then, that torture is easier to justify than killing even from a first-person perspective.

© Society for Applied Philosophy, 2006

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

340 Uwe Steinhoff

Moreover, the feature that in Sussman's eyes makes torture so bad is by no means unique to torture. We find it, in fact, even in armed robbery. If a robber points a gun at a victim and threatens to kill him if the victim does not give his money to the robber, the robber, if successful, also turns the victim's agency against the victim himself. He makes the victim's fear express his, the robber's, will; and the victim, in handing over the money for fear of death, is 'complicit' in his own violation.

It is important to note here that the victim will also normally not react to the threat with a utility/risk-calculation, i.e. with rational deliberation. If, especially at night in a dark alley, one has a gun pointed at one's head and hears: 'Give me your f. . . money or I'll f. . . kill you', one will hardly engage in deliberations of this kind: 'The probabil ity that his threat is meant seriously lies around x; my options to defend myself or to escape are this and that, with this and that probability of success; if, on the other hand, I give him the money, this will cost me certainly y money units, which in turn leads with probability z to my not being able to pay in time for this and that thing, which in turn . . . and therefore I will. . .'. Rather, many people threatened with deadly violence will think: 'Oh my God, he wants to shoot me, he wants to shoot me . . . I'll give him the money!': thus, the sudden death threat and the perceived time pressure will in many if not most cases severely disturb the ability to deliberate rationally if they do not completely undermine it. In short, fear may overwhelm one as pain overwhelms the victim of torture. There may be differences of degree but no qualitative ones. Besides, at least some forms of torture will undermine the ability of some torture victims to think rationally less than some forms of death threat would of some threatened people.

Thus, Sussman is not able to show that there is a higher burden of justification on torture than on killing. Conversely, thought experiments like those of the ten prisoners strongly suggest that it is the other way around.

Of course, it is true that emotional reactions to displays of torture, for example in movies, are much stronger than those to displays of killing. And I suspect that much of the conviction that torture is somehow more terrible than death is sustained by these immediate emotional reactions. In fact, however, little of moral relevance follows from them. After all, the emotional reactions to scenes involving large amounts of faeces or cockroaches, for example, are mostly also stronger than to scenes involving killing, at least when it is a quick killing. When the killing is prolonged the reactions are stronger, but they are then prompted by the displayed suffering — sometimes simply in the form of fear — not by the impending death itself. The point is that we simply cannot empathise with the dead. There is nothing we can empathise with. We do not know how it is to be dead, and there is certainly no possibility to feel dead (in any non-metaphorical sense). We do know, however, what it is to feel pain. Therefore, if we see how pain is inflicted on another person, we can feel quite literally corn-passion. There is no such compassion possible with the dead (we can only feel sorry for them, but that is a completely different and far less intense feeling). In other words, our emotional re actions to displays of torture or killing respectively are distorted by certain limits of empathy. What are not distorted are our emotional reactions to the prospect of our own death. Most of us fear it more than torture (that is limited in time). Whether killing is worse than torturing, then, should be measured against the preferences of people facing the alternatives of torture (over a certain limited time) and death — most prefer torture. To measure it against emotional reactions to displays of torture would by way of extension also mean that it is better to shoot someone dead than to push him

Society for Applied Philosophy, 2006

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

Torture — The Case for Dirty Harry and against Alan Dershowitz 341

into a bathtub filled with cockroaches. But such an implication obviously reduces the premise to absurdity.

Thus, it seems the argument in defence of torture given at the beginning is not such a weak one after all. We can now slightly rephrase it in the light of our findings: If it is permissible to kill a defenceless enemy combatant in order to avoid own casualties, why should it not be permissible to torture a defenceless terrorist in order to avoid own casualties?

One might try to answer this by saying that the enemy soldiers in my example may be defenceless in the situation in which they are shelled or attacked with missiles, but that they still pose a threat — they threaten our advancing troops, which is precisely the reason why we take them out by artillery or rockets. The detained terrorist, on the other hand, does not pose a threat any more. For what can he do?

Sanford Levinson states that 'the defender of torture would dispute the premise that the captured prisoner is no longer a threat, at least if the prisoner possesses important information about the future conduct of his fellow terrorists'.15 The case of the de

fender of torture can be made much stronger, however, if we take the infamous ticking bomb case. The terrorist has somewhere hidden a bomb that will, if it goes off, kill a high number of innocents (maybe it is even a nuclear bomb that would destroy one of the ten biggest cities in the world). He does not want to say where it is. Would it not be justified to torture him in order to make him disclose the vital information and find the bomb in time to disable it?

Some might be inclined to argue that withholding vital information (i.e. an omission) couldn't reasonably be described as a threat. Perhaps, perhaps not. However, people culpably posing a threat and aggressors are not the only persons that can be legiti mately killed. For it seems, as Phillip Montague has pointed out, that, in cases in which one cannot save oneself from the negative consequences of another person's acts in any other way than by diverting or inflicting them on this person, one is justified in doing so. Montague states the following conditions for this:

(i) individuals Xj . . . Xn are situated so that harm will unavoidably befall some but not all of them; (ii) that they are so situated is the fault of some but not all members of the group; (iii) the nature of the harm is independent of the individuals who are harmed; (iv) Y, who is not necessarily included in Xj . . . X, is in a position to determine who will be harmed.16

The inflicting of harm (for example in the form of an attack or in the form of torture) on a certain person, thus, is not legitimised here by a present aggression but by the person's culpably causing the threat of harm. This cause can lie in the past. Montague adduces the example of a doctor who wishes for the death of a cardiac patient and therefore swallows the pacemaker the patient needs. The only chance to get the pace maker quickly enough to save the patient's life is to perform a hasty operation on the doctor, which, however, would be fatal to him. Hence, one faces here a situation in which either the innocent patient or the malicious doctor has to die. However, if one of the two has to die anyway and one has to decide who it is going to be, then one clearly should, says Montague, decide against the person who has culpably brought about this situation. That seems to be a pretty reasonable stance, and I agree with him. The application to the case of the terrorist who has hidden a bomb to kill innocents is obvious.

© Society for Applied Philosophy, 2006

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

342 Uwe Steinhoff

It even suffices when the bomb would kill only one innocent. We do not have to assume a ticking nuclear bomb, killing thousands or even millions. Consider the Dirty Harry case. In the Don Siegel movie Dirty Harry someone kidnaps a female child and puts her in a place where she will suffocate if not rescued in time. There is not much time left, according to the very claims of the kidnapper. The police officer Harry (Clint Eastwood) is to deliver the ransom to the kidnapper. When they finally meet at night in a park, the kidnapper knocks Harry down with his gun and tells him that he will let the girl die anyway. He also tells Harry that he wants him to know that before he kills him too. In the moment he is about to shoot Harry, who lies defenceless and badly beaten to his feet, Harry's partner interferes (and is shot). The kidnapper can escape, wounded. Harry pursues him. Finally he corners him, and the kidnapper raises his arms to surrender. Harry shoots him in the leg. The kidnapper is frightened to death, tells Harry not to kill him and that he has rights. Harry asks him where the girl is. The kidnapper talks only about his rights. Harry sees the kidnapper's leg wound and puts his foot on it, torturing the kidnapper. The camera retreats. In the next scene, the girl is saved.

The Dirty Harry case, it seems to me, is a case of morally justified torture. But isn't the kidnapper right? Does not even he have rights? Yes, he has, but in these circum stances he does not have the right not to be tortured. Again, the situation is analogous to self-defence. The aggressor does not lose all of his rights, but his right to life weighs less than the innocent defender's right to life. The aggressor culpably brings about a situation where one of the two — he or the defender — will die. It is only just and fair that the harm that will befall in this situation upon one of the two is diverted to the person who is responsible for the harm — the aggressor. In the Dirty Harry case, the kidnapper brings about a situation where a person is tortured or will continue to be tortured until death (being slowly suffocated in a small hole is torture). It is only just and fair that this harm befalls the person responsible for the situation — the kidnapper. Moreover, the choice is made even easier by the fact that being tortured for a small period of time is better than being tortured until death. Harry made the right decision.

Two replies might be made at this point. The first one — repeated like a litany by certain opponents of torture — is that interrogative torture simply does not work. That, however, is simply wrong. Sometimes interrogative torture does work, and the torturer gets the information he was looking for.17

Well, one might say, but at least interrogative torture is not very reliable. Apart from the fact that even that is not so clear, it would not even help. Consider the following case: An innocent person is being attacked by an aggressor, who fires a deadly weapon at him (and misses him at first but keeps firing). The attacked person's only possibility to save his life is by using the One-Million-Pains-To-One-Kill-Gun that he happens to have with him. On average, you have to pull the trigger of this gun one million times in order to have one immediately incapacitating projectile come out of it (it incapaci tates through the infliction of unbearable pain). All the other times it fires projectiles that only ten seconds after hitting a human target cause the target unbearable pain. Thus, firing this gun at the aggressor will certainly cause the aggressor unbearable pain, but the probability that it will save the life of the defender is only 1:1,000,000. I must admit that I cannot even begin to make sense of the suggestion that, given these odds, the defender should not use the gun against the aggressor. Yes, the pain inflicted by the weapon on the aggressor is extremely unlikely to secure the survival of the

Society for Applied Philosophy, 2006

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

Torture — The Case for Dirty Harry and against Alan Dershowitz 343

defender, but there still is a chance that it will, so why should the defender forgo this chance for the benefit of the aggressor? Obviously, there is no reason (at least none that I could see). Again, the application to the torture of ticking bomb terrorists and Dirty Harry kidnappers is obvious.

What is the second reply? Richard H. Weisberg takes issue with the example of the ticking bomb case:

. . . the hypothetical itself lacks the virtues of intelligence, appropriateness, and especially sophistication. Here, as in The Brothers Karamazov — pace Sandy Levinson — it is the complex rationalizers who wind up being more naive than those who speak strictly, directly, and simply against injustice. 'You can't know whether a person knows where the bomb is', explains Cole in a recent piece in the Nation, 'or even if they're telling the truth. Because of this, you wind up sanctioning torture in general'.18

To begin with, by allowing torture in the ticking bomb case one does not necessarily wind up sanctioning it in general. Killing (of certain people) is sanctioned in war but not in general. The actual second reply I was referring to, however, is that you do not know whether you have the right person. (That you do not know whether the person speaks the truth is simply the first reply. We have already dealt with it.) But what does it mean: 'You don't know'? Does it mean you do not know for certain? If not knowing for certain whether you have the right person would be sufficient reason not to harm that person, we would not only have to abstain from self-defence but also from punish ment. You never know for certain!

Take the example of a man who draws a gun in front of a head of state and aims at him. The bodyguards simply cannot know (for certain) whether this person wants to shoot, they cannot even know whether it is a real gun. Maybe the 'attacker' is only a retard with a water pistol. So the bodyguards of the head of state (whom we want to assume innocent) should not shoot at such a person who, for all they can know, seems to be attacking the person they are to protect? Actually, if shooting (and probably killing) him is the only way to make sure that he is not able to pull the trigger, they should shoot him.

One might say that this person, even if he does not really want to shoot and has no real gun, at least feigns an attack, and this makes him liable to counter-attack. Whoever credibly feigns an attack on another person cannot later, after having suffered from severe countermeasures, complain that he 'only' feigned it. He shouldn't have feigned it at all. We can, however, have a comparable situation with a terrorist. If a person says to other persons that he is going to build a powerful bomb to blow up a kindergarten and has the necessary skills and buys the necessary chemicals, he had better not, when the security service storms his hideout where he is surrounded by his bomb-making equipment, sneeringly say: 'You are too late. I have already planted the bomb. It will go off in 12 hours and kill hundreds of children'. If he then is tortured by the security service, which wants to find out where the bomb is, he is not, it seems, in a particularly good position to complain about that even if he has not planted a bomb.19 Moreover, even if a real or supposed terrorist has not made that particularly threatening state ment, hanging around with the wrong people in the wrong situations can also make you liable to attack. Suppose an innocent woman is hunted by a mob. Maybe they do not like her skin colour, her ethnic group, her religion or whatever. The mob has

© Society for Applied Philosophy, 2006

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

344 Uwe Steinhoff

already killed other people for the same reason. She hides with the hand grenade she fortunately has, behind a bush. Suddenly one of the group sees her, points his finger at her, shouts 'There she is', and the armed members of the group raise their guns to shoot at her. Not all members of the group have guns. Some are unarmed and shout: 'Kill her, kill her!' Others do not even shout but sneer, foaming at the mouth (I am not talking about completely innocent people, who just 'happen' to be there for no fault of their own). The only way she can save herself is to throw the grenade at the mob, which will kill all of them, including the unarmed ones. Is she justified in doing so? I would think so. Being a member of certain groups that collectively undertake aggres sive acts or intentionally pose a threat to innocent people makes one liable to severe countermeasures. Consequently, a member of a terrorist group might be liable to torture in the ticking bomb case, even if he does not know were the bomb is.

It helps, by the way, very little to aver at this point that torture is simply not compatible with liberalism. David Luban, for example, claims that torture aims 'to strip away from its victim all the qualities of human dignity that liberalism prizes' and that 'torture is a microcosm, raised to the highest level of intensity, of the tyrannical political relationships that liberalism hates the most'.20 However, prisons are also 'mi crocosms' of tyranny; yet, most liberals do not find them incompatible with liberalism. Where is the difference? Maybe it lies in the fact that in torture tyranny is 'raised to the highest level'. But, first, it is far from clear that one hour of torture is more tyrannical than 15 years of prison. Second, even if torture were more tyrannical than prison, and liberalism abhorred tyranny, there remained still the fact that liberalism can accommo date quite intense forms of tyranny, such as incarceration for life (or for a decade and more). Why should it not also be able to accommodate the most extreme form of tyranny? 'Because it is the most extreme form' is in itself no answer to this question.

More importantly, liberalism is not so much about 'dignity' — which is a quite elusive concept, anyway (in particular, I deny that the dignity of the culpable aggressor is violated by Dirty Harry's action any more than it would be violated by Dirty Harry's killing him in self-defence) — but about liberty. It is called liberalism, not 'dignism'. It is also not about just anybody's liberty. It is about the liberty of the innocent. This is why there is no particular problem in liberalism to kill aggressors or to deprive them of their liberty if this is the only way to protect innocent people from these aggressors. The core value of the liberal state is the protection of the liberty and the rights of innocent individuals against aggressors. The state can be such an aggressor, but the state can and must also protect against other aggressors. To keep Dirty Harry in the situa tion described from torturing the kidnapper, therefore, would run against the liberal state's own raison d'etre. The state would help the aggressor, not the victim; it would help the aggressor's tyranny over the innocent and therefore actually abet the relation ship it hates the most.

Since my description of the core value of liberalism, as I submit, is at least as plausible as Luban's (and I think it is historically much more plausible), the appeal to liberalism cannot help absolute opponents of torture. To claim that liberalism 'cor rectly understood' absolutely prohibits torture simply engages in an attempt of persua sive definition and begs the question. Besides, why could liberalism, 'correctly understood', not be wrong?

But — speaking about the innocent — what about the risk of torturing a completely innocent person, a person that made itself not liable? Yes, that risk exists, as it does in

Society for Applied Philosophy, 2006

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

Torture — The Case for Dirty Harry and against Alan Dershowitz 345

the case of punishment. In the latter case, the risk of punishing innocent persons has to be weighed against the risk of not at all punishing the non-innocent and of not at all deterring potential criminals. In the case of interrogative torture in the context of a ticking bomb situation, the risk of torturing an innocent person has to be weighed against the risk of letting other innocent persons die in an explosion. If the weighing process in the former case can justify punishment, it is unclear why the weighing process in the latter case could not sometimes justify torture. If the odds are high enough, it does. In fact, the justification in the latter case might even be easier — easier at least than justifying capital punishment, for death, as already noted, is worse than torture (at least for most people who are confronted with a decision between their death and being tortured for a limited time). It might even be easier than justifying incarceration for one or two decades, for it is not clear that many persons would not prefer some hours or even days of torture to that alternative.

To sum up the discussion so far: A compelling argument for an absolute moral prohibition of torture cannot be made. Under certain circumstances torture can be justified. Justified, not only excused. I emphasise this because some philosophers claim that situations of so-called necessity or emergency can only excuse torture (and some other extreme measures). But there is nothing in the meaning of the terms 'necessity' or 'emergency' themselves that could warrant that view. For example, the German penal code distinguishes between 'justifying emergency' and 'excusing emergency'. § 34 reads:

Whosoever, in order to avert a not otherwise avoidable present danger to life, body, freedom, honour, property, or another legally protected interest, acts so as to avert the danger to himself or others, does not act illegally if, upon consideration of the conflicting interests, namely of the threatened legally protected interests and of the degree of the threatened danger, the protected interest substantially outweighs the infringed interest. This, however, is true only if the act is an adequate [in the sense of means-end rationality ('angemessen')] means to avert the danger.21

He does not act illegally, and that means his act is legally justified. If the protected interests do not substantially outweigh the infringed interest, however, he can at best be excused (§ 35). The moral case is not different. There can be situations where torture is an instrumentally adequate and the only means to avert a certain danger from certain morally protected interests and where the protected interests substantially outweigh the infringed ones. Therefore, if the odds are high enough, torture can be not only excused but morally justified.

No doubt, an absolutist opponent of torture will not be particularly impressed by the argument offered so far. In fact, absolutists normally (although perhaps not always) do not even try to refute the arguments adduced against their absolutist positions; they tend to just persistently and dramatically reaffirm their positions. The writer and poet Ariel Dorfman is a good example:

I can only pray that humanity will have the courage to say no, no to torture, no to torture under any circumstance whatsoever, no to torture, no matter who the enemy, what the accusation, what sort of fear we harbor; no to torture no matter what kind of threat is posed to our safety; no to torture anytime, anywhere; no to torture anyone; no to torture.22

© Society for Applied Philosophy, 2006

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

346 Uwe Steinhoff

Moral absolutism is a dangerous and mistaken view. If, for example, humanity would face the choice (maybe posed by some maniac with the ultimate weapon or by an alien race, or what have you) between being exterminated or torturing one particularly bad man (let us say Idi Amin) for an hour or a day, it is far from clear why any person in his right mind — both intellectually and morally — should pray that humanity said 'no to torture'.23 And what, by the way, if the choice is between all human beings (that includes children) being tortured by the alien race or only one particularly bad man being tortured by some humans? Consequences count; they cannot simply be ignored for the benefit of some allegedly absolute rule, especially if they might be catastrophic. Fiat justitia, pereat mundus is an irrational and immoral maxim.

To say it again: A compelling argument for an absolute moral prohibition of torture cannot be made. But what about the legal prohibition of torture? If torture can be morally justified under certain circumstances, should it also be legalised?

It could seem that torture is already legal under the German penal code. For if, as I claimed, the interests which are protected by torturing a terrorist can substantially outweigh the infringed interests (most notably of the terrorist), then torture must be legal. However, the fact that this outweighing can occur from a moral perspective does not yet mean that it can also occur from the legal perspective of a certain penal code. Each system of laws is in principle free to stipulate an absolute prohibition of torture. The German law does this in so far as it accepts certain international absolute prohibi tions of torture as binding. That is, according to German law nothing can (legally) outweigh the interest of not being tortured (or at least of not being tortured by state officials or agents acting on behalf of a state). That torture is illegal under all circum stances in a given system of law, however, does not exclude the possibility that the practice might under some circumstances be excused by the law. It seems to me that it would be reasonable to excuse it under some circumstances.24 I shall, however, say nothing further on this topic here.

Instead, I shall say something about the proposal to justify torture before the act (rather than excusing it ex post). The lawyer Alan Dershowitz made the infamous suggestion to introduce legal 'torture warrants', issued by judges.

... it is important to ask the following question: if torture is being or will be practiced, is it worse to close our eyes to it and tolerate its use by low-level law enforcement officials without accountability, or instead to bring it to the sur face by requiring that a warrant of some kind be required as a precondition to the infliction of any type of torture under any circumstances?25

And he states:

My own belief is that a warrant requirement, if properly enforced, would probably reduce the frequency, severity, and duration of torture. I cannot see how it could possibly increase it, since a warrant requirement simply imposes an additional level of prior review. .. . here are two examples to demonstrate why I think there would be less torture with a warrant requirement than without one. Recall the case of the alleged national security wiretap being placed on the phones of Martin Luther King by the Kennedy administration in the early 1960s. This was in the days when the attorney general could authorize a national security wiretap without a warrant. Today no judge would

Society for Applied Philosophy, 2006

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

Torture — The Case for Dirty Harry and against Alan Dershowitz 347

issue a warrant in a case as flimsy as that one. When Zaccarias Moussaui was detained after trying to learn how to fly an airplane, without wanting to know much about landing it, the government did not even seek a national security wiretap because its lawyers believed that a judge would not have granted one.26

A few things must be said concerning this argument. First, closing one's eyes to the practice of torture is not the only alternative to the introduction of torture warrants. Unfortunately, Dershowitz seems to have difficulties grasping the difference between closing one's eyes on the one hand and exposing and condemning on the other. To wit, he criticises William Schulz, the executive director of Amnesty International USA, who asks whether Dershowitz would also favour brutality warrants and prisoner rape warrants. (Dershowitz answers with a 'heuristic yes', whatever that is supposed to be.)27 And he quotes himself saying: 'My question back to Schulz is do you prefer the current situation in which brutality, testilying and prisoner rape are rampant, but we close our eyes to these evils?'28 Who is 'we'? Certainly not Schulz or Amnesty International.29

Second, Dershowitz admits that he 'certainly cannot prove . . . that a formal require ment of a judicial warrant as prerequisite to nonlethal torture would decrease the amount of physical violence directed against suspects'.30 It seems, however, that Dershowitz should offer something more than his personal 'belief' and two examples to back the quite grave proposal to legalise torture. That he does not displays a lightness about the matter which is out of place. To be sure, he also adduces John H. Langbein's historical study of torture,31 and although he concedes that it 'does not definitely answer' 'whether there would be less torture if it were done as part of the legal system', he thinks that it 'does provide some suggestive insights'.32 Yet, before drawing 'suggestive insights' from Langbein's study and from history, one should get both straight. Dershowitz does not.33 In fact, Langbein leaves no doubt that torture was not part of the judicial system in England. Not only 'law enforcement officers' but also the courts (and judges) could not warrant torture. Langbein even states:

The legal basis, such as it was, for the use of torture in the eighty-one known cases appears to have been the notion of sovereign immunity, a defensive doctrine that spared the authorities from having to supply justification for what they were doing.34

The facts, then, are that torture was never part of the English judicial system (if it was ever legal in England at all) whereas it zvas part of the Continental legal system. Extensive (not to say epidemic) use was made of torture on the Continent but not in England. Obviously, these facts suggest insights quite different from the ones Dershowitz comes up with.

Moreover, it is also funny that Dershowitz thinks that his two examples support his case. What his examples show (if they show anything) is that an attorney general who is authorised to put a wiretap without judicial warrant is more likely to put a wiretap than an attorney general who does need a warrant. However, the question to be answered is whether torture would be less likely under a requirement of a judicial warrant than under a total ban. To suggest a positive answer to this question by way of an analogy, Dershowitz would have to compare a legal arrangement in which the

Society for Applied Philosophy, 2006

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

348 Uwe Steinhoff

attorney general is prohibited from putting a wiretap with a legal arrangement where he is authorised to do so if he has a warrant. Dershowitz does not do that. It is he who

engages in 'tortured reasoning', to use his term,35 not his critics. Finally, why shouldn't state agents who do not get a warrant torture anyway? They

do not get a warrant today, and some of them torture anyway. Dershowitz answers that:

. . . the current excuse being offered — we had to do what we did to get information — would no longer be available, since there would be an author ized method of securing information in extraordinary cases by the use of extraordinary means.36

First, people who escape detection are not in need of excuses to wriggle out of punish ment in the first place. Besides, the excuse would be available. It would be: 'Since the judge didn't give us the warrant — he did not realise the seriousness of the situation (or there wasn't enough time) — we just had to torture under these circumstances without a warrant in order to get the information and to avoid a great evil'.

In short, Dershowitz has not offered the slightest bit of evidence — not even anecdotal — for his bold claim that the introduction of torture warrants would reduce torture or

as much as increase accountability. Yet there is very good evidence to the contrary. Since Dershowitz invited us to draw suggestive insights from history, especially on the basis of Langbein's study, it might be worthwhile to note what Langbein himself has to say:

Another insight from history is the danger that, once legitimated, torture could develop a constituency with a vested interest in perpetuating it.37

And that, to draw the conclusion Dershowitz isn't able to draw, would hardly help to reduce torture or to increase accountability.

But why should we try to reduce torture? After all, in the first part of this paper I have argued that no compelling argument for an absolute moral prohibition of torture can be made; yes, not even for a prohibition in the Dirty Harry cases. I have also argued that torture is not worse than death and probably not worse than a decade of incar ceration. So since we have legal incarceration, why shouldn't we have legal torture too?

One very straightforward answer is: because we don't need it. The ticking bomb case or the Dirty Harry case is a very rare case. In fact, it is safe to assume that all the torture that happened or happens in Abu Ghraib, Afghanistan and Guantanamo sim ply has nothing to do with ticking bombs or hostages who are about to die. The same holds for the overwhelming majority of all other cases of torture. Ticking bomb and Dirty Harry cases are exceptions. An emergency or necessity paragraph along the lines of § 35 of the German penal code can deal with such exceptions, and perhaps not even that is needed. If the stakes are high enough and no other option is available, police officers or other state agents will probably use torture even if they face prosecution if caught (that is, incidentally, what Dershowitz himself claims). Besides, if punished, they might still be allowed the benefit of mitigating circumstances.

Second, that being tortured (or torturing someone) is not necessarily worse than being killed or incarcerated for a long time (or than killing someone or incarcerating him for a long time) does not imply that introducing a wider practice of torture is not worse than introducing or maintaining a wider practice of incarceration or killing. Dershowitz, for example, acknowledges:

© Society for Applied Philosophy, 2006

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

Torture — The Case for Dirty Harry and against Alan Dershowitz 349

Experience has shown that if torture, which has been deemed illegitimate by the civilized world for more than a century, were now to be legitimated — even for limited use in one extraordinary type of situation — such legitimation would constitute an important symbolic setback in the worldwide campaign against human rights abuses.38

However, he thinks:

It does not necessarily follow from this understandable fear of the slippery slope that we can never consider the use of nonlethal infliction of pain, if its use were to be limited by acceptable principles of morality. After all, imprison ing a witness who refuses to testify after being given immunity is designed to be punitive — that is painful. Such imprisonment can, on occasion, produce more pain and greater risk of death than nonlethal torture.39

It does indeed not follow that we can never consider the use of non-lethal infliction of

pain, but it does follow that institutionalising torture — for example with torture war rants — is a bad idea. In particular, the analogy with the practice of coercing witnesses through imprisonment into testifying is misleading. The practice is designed to be punitive, yes, but that is not the same as being designed to be painful. Not every aversive treatment causes pain. It is important not to blur the distinctions. Further, the very fact that imprisonment produces only on occasion more pain and greater risk of death than non-lethal torture (although I suppose that non-lethal imprisonment would carry no risk of death) shows that it is not designed to produce pain and death. After all, being released can, on occasion, also produce more pain and greater risk of death than non-lethal torture. But how is that supposed to support the case for torture or for torture warrants? Thus, by using imprisonment as a method of punishment we are not already on the slippery slope.

Even if legalising torture puts us on a slippery slope, couldn't we stop the slide downwards? Dershowitz proposes a 'principled break':

For example, if nonlethal torture were legally limited to convicted terrorists who had knowledge of future massive terrorist acts, were given immunity, and still refused to provide the information, there might still be objections to the use of torture, but they would have to go beyond the slippery slope

40

argument.

Actually, one argument that could be made here is that a convicted terrorist will hardly be a ticking bomb terrorist, unless, of course, he has set the time fuse on a few months or even years in the future or his conviction was made without due process. Giving up due process, however, does not look very much like a 'principled break', at least if the principle is supposed to be compatible with the rule of law. That notwith standing, it has to be admitted that 'massive terrorist acts' will have to be planned long enough in advance so that a convicted terrorist might have knowledge of them. Conse quently, torturing him might be a means to thwart the attacks.

However, Dershowitz's talk about a 'principled break' does, in fact, not address the problem of an 'important symbolic setback in the world-wide campaign against human rights abuses' at all. The symbolic setback consists precisely in undermining the abso lute prohibition on torture and cannot be compensated, probably not even mitigated,

© Society for Applied Philosophy, 2006

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

350 Uwe Steinhoff

by recourse to alleged 'principled breaks'. Moreover, the whole idea of a 'principled break' in connection with 'security laws' that cut down on civil liberties and individual rights is rather naive. (I put 'security laws' in quotation marks because cutting down on civil liberties and individual rights hardly increases an individual's security from the state — the political entity, it should be remembered, that has slaughtered more people than any other political entity in history and is certainly more dangerous than any sub national terrorist organisation.) Experience shows that measures introduced against putative terrorists in alleged conditions of emergency tend to be doubly extended, namely, beyond the emergency and to crimes or offences of lesser seriousness. In the UK, for example, emergency anti-terrorist measures, such as limitations on the right to silence, admissibility of confession evidence and extended periods of pre-judicial detention, have infiltrated ordinary criminal law and procedure.41 Once advertised as being targeted only against terrorists, they can now befall any citizen who gets involved in criminal procedure.

It is to be expected, then, that the legalisation of torture for certain specific circum stances will, just like other so-called security laws, come with an inherent 'metastatic tendency'42 that in time extends it beyond those circumstances. Apart from this dan gerous tendency of 'security laws' in general there is, in addition, something very special about torture. Jeremy Waldron has argued that the prohibition of torture is archetypical of the idea:

. . . that even where law has to operate forcefully, there will not be the connec tion that has existed in other times or places between law and brutality. People may fear and be deterred by legal sanctions . . . they may even on occasion be forced ... to do things or go places against their will. But even when this happens, they will not be herded like cattle or broken like horses; they will not be beaten like dumb animals or treated as bodies to be manipulated. Instead, there will be an enduring connection between the spirit of law and respect for human dignity — respect for human dignity even in extremis, where law is at its most forceful and its subjects at their most vulnerable.43

That the prohibition of torture is a legal archetype means that it has 'a significance stemming from the fact that it sums up or makes vivid to us the point, purpose, principle, or policy of a whole area of law'.44 For example, Waldron shows that decisive court rulings against lesser forms of police brutality — lesser, that is, than torture — were made with reference to torture. The similarities with torture were invoked to

reject those other brutalities. This, of course, would not be possible if torture itself became regularised and justified by law, for the similarity with a regular legal practice could hardly count against some other practice. As Waldron puts it:

The idea is that our confidence that what lies at the bottom of the slope (torture) is wrong informs and supports our confidence that the lesser evils that lie above torture are wrong too.45

Thus, by undermining the archetype of the prohibition of torture one also undermines the prohibition of lesser forms of brutality. The whole set of injunctions against brutal ity would unravel and the character of the legal system would be corrupted.46

What is so frightening about such a brutalisation of the legal system is that it is also the brutalisation of its enforcer — which, in modern societies, is ultimately the state. It

Society for Applied Philosophy, 2006

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

Torture — The Case for Dirty Harry and against Alan Dershowitz 351

is one thing to grant to an individual in a certain situation the moral justification to torture another individual; it is a completely different thing to allow the state to legally institutionalise torture in certain circumstances. Dirty Harry has to justify himself not only morally but also legally. He might face legal charges, and that might make him think twice before he tortures someone. This, in fact, ensures that the slope of the moral permission of torture in certain cases does not become too slippery to be accept able. Dirty Harry takes his decision as an individual, not as an agent of the state. The state is not behind him. But if law enforcers can resort to torture knowing in advance that the state is behind them, the worst has to be expected — on a large and inevitably growing scale. Here it is worth noting that the argument that the prohibition of torture is an archetype and the argument that the legal introduction of torture would have a metastatic tendency reinforce each other. The further the practice of torture extends, the more it will undermine the archetypical character of the prohibition; the more that happens, the further the practice will extend. It is not only a slippery slope but also a slope that on its way down gets exponentially steeper. One of the functions of the rule of law is to keep the power of the state under control. But this doesn't work with any law. It doesn't work with a brutal or brutalised one. Torture warrants are indeed a

'stunningly bad idea'.47

Uwe Steinhoff, Department of Politics & International Relations, University of Oxford Manor Road, Oxford, OX1 3UQ, UK. [email protected]

NOTES

1 H. Shue, 'Torture' in S. Levinson (ed.) Torture: A Collection (Oxford: Oxford University Press, 2004), pp. 49-60, at p. 49.

2 Ibid.

3 Ibid., pp. 5Of. 4 Ibid., p. 51. 5 Henry Shue has meanwhile in conversation distanced himself somewhat from drawing a parallel between

the prohibition of torture and jus in bello requirements.

6 David Rodin claims that, contrary to torture victims, soldiers in trenches are not entirely defenceless, for they could surrender, retreat or dig into deeper trenches; and, had they been better prepared, they would have found more protective shelter or been equipped with longer-range guns (personal communication). However, had the torture victim been better prepared she might have escaped or killed herself in order not to become a torture victim in the first place. But if she indeed has become a torture victim, she is defenceless under the torture, as are the soldiers in the trenches who are not lucky (or good) enough to have suitable weapons. As to the first point, the torture victim can surrender too and disclose the information. Such surrender might not always be accepted, but likewise the surrender of soldiers in trenches is not always seen by artillerists miles away or hidden behind a cloud of smoke. Finally, to suggest that trench soldiers subjected to heavy artillery fire should in that situation be able to dig into deeper trenches seems to be as unrealistic as to suggest that a torture victim should protect herself by yoga or meditation techniques.

7 Jeff McMahan thinks that the relevant principle here is not proportionality but minimal force. (Personal communication) I disagree, but the disagreement may simply be a semantic one, namely one about what 'proportionality' means within the context of jus in bello. Be that as it may, the important point for the purposes of the present discussion is that an alleged principle of the immunity of the defenceless plays no role here.

8 H. Miinkler, Die neuen Kriege (Reinbek bei Hamburg: Rowohlt, 2003), p. 234, my translation. 9 D. Sussman, 'What's wrong with torture?', Philosophy and Public Affairs 33, 1 (2005): 4.

10 Ibid., n. 10.

© Society for Applied Philosophy, 2006

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

352 Uzve Steinhoff

11 Ibid., p. 4. 12 Ibid., p. 21. 13 /tai., p. 30. 14 The example is a variant of Bernard Williams's famous example of Jim and the twenty Indians. 15 S. Levinson, 'The debate on torture: war on virtual states', Dissent (Summer) (2003): 79-90, at pp. 83f. 16 P. Montague, 'The morality of self-defense: a reply to Wasserman', Philosophy and Public Affairs 18

(1989): 81-89, at pp. 81f. Ryan advocates a similar position to Montague's. See C. C. Ryan, 'Self defense, pacifism, and the possibility of killing', Ethics 93 (1984): 508-524, at pp. 515ff.

17 See S. Levinson, 'Contemplating torture: an introduction' in Levinson (2004) op. cit. (see n. 1), pp. 23 43, at pp. 33ff., and the further references there.

18 R. H. Weisberg, 'Loose professionalism' in Levinson (2004) op. cit., pp. 299-305, at p. 304. 19 Jeff McMahan agrees that this person cannot complain under the circumstances but thinks that this still

does not make him liable to be tortured, the reason being that 'torturing him serves no purpose under the circumstances'. (Personal communication) However, if he cannot complain, he cannot be being wronged (for then he obviously could complain); and I think that the only possible reason why someone is not wronged by an attack (for example in the form of torture) is that he is liable to the attack. In fact, this seems to be pretty much the meaning of 'liable to attack'. Besides, if someone unjustly shoots at me and I shoot back, hitting him in the shoulder, and he continues shooting and kills me, then my counterattack has served no purpose under the circumstances (if it had, I would not be dead). That makes my counter attack hardly unjust or the attacker not liable to be shot at (otherwise every unsuccessful defender would also be an unjust defender, which is absurd).

20 D. Luban, (2005) 'Liberalism and the unpleasant question of torture', http://ethics.stanford.edu/ newsletter/_old/december/Liberalism%20and%20the%20Unpleasant%20Question%20Question%20of% 20Torture.doc, accessed on 2 October 2005, electronic resource. A comparable argument is put forward by K. Seth, 'Too close to the rack and the screw: constitutional constraints on torture in the war on terror', University of Pennsylvania Journal of Constitutional Law 6 (2003-2004): 278-325.

21 The translation is mine.

22 A. Dorfman, 'The tyranny of terror: is torture inevitable in our century and beyond?' in Levinson (2004) op. cit., pp. 3-18, at p. 17.

23 Torturing this person would, of course, be a case of self-preservation and not of self- or other-defence (or something close to it) as in the Dirty Harry case.

24 That seems to be the position of Shue op. cit., pp. 58f., and R. A. Posner, R. A. 'Torture, terrorism, and interrogation' in Levinson (2004) op. cit., pp. 291-298, at pp. 297f.; and it is the position of O. Gross, O. 'The prohibition on torture and the limits of the law' in Levinson (2004) op. cit., pp. 229-253, esp. at pp. 231 and 239-250.

25 A. Dershowitz, 'Tortured reasoning' in Levinson (2004) op. cit., pp. 257-280, at p. 257. He emphasises that that was his question and not the 'old, abstract' one 'over whether torture can ever be justified', and he complains about 'misleading' descriptions of his proposals. Ibid., p. 266. Maybe the next time he addresses the former question instead of the latter he could help to avoid 'misleading' descriptions of his intentions by not using titles like 'Should the Ticking Bomb Terrorist Be Tortured?' See A. Dershowitz, Why Terrorism Works: Understanding the Threats, Responding to the Challenge (New Haven, CT: Yale University Press, 2002) p. 131.

26 Dershowitz (2004) op. cit., pp. 270f. 27 Ibid., pp. 266f. 28 Ibid., p. 267. ' "Testilying" is a term coined by New York City police to describe systematic perjury

regarding the circumstances that led to a search, seizure, or interrogation'. Ibid., p. 278, n. 13. 29 Compare E. Scarry, 'Five errors in the reasoning of Alan Dershowitz' in Levinson (2004) op. cit.,

pp. 281-290, at p. 288. 30 Dershowitz (2002) op. cit., p. 158. 31 J. H. Langbein, Torture and the Law of Proof: Europe and England in the Ancien Regime (Chicago, IL:

University of Chicago Press, 1977). 32 Dershowitz (2002) op. cit., p. 158. 33 On Dershowitz's misreading of Langbein see also J. Waldron, 'Torture and positive law: jurisprudence for

the White House', Columbia Law Review 105 (2005): 1739, n. 250. 34 J. H. Langbein, 'The legal history of torture' in Levinson (2004) op. cit., pp. 93-103, at p. 100. 35 Dershowitz (2004) op. cit., p. 257.

© Society for Applied Philosophy, 2006

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms

Torture — The Case for Dirty Harry and against Alan Dershowitz 353

36 Ibid.j p. 276. 37 Langbein (2004) op. cit., p. 101. 38 Dershowitz (2002) op. cit., p. 145. 39 Ibid., p. 147. 40 Ibid.

41 P. Hillyard, 'The normalization of special powers from Northern Ireland to Britain' in N. Lacey (ed.) A Reader on Criminal Justice (Oxford: Oxford University Press, 1994); O. Gross, 'Cutting down trees: law-making under the shadow of great calamities' in R. J. Daniels, P. Macklem and K. Roach (eds.) The Security of Freedom: Essays on Canada's Anti-Terrorism Bill (Toronto: Toronto University Press, 2001), pp. 39-61, esp. at 47ff. I owe the references to these articles to L. Zedner, 'Securing liberty in the face of terror: reflections from criminal justice', pp. 7 and 15, and to C. Warbrick, 'Terrorism, counter-terrorism,

international law', p. 9, unpublished papers held at the colloquium Moral and Legal Aspects of Terrorism, Corpus Christi College, Oxford, 5 March 2005.

42 Shue op. cit., p. 58. 43 Waldron op. cit., pp. 1726f. 44 Ibid., p. 1723. 45 Ibid., p. 1735. 46 Ibid., p. 1728-1739. 47 J. B. Elshtain, 'Reflections on the problem of "dirty hands"' in Levinson (2004) op. cit., pp. 77-89, at

p. 83. I owe thanks to Patrick Lenta, Jeff McMahan and David Rodin for helpful comments on an earlier draft of this paper.

© Society for Applied Philosophy, 2006

This content downloaded from 160.253.64.26 on Wed, 10 Oct 2018 18:29:02 UTC All use subject to https://about.jstor.org/terms