~ gabriella giudici

Why is torture morally wrong? This question has been neglected or avoided by recent moral philosophy, in part because torture is by its nature especially difficult to discuss. Torture involves degrees of pain and fear that are often said to be utterly indescribable; indeed, these experiences are sometimes said to destroy in their victims the very hope of any sort of communication or shared experience whatsoever.¹ Torture has proved surprisingly difficult to define.² There is no clear agreement on the distinction between torture, coercion, and manipulation, or whether such techniques as sleep and sensory deprivation, isolation, or prolonged questioning should count as forms of torture.³ In addition, we may be fearful of deriving some sort of perverse titillation from the subject, or of being able to dispassionately contemplate the agonies of real victims of torture. Those who have not suffered torture may well feel it is not their place to offer any very substantive reflections on the practice, leaving the issue to those who unfortunately know what they are talking about. We might also worry that in just raising the question, we inadvertently give aid and comfort to torturers, if only by supplying materials for disingenuous self-justification.

On the other hand, if we approach the question of torture’s justifiability in good faith, it may seem so easy to answer as to be hardly worth asking. Torture involves the intentional infliction of extreme physical pain or psychological distress on a person, for such ends as inducing the betrayal of some cause or intimate, intimidating actual or potential opponents, or as an exercise of dominance or sadism simply for its own sake. Since at least Beccaria there has been a broad and confident consensus that torture is uniquely “barbaric” and “inhuman”: the most profound violation possible of the dignity of a human being. In philosophical and political discussions, torture is commonly offered as one of the few unproblematic examples of a type of act that is morally impermissible without exception or qualification.⁴

Yet in the current “war on terrorism,” the thought that torture may be an appropriate means of combating terrorists has emerged in respectable political and legal discussions in the United States and elsewhere.⁵ We would not hope to justify torture that was meant to terrorize innocent civilians or dehumanize a population pursuant to expulsion or genocide.⁶ But could we justify the torture of a terrorist in order to find out the location of a bomb he has planted, or of a kidnapper to make him reveal the location of his captive while she might still be alive?⁷ Here, torture might be understood as the sort of violence that can be permissible as part of the prosecution of a just war or legitimate police action, especially against people who have shown nothing but contempt for the laws of war and the rights of their victims. Nor is it entirely obvious why we should refuse to consider torture as a possible form of punishment, especially when we are willing to allow punishments of lengthy incarceration and perhaps even death. Yet while a convict might reasonably prefer torture to these other punishments, punitive torture remains officially beyond the pale even in an America that countenances the executions of juvenile and mentally retarded offenders. What is it about torture that sets it apart even from killing, maiming, or imprisoning someone, such that the circumstances that might justify inflicting such harms would not even begin to justify torture?⁸

In this article, I defend the intuition that there is something morally special about torture that distinguishes it from most other kinds of violence, cruelty, or degrading treatment. Torture is all these things, of course, and is morally objectionable simply as such. What I deny, however, is that the wrongness of torture can be fully grasped by understanding it as just an extreme instance of these more general moral categories. I argue that there is a core concept of what constitutes torture that corresponds to a distinctive kind of wrong that is not characteristically found in other forms of extreme violence or coercion, a special type of wrong that may explain why we find torture to be more morally offensive than other ways of inflicting great physical or psychological harm.⁹

My account is not meant to provide any very immediate way of resolving the “ticking bomb” or “Dirty Harry” dilemmas mentioned above. Instead, I am trying to articulate more clearly the moral structure of these dilemmas and the special reluctance we have to consider torture even in the face of such pressing claims. I do not here contend that torture is categorically wrong, but only that it bears an especially high burden of justification, greater in degree and different in kind from even that of killing.¹⁰ Establishing such a special burden of justification may be a necessary step in defending a categorical proscription of torture, but it is not by itself sufficient for such a defense.

While not providing any immediate answers, this account will at least give us reason to resist assimilating these dilemmas to the problems posed by other uses of force in war and police action. My approach is broadly Kantian, but I do not construe the wrong of torture as just that of disregarding, thwarting, or undermining the victim’s capacities for rational self-governance. Instead, I argue 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, torture 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.

My discussion focuses on interrogational torture, i.e., torture that involves a protracted process of inflicting or threatening pain in a context of helplessness and dependence, so as to make its victim provide information, confessions, denunciations, and the like. Such torture is prevalent in the world today, and it is the sort of torture that seems most likely to be justifiable in sufficiently dire circumstances and against sufficiently ruthless and culpable foes. I foreground such torture because in it the central idea of a relationship that is not only morally wrong but also morally perverted finds its clearest illustration. Once this idea of a moral perversion is on the table, I turn to the other context where there may sometimes seem to be prima facie grounds for torture: legal punishment. Such torture is also objectionable as a travesty of the most basic practical relations between embodied agents, but the character of this distorted relation will differ in some important respects from that found in other forms of torture. Again, this examination is not intended to provide any immediate answers about whether torture might ultimately be justified in any particular case. Instead, it is meant only to expand our theoretical vocabulary, so that we can start to say what is morally special about torture without making torture so unique as to be beyond profitable philosophical discussion.

I

The United Nations Convention against Torture and Other Cruel, Inhuman, or Degrading Treatment or Punishment defines torture as

[a]ny act by which severe pain or suffering, whether physical or mental, is intentionally inflicted on a person for such purposes as obtaining from him or a third person information or a confession, punishing him for an act he or a third person has committed or is suspected of having committed, or intimidating or coercing him or a third person, or for any reason based on discrimination of any kind, when such pain or suffering is inflicted by or at the instigation of or with the consent or acquiescence of a public official or other person acting in an official capacity. It does not include pain or suffering arising only from, inherent in or incidental to lawful sanctions.¹¹

At a minimum, torture involves the deliberate infliction of great pain or some other intensely distressing affective state (fear, shame, disgust, and so forth) on an unwilling person for purposes that person does not and could not reasonably be expected to share. While one might accidentally kill or inadvertently maim, one cannot accidentally or inadvertently torture. But there are many ways of deliberately inflicting pain that we would hesitate to call torture. I might punch a stranger in the face, breaking his nose, and then run away. The victim here may experience great pain, but we would not normally say he had been subjected to torture. Two people might simultaneously inflict great pain on each other (e.g., wrestlers each applying submission holds to each other), but we would not normally describe this as torture either.

In addition to the intentional infliction of great pain, torture seems to require that its perpetrators and victims be placed in a distinctive kind of social setting and relationship to one another. Victims of torture must be, and must realize themselves to be, completely at the mercy of their tormentors. This condition involves two distinct elements. First, being at another’s mercy requires that there be a profoundly asymmetric relation of dependence and vulnerability between the parties. The victim of torture must be unable to shield herself in any significant way, and she must be unable to effectively evade or retaliate against her tormenter. I may intentionally inflict great pain in a fight in order to make my foe do something; I may gouge his eyes in order to get him to stop choking me. Nevertheless, insofar as my opponent is not helpless before me, my eye-gouging is not an instance of torture, even though I am trying to force him to comply with my desires by inflicting pain on him. Police who use tear gas to disperse a crowd are not engaging in torture, regardless of how painful the gas may be.

In neither case are the victims forced to be passive before the infliction of suffering, their avenues of response limited to those narrowly defined by their tormentors. Instead, these victims still have it within their power to resist or mitigate the violence done to them: by retreating, devising ways of protecting themselves, or countering their assailants with new threats of their own. In such cases, the attackers are in the position of having to anticipate how their opponents, as free agents, might try to alter or upset the conditions that frame their conflict. In combat, each party recognizes the other to be capable of reshaping the practical task before them in an indefinitely wide variety of ways. In contrast, the torture victim realizes that he has no room to maneuver against his antagonist, no way to fight back or protect himself, and he must realize that his antagonist is aware of this as well. The victim may ultimately comply or not, but he has no prospect of surprising his tormentor in a way that might change the basic shape of the antagonism between them.

Second, the torture victim must see herself as being unable to put up any real moral or legal resistance to her tormentor. The victim takes her tormentor to be someone who can do anything he wants to her, who does not have to worry about answering any challenges that the victim (or her representatives) might put to him. The torturer confronts no moral or legal impediments stemming from his victim’s will, but evidently takes himself to be limited only by his own desires and interests, or the desires and interests of those he serves as an agent. Yet the torturer is in a position to demand anything from his victim as if by an enforceable right. The most intimate and private parts of a victim’s life and body become publicly available tools for the torturer to exploit as he will. The victim is completely exposed, while the torturer is free to conceal anything he likes, even those things to which a victim clearly has a right and a profound interest. Typically, victims are kept in the dark about where they are, why they are being tortured, who might be making the ultimate decisions about their fates, how long they have been confined, or even whether it is day or night. The asymmetry of power, knowledge, and prerogative is absolute: the victim is in a position of complete vulnerability and exposure, the torturer in one of perfect control and inscrutability.

Characteristically, the torture victim finds herself to be not only physically and morally defenseless, but exposed to a will that appears largely if not completely arbitrary.¹² The victim’s greatest interests are completely subject to the caprice of her torturers, who normally conceal just what it is they want or what their ultimate plans are, or represent their goals in inconsistent and ever-shifting ways. Of course, a victim might know that she is being tortured for a specific purpose (to obtain some particular piece of information, perhaps, or to incriminate someone) or that her torturers operate under some significant restrictions (perhaps they have orders not to kill or leave any permanent marks on the victim). Yet even in these cases, the victim’s only grounds for such beliefs about her tormentors’ ends and intentions come from how these tormentors choose to present themselves to her. Typically, a torture victim has no independent way of corroborating any admissions or assurances of her torturers. Insofar as she is able to form any estimate of their motives and intentions, the victim must trust in the sincerity of people who have already shown that they have no scruples about how they treat her. A torturer may seem to want a particular piece of information, or recognize certain kinds of treatment to be off-limits, but such self-imposed restrictions might be abandoned or revised without warning at any moment.

The victim knows nothing of her torturer other than what he wants her to know, save that he is at best indifferent to her rights and interests. Even if the victim is willing to supply the information or confession that seems to be wanted, she has no reason to believe that her tormentor will accept it as accurate and complete. Perhaps she will continue to be tortured “just to make sure,” or for some other reason entirely, or for no reason at all. She can neither verify any claims her tormentor makes, nor rely on any promises or assurances he offers. Instead, the victim can only guess at the real motives and intentions of her torturer, being forced into the position of trusting someone in a context that makes the very idea of trust seem insane.¹³

II

Torture should be distinguished from both coercion and brainwashing, even though all three may often overlap in particular cases. What is distinctive about torture is that it aims to manipulate its victims through their own responses, as agents, to the felt experience of their affects and emotions in a context of dependence, vulnerability, and disorientation. Coercion, in contrast, need only exploit the agent’s rational responses to the cognitive content of these feelings. The coercer tries to influence his victims through their own appreciation of their reasons for action. Coercion presupposes that its victim thinks that his coercers intend to act against his interests should he act or fail to act in a particular way, and that the reason they have adopted such conditional intentions is to give him a stronger reason to do something than he had before. Coercion requires only that its victim have the capacities needed for practical reasoning and intentional action, and that he be able to recognize the expression of these powers in those who are trying to pressure him. Affect and emotion are not required, even though agents will normally take themselves to have very strong interests in avoiding certain kinds of affective experiences. In principle, we could coerce a being that has no emotional life at all (e.g., a corporate agent such as a state or a university), so long as this being had determinate interests that it rationally pursued in part by anticipating the intentions and actions of other rational agents. But we could not in principle torture this sort of artificial person who lacks any distinct sort of emotional or affective life.

Coercion, as a kind of hard bargaining by means of threats, involves too direct an appeal to its victim’s rationality to count as torture. Brainwashing, in contrast, diverges from torture in failing to appeal to its victim’s rational judgment at all. Like the use of drugs or sleep deprivation to put a victim into a hyper-suggestible state, brainwashing exploits the victim’s affects and bodily responses so as to directly subvert or restructure his rational capacities and commitments. It is essential to this process that the victim cannot be fully aware of what is going on. The victim’s beliefs, desires, and perceptions are supposed to be reshaped in a way that reflects the brainwasher’s designs, while still appearing to the victim as being properly responsive to the world and his own authentic concerns. To recognize that one has been brainwashed is to begin to undo the process. In contrast, not only is it possible for someone to realize that she is being coerced, but such a realization is a constitutive feature of coercion as such. To be coerced, I need to do more than realize that someone else will do something I find undesirable should I fail to do what he wants. I must also realize that my coercer has adopted this intention in order to get me to do what he wants, and that I can expect him to anticipate and block my attempts to escape or modify this situation. To be coerced I must see myself as threatened in this way, as confronting another will that intends to consistently frustrate my actions for the sake of its interests. I experience the will of my coercer not just as distinct from my own, as an obstacle or impediment, but as my will’s opponent, as a counter-will that is the systematic negation of my own in some area of activity. In contrast, successful brainwashing requires that its victim ultimately take up and identify with the will of another person, thereby losing any sense that this will is different from or opposed to her own.

Torture exists somewhere between coercion and brainwashing. Like coercion, torture requires its victim to have some minimal understanding of what is being done to him. The victim must realize himself to be at the mercy of someone else who is deliberately trying to get him to act against his own choices and commitments. Yet the torturer is not merely constructing a harsh set of options for the victim to rationally navigate as best he can. The felt experience of pain, fear, and uncertainty are essential elements of torture. Yet the torturer does not set out to exploit the immediate causal consequences of these feelings, as if pain were just a particularly cheap or effective form of truth serum. Rather, in torture the victim must confront his own feelings as a problem, as something he must respond to, where this response is something for which he may see himself to be in some way accountable. Torture is normally accompanied not just by pain and the constant threat of pain and death, but also with by relentless (if evidently pointless) questioning. The victim is presented with a dilemma about submission or resistance (which can also be a dilemma about how or when to properly or effectively submit). And he must confront this dilemma not merely with respect to the disvalue of pain and fear, but while caught up in the experience of these very feelings themselves.

There need not always be a determinate answer as to whether some piece of manipulation is really one of coercion, torture, or brainwashing. A particular act might simultaneously be an instance of more than one type of abuse. The torture victim subjected to repeated electrical shock is also being coerced. Even if he were completely dispassionate, he would realize that he was being given some very compelling reasons to submit to his tormentor. Torture will also sometimes approach brainwashing, insofar as the protracted experience of pain and terror seldom leaves an individual’s rational capacities intact. The experience of torture may simply shatter a mind, making it immediately responsive to the suggestions of an interrogator, just as one may be brought to babble without inhibition or control by the influence of drugs. Any particular act of torture will tend to shade off in one direction or another: the more effectively the torture undermines the victim’s rational capacities, the less effectively it can also be coercing him by appeal to his incentive structure (and vice versa). Yet while there can be such irreducibly overlapping cases, this does not entail that coercion, torture, and brainwashing all exhibit the same basic kind of wrong that varies only in terms of quantitative dimension.¹⁴ As the roles of reason and sensibility shift along this spectrum, the relation of the victim to her tormentor and to herself can assume new shapes that take on different kinds of moral significance.

III

The question now emerges whether torture, so understood, represents a morally distinct and interesting category of action. This question has two aspects. The first is whether torture represents any sort of morally unified category at all, or whether it only refers to a heterogeneous collection of acts that bear various resemblances to one another. The worry here is that even if such acts are individually objectionable in some way or other, there may be no interesting type of wrong that they all involve.¹⁵ The second aspect of the question concerns whether torture, if indeed a morally unified category, is sufficiently basic to be of analytical interest. Is there a special type of moral wrong characteristic of torture as such, or can this wrong (or wrongs) be fully captured by a broader description that might just as readily apply to different kinds of action? If there is such a basic wrong, does appeal to it illuminate what is fundamentally objectionable about torture, or does this appeal turn out to be only a way of repeating our basic conviction that torture is especially or uniquely bad?

The naive utilitarian objection to torture is that it produces tremendous suffering that typically fails to be sufficiently offset by any resulting benefits. A sufficiently nuanced utilitarianism may appeal not just to the intensity and quantity of pain that torture involves but also to the special disutilities associated with the “higher pains” only available to creatures that can experience dread, anguish, and self-disgust.¹⁶ This utilitarian can also point to the lasting psychological and political effects of torture on human beings and institutions. The agony of torture typically continues to reproduce itself in the lives of victims and those close to them long after the physical torments stop.¹⁷ Politically, torture tends to become an entrenched, ever-widening practice, progressively divorced from whatever legitimate aims it might have originally served. Torture that is resorted to as an emergency measure frequently becomes a permanent feature of a regime of terrorization for its actual and potential victims, an education in brutality for its perpetrators, and a corrosive that progressively dissolves the rule of law.¹⁸ The sophisticated utilitarian will also point to the typical inefficiencies and self-defeating effects of torture. Torture is a notoriously unreliable way of gathering intelligence (although perhaps not all that more unreliable than the alternatives usually available). Torture is usually a counterproductive strategy of political control, undermining respect for legal authority and in the long run leaving a subject population more alienated and radicalized than cowed.

The utilitarian focuses on the actual harms involved in torture, and in so doing clearly captures an essential element of what is morally objectionable about such practices. However, utilitarianism will have trouble explaining the moral significance of the social and intentional structure of the “drama” that torture enacts. For it seems that all the harms that typically result from torture might just as readily result from what can sometimes be morally legitimate forms of warfare. It is doubtful that there is any form of pain or injury that can be delivered by a torturer that is categorically worse than the harms that can result from bullets and bombs. The torture victim may experience such terror and helplessness as to leave him permanently shattered psychologically, but so too may besieged soldiers subjected to artillery or aerial bombardments intended to destroy their morale.¹⁹ Yet such tactics are not normally met with the sort of categorical condemnation that the torture of enemy soldiers receives.

If the special wrongness of torture reflected only the special badness of harms it inflicts, then a soldier who could rescue either a number of troops from a firefight or the same number of POWs from torture would be strictly obligated to do the latter, other concerns being equal. I take it that this conclusion does not correspond to how we intuitively understand the special moral status of torture. While we have strong moral reasons to prevent torture, they do not seem to be different in kind from the reasons we have to prevent the traumas of war. But we do seem to have a moral reason not to serve as someone’s torturer that is qualitatively different, and more stringent, than the reasons we have not to make war on him. There seems to be something about the distinctive structure of the relationship of torturer to victim that is intrinsically objectionable and that goes beyond the badness of its usual effects.

Kantian moral theory may seem better suited to capture the distinct moral considerations posed by the structure of the relationship between torturer and victim. The Kantian argues that what is essentially wrong with torture is the profound disrespect it shows the humanity or autonomy of its victim. Here, torture is wrong as the most extreme instance of using someone as a mere means to purposes she does not or could not reasonably share. Although this explanation clearly captures part of what is morally significant about torture, the account is importantly incomplete. The Kantian’s problem complements that of the utilitarian. The utilitarian focuses on the badness of the victim’s agony but cannot readily grasp the significance of the characteristic interpersonal structure of torture. The Kantian can begin to make sense of that structure, but in turn has difficulty explaining why torture seems morally special because it specifically involves pain and other unpleasant feelings, rather than some other way that our ends might be frustrated or our agency disrupted.

For the orthodox Kantian, what is fundamentally objectionable about torture is that the victim, and the victim’s agency, is put to use in ways to which she does not or could not reasonably consent. The fact that it is pain that is characteristically involved is of only indirect importance. What immediately matters to the Kantian is that the victim may reasonably and strongly object to such treatment. The use of pain is significant only insofar as pain is something someone may reasonably and strongly refuse to undergo. But just as the social setting of torture adds a special dimension to its wrongness, so too must the fact that torture involves pain and action directly upon the victim’s body rather than some other intensely unwanted imposition. I may deeply desire that some compromising photographs that have been stolen from me never be made public, and I might even be willing to endure great physical pain in order to prevent this. Yet my blackmailer is not doing anything of a piece with torturing me, even though she is thwarting my will through a means to which she has no right.

The orthodox Kantian can go a little farther toward accommodating the special significance of pain. Unlike other kinds of unwanted imposition, pain characteristically compromises or undermines the very capacities constitutive of autonomous agency itself. It is almost impossible to reflect, deliberate, or even think straight when one is in agony. When sufficiently intense, pain becomes a person’s entire universe and his entire self, crowding out every other aspect of his mental life. Unlike other harms, pain takes its victim’s agency apart “from the inside,” such that the agent may never be able to reconstitute himself fully. The Kantian can thus recognize that torture is not only a violation of the value of rational agency, but a violation that is accomplished through the very annihilation of such agency itself, if only temporarily or incompletely.

This account faces the challenge of showing how disrupting rational self-governance through pain is interestingly different from disrupting self-governance through intense physical pleasure (or any other kind of affect that is not painful). Of course, intense pleasure is usually welcomed by its recipients, and as such may be fully consistent with their own autonomy. Moreover, it is normally quite difficult to “inflict” pleasure on someone against his will. Pain is certainly a simpler and more effective tool for such manipulation, particularly in the absence of sophisticated medical technology. Yet there seems to be no barrier in principle to undermining someone’s agency through ecstasy rather than through suffering. Certain drugs might induce intense euphoria in someone regardless of what he wants and in a way that, like pain, makes it impossible to think or care about anything else. Perhaps such pleasure would make its victim liable to suggestion, such that he would reveal anything under questioning. If of a sufficiently ascetic or puritanical bent, this victim might well object as strongly to such techniques as other people would to electric shock. While such manipulation would clearly be a profound violation of agency, the orthodox Kantian seems compelled to say that it is wrong in precisely the same way as the normal agonistic forms of torture. The Kantian seems unable to do justice to what we would normally take to be a clearly nonaccidental truth: the fact that torture hurts.

Even if the Kantian can account for this asymmetry between pleasure and pain, her view faces yet another hurdle. Any approach that condemns torture as a disruption of agency will have difficulty distinguishing torture from killing. After all, nothing could compromise or undermine rational self-governance more completely than the very death of the agent. Yet while there is a very strong moral presumption against both killing and torturing a human being, it seems that we take the presumption against torture to be even greater than that against homicide. This intuition seems to be just the opposite of what the Kantian account would lead us to expect. We normally think that we may kill in self-defense or as part of the prosecution of a just war. But we are much more reluctant to accept that we may torture in self-defense or as part of proper combat. Although the Kantian might be able to show that torturing a person is not much better than killing him, she seems unable to explain how inflicting torture might actually be more objectionable (or objectionable in a fundamentally different way) than ending his life.

IV

In “Torture,” Henry Shue allows that the circumstances that justify killing could in principle justify torture.²⁰ However, Shue argues that whenever torture is a real option for an agent, that agent cannot be in any such circumstances. We may be justified in using deadly force only against someone who is actively posing some kind of substantial threat to someone or something. Yet if we are in a position to torture someone, Shue observes, we must be in a position of complete power over him, thus able to counter any danger he might pose. We might sometimes have to kill in self-defense, but we could not find ourselves in a situation where we would have to torture in self-defense. For Shue, torture, unlike killing, necessarily violates a basic principle of just combat: the prohibition against attacking the defenseless.

Must potential victims of torture be defenseless in this way? Consider again the captured terrorist who we know to have planted a powerful bomb in some crowded civilian area. Although the terrorist is in our power, he refuses to reveal the bomb’s location, hoping to strike one last blow against us by allowing a train of events that he has set in motion to come to its intended conclusion. In one sense, the terrorist is indeed defenseless. We can do anything we like to him, and there is nothing he can do to resist or shield himself against us. But such helplessness means neither that the terrorist has ceased to engage in hostilities against us, nor that he is no longer an active military threat. His placing of the bomb was the beginning of an attack on us; his silence, although not any kind of further overt act, is nevertheless voluntary behavior undertaken for the sake of bringing that act to completion. His continued silence thus might well be considered a part of his attack, understood as a temporally extended action.

Consider a case where the police confront a very obese man who is trying to suffocate another by sitting on his chest. Like the terrorist, the fat man is defenseless before the police, who can wound or even kill him easily. Yet if the fat man dies or loses consciousness, the police lack the strength to shift him off his victim. Only the fat man can end his attack, even though he does not now need to do anything further in order for it to succeed. Here it seems that the police might well be justified in macing the fat man or twisting his arm (or threatening to kill him) in order to get him to derail a train of events that he has intentionally set in motion, even though he no longer has to actively contribute anything to it. The terrorist relies on a bomb’s mechanism to accomplish his goal, the fat man on his weight, and it is hard to see how this difference of method could be of any great moral significance.

In response to such worries Shue argues that what is important is not only whether a potential target is defenseless, but also whether he has a real opportunity to surrender. In order for an enemy to be a legitimate target, there must be some way he can renounce hostilities, some form of exit from relations of conflict that he knows about and that can be legitimately demanded of him. Yet in the case of interrogational torture, it seems possible to satisfy this condition. After all, the bomb-planting terrorist knows very well what we want (the information), has the power to provide it, and knows that it is only for the sake of such information that we are moved to torture him. Telling us what we want to know might then appear to be something like the laying down of arms that must accompany any real surrender.

Shue replies that an interrogator cannot know for sure in any given case whether his prisoner actually possesses the information wanted. The ignorant terrorist (or unfortunate bystander) would have no way to effectively surrender, and thus no way to exit the relationship of active hostility in which he finds himself. Shue may be right about such cases. However, this objection would establish the general impermissibility of interrogational torture only insofar as our belief that we had the right person needed to meet some standard of proof that we could never reasonably hope to attain. If we had to be absolutely certain that our prisoner has the information we want (and thus the capacity to surrender) then we might indeed never be in a proper position to torture. Yet this would be to hold torture to an epistemic test that we would never use in other areas of just combat or criminal punishment. Soldiers may continue to attack even when they cannot know with absolute certainty whether their enemy is still physically and psychologically capable of surrendering (especially true when warfare is conducted at great speed and distance, as with aerial or artillery bombardment). Yet if we scale back the demand for certainty to keep it in line with other forms of just combat, then torture may turn out to be justified in those cases where we have sufficiently strong evidence that our suspect possesses the information we want. By itself, the mere possibility of torturing the ignorant is no more of an objection to the permissibility of torture than the mere possibility of killing a noncombatant is to just warfare or the bare chance of punishing the innocent is to legal sanctions.

Shue contends we must not only afford our foes the opportunity of surrender, but that opportunity must also be such they can exercise it without having to make any unreasonable sacrifice. We hardly offer our foes any real opportunity for surrender if to do so they would have to allow their families to be killed, or humiliate themselves for a minor military purpose. In a similar vein, Shue argues that we could not require of a prisoner that he betray a cause to which he is committed in order to escape torture. Such a betrayal would so violate the prisoner’s integrity that it would give him no more of a real opportunity for surrender than would the option of suicide. What remains unclear, however, is why we should be so solicitous of the terrorist’s sense of his own integrity. The terrorist disregards the principles of just combat, striking at his enemies’ loved ones simply because they are dear to him. The terrorist makes no effort to distinguish himself from civilians and other noncombatants, forcing his foe into the terrible choice of either waging war against innocents or failing to protect himself and those near to him. Given that the terrorist attacks his enemy’s own integrity this way, it is hard to see how he is entitled to terms of surrender that do not require him to in any way compromise his cause. Plausibly, such terms should be reserved for combatants who accept certain risks (by wearing uniforms, living apart from civilian populations, and so on) and do so in order to allow fighting to proceed without forcing combatants to make such self-disfiguring choices. At least, Shue has given us no reason why the principles that govern the surrender of legitimate combatants should also be applied without modification to those who make no pretense of being bound by anything like the rules of war. Shue’s argument also has the paradoxical consequence that it would be less objectionable to torture a morally sensitive collaborator who is ambivalent about waging war against us (and hence has no very firm or deep commitment to betray) than to torture the hate-filled fanatic who is wholeheartedly bent on our destruction.

V

Our problem has been to understand how torture is essentially different from other kinds of intentional injury, coercion, or killing, and why torture seems harder to justify than even such acts. Whatever solutions we are to offer, they must do justice to the different features of torture that the orthodox Kantian and utilitarian separately call to our attention. Whatever makes torture distinctively bad must have something to do with the sort of interpersonal relationship it enacts, a relationship that realizes a profound violation of the victim’s humanity and autonomy. Yet the special iniquity of torture must also have something to do with the fact that torture characteristically involves the infliction of pain and the use of force directed immediately against the body and the emotions. If torture is morally distinctive in the ways that our intuitions suggest, there must be something about what pain is, and about its special relation to our own agency, that makes some important moral difference. We can preserve torture’s special moral status only if we can find there to be some significant moral difference between being used as a mere means in general, and being used as such a means through one’s own distressing affects and bodily responses.

Let me now sketch the beginnings of an account that might satisfy this burden. My suggestion is essentially an extension of the Kantian thought that torture fails to respect the dignity of its victim as a rationally self-governing agent. What is distinct about torture, however, is that it does not just traduce the value such dignity represents by treating its subject as a mere means. Rather torture, even in the “best” case, involves a deliberate perversion of that very value, turning our dignity against itself in a way that must be especially offensive to any morality that fundamentally honors it. To see this, we should consider what sort of experience of her own agency the victim must suffer in order for the torturer to succeed. The torturer wants something from his victim, something that the victim would not normally provide (information, confession, pleading, and so on). In response, her captors begin a protracted process of threatening and inflicting pain. Physical pain has a peculiar quality. On the one hand, we experience it as not a part of ourselves: it is something unbidden whose very nature is such that we want to expel it from of ourselves, to abolish it or drive it away. In an obvious way, we are passive before physical pain; it is something that just happens to us, neither immediately evoked nor eliminated by any decisions or judgments we may make.

On the other hand, pain is also a primitive, unmediated aspect of our own agency. Pain is not something wholly alien to our wills, but something in which we find ourselves actively, if reluctantly, participating. My pain is, after all, my pain, an experience that in its very nature seems to demand, wheedle, or plead with me. Insofar as the experience of pain has any content, it seems to be that of a pure imperative. To feel pain is to confront something like a bodily demand to change something about one’s condition, to do something to silence this very demand. Often a pain makes clear just what needs to be done or avoided, but such specificity is not essential to its character as a pain. We can have acute nonspecific pains that we have no idea how to deal with, such as the visceral pains we might feel deep within our bodies after surgery. Yet even here we feel a need to do something in response: we writhe, grimace, or groan. Here pain still seems to be experienced as a kind of imperative, but one that refers only to itself, like someone incessantly screaming “Shut me up!” In this respect, pain seems to be importantly different in form from even physical pleasure. Such pleasure, while an important aspect of the will’s embodiment, does not have the same kind of imperatival quality as pain, the same self-referential demandingness. To feel physical pleasure is to enjoy or luxuriate in one’s state, to be content, satisfied, at rest. To be in pain is to confront some sort of insistent plea to do something; unlike pleasure, its very nature calls for a response from us.

Following Elaine Scarry, we might construe pain as something like the “voice” of the body.²¹ The comparison to language is illuminating. In many respects, pain is like a sensation, but a sensation that seems to have a kind of immediate significance in which the agent already finds his attention and will to be invested. Normally, one cannot adopt a purely contemplative attitude toward one’s own pain. In these respects, feeling pain resembles hearing the utterance of a meaningful sentence. I cannot hear a minimally grammatical English sentence as just noise; I cannot just choose whether or not to grasp its meaning (although I may try with some effort to divert my attention to other things). In hearing the sentence as meaningful, I find myself to be passively receiving an experience in which my active powers of understanding are already in play. If someone tells me to “Go away!” I cannot first contemplate the utterance and then consider how I might respond to it. Rather, in hearing the command, the possibility of going away has already become real for me, as something I must make some effort to disobey or disregard. I may challenge or ignore the order, but this is always a “second move” in response to an initial proposal in which I find myself to have already begun to participate. Similarly, I might ignore or disregard my pain, but to do so I must counter some motivational and evaluative “proto-commitments” that have already begun to engage my will by the time I can actively confront the question of how to respond.²²

Pain resembles a kind of primitive language of bodily commands and pleas that makes the same kind of insistent demands on our attention and response as our children’s shrieking and whining. Understood as a kind of expressive voice, my pain is not unproblematically an exercise of my own agency (the way my reflectively adopted commitments might be), but neither is it something fully distinct from such agency. It is this peculiar duality that the torturer sets out to exploit. 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. The resisting victim is committed to remaining silent, but he now experiences within himself something quite intimate and familiar that speaks for the torturer, something that pleads a case or provides an excuse for giving in.²³ My suffering is experienced as not just something the torturer inflicts on me, but as something I do to myself, as a kind of self-betrayal worked through my body and its feelings. As Scarry observes,

The ceaseless, self-announcing signal of the body in pain, at once so empty and undifferentiated and so full of blaring adversity, contains not only the feeling “my body hurts” but the feeling “my body hurts me”. . . .²⁴

In a similar vein, Susan Brison remarks that after being beaten, raped, and repeatedly strangled into unconsciousness, “My body was now perceived as an enemy, having betrayed my newfound trust and interest in it.”²⁵

It is perhaps not accidental that many of the most common forms of torture involve somehow pitting the victim against himself, making him an active participant in his own abuse. In Abu Ghraib, captives were made to masturbate in front of jeering captors. Here the captive was forced into the position of having to put his most intimate desires, memories, and fantasies into the service of his torturers, in a desperate attempt to arouse himself for their amusement. The U.S. soldiers could beat and kill their prisoners, but only the prisoner himself could offer up his own erotic life to be used against himself in this way. A ubiquitous form of torture is the denial of regular access to toilet facilities. The torture here is not just the infantilizing and dehumanizing disgrace of soiling oneself, but the futile struggle against one’s own body not to do so. The victim confronts the question of whether she was simply forced to soil herself, or whether she allowed herself to do so, discovering herself to be willing to purchase some comfort at the price of public or personal humiliation.²⁶

Torturers often force their victims to stand or maintain contorted postures (“stress positions”) for prolonged periods of time. In these cases, the victim’s own efforts to remain in a particular position serve as the immediate source of his suffering.²⁷ One of the most common forms of contemporary torture is “the submarine,” a technique that involves repeated partial drownings.²⁸ I take it that the torture here is not just the agony of inhaling water, but the hopeless struggle against one’s own desperate urge to breathe that precedes it. Not only does the victim find himself hurt by his body, but he also finds himself to be the one hurting his body as well, in some way pushing it against itself. The relationship of torturer to victim is thus replicated in the victim’s own consciousness of himself as an embodied agent. In the most intimate aspects of his agency, the sufferer is made to experience himself not just as a passive victim, but as an active accomplice in his own debasement.²⁹

When inflicted as part of torture, physical pain is importantly different from other kinds of agony that might result from disease or injury. In the case of natural pain, my body may seem to rebel against me, but it does not seem to rebel out of some other allegiance. The pain of disease may transfix me, but it does not seem to do so for any point of its own. Naturally caused pain is relatively inarticulate, manifesting itself as only the blank negation of my will by something that is nevertheless an intimate part of it. Torture is different, in that the sufferer experiences his pain as having a point. There is a will lurking behind his suffering, a will with a project that is somehow meant to be served by all the various torments. Part of the characteristic dynamic of torture is that the victim is led to hope (however falsely or unreasonably) that there might be something he could do to appease or mollify his tormentor. The victim almost invariably finds himself trying to anticipate his torturer’s will, to figure out what it wants.³⁰ If the victim is in a position to provide what is wanted (say, some specific information), then he will find himself considering (despite himself), whether it might not be so bad to reveal the information, whether the betrayal would be so great, whether he has already endured more than could be asked of him. To head off this sort of dynamic, Jacobo Timerman recommends complete mental passivity, since any sort of active thought served only to aid his torturers:

[I]t’s best to allow yourself to be led meekly toward pain and through pain, rather than to struggle resolutely as if you were a normal human being. The vegetable attitude can save a life. . . . More than once I was brusquely awakened by someone shouting: “Think. Don’t sleep, think.” But I refused to think. . . . To think meant becoming conscious of what was happening to me, imagining what might be happening to my wife and children; to think meant trying to work out how to relieve this situation, how to wedge an opening in my relationship with the jailers. In that solitary universe of the tortured, any attempt to relate to reality was an immense painful effort leading to nothing.³¹

Despite his conscious commitments, the victim experiences within himself a dialectic where some part of him serves as the eager agent of his tormentor. Even if the victim is not in a position to provide the information or denunciations wanted, he still finds a part of himself taking his torturer’s side. The ignorant victim must ask himself how he can best display his ignorance, or offer up some lie or substitute information that might appease his tormentor, often in a feat of great inventiveness.³² In either case, the victim finds in his pain, and his own immediate responses to that pain, a surrogate for the torturer. The victim’s own voice, the voice of his body, has come in part to speak the torturer’s mind. This is why torture requires there to be some protracted process in which pain is both inflicted and withheld (perhaps even assuaged) in a capricious and unpredictable way. This continuous dynamic of inflicting and withholding pain, seemingly at the whim of an omnipotent tormenter, puts the victim in the unavoidable position of betraying or colluding against himself, an experience the victim undergoes whether or not he actually informs or confesses.³³

The situation is even worse in the case of terroristic torture. In the case of terroristic torture, there is nothing in particular that the torturer consistently demands in order to stop the torment. Yet such torture invariably involves relentless questioning, even if it is directed toward information that is obviously valueless or already known to the torturer. These demands are without any evident point, but they also suggest to the victim that something is wanted from him, but something of which he cannot make sense. Irena Martinez, another victim of the Argentine junta, recalls:

Basically, they want you to feel that your life depends on them, that they are omnipotent, and that every nice thing that you do for them will have some influence on their decision about your life or death.³⁴

The victim is forced into a position where she must try to anticipate and understand every little mood and quirk of her torturers. Despite herself, she finds herself trying to grasp her torturer’s interests, anticipate his demands, and present herself in a way that might evoke pity or satisfaction from him.

Consider the climax of Winston Smith’s torments in 1984. When Smith is confronted with the rat-mask, there is nothing that he is called on to say or do to avoid this horror. Rather, it is Smith himself who, in an exercise of considerable ingenuity, has to come up with a self-violation extreme enough to satisfy O’Brien. Smith is undone not merely because his cry of “Do it to Julia!” is sincere; both Smith and Julia fully expected, and accepted, that they would betray each other again and again. Smith is undone because he takes the desperate initiative of putting his intimate knowledge of his own deepest loves and commitments into the service of O’Brien’s torments. Despite himself, Smith ends up doing for O’Brien the one thing that O’Brien could not himself do, and so becomes an active part of the very apparatus of torture that is being applied to him.

The victim of torture finds within herself a surrogate of the torturer, a surrogate who does not merely advance a particular demand for information, denunciation, or confession. Rather, the victim’s whole perspective is given over to that surrogate, to the extent that the only thing that matters to her is pleasing this other person who appears infinitely distant, important, inscrutable, powerful, and free.³⁵ The will of the torturer is thus cast as something like the source of all value in his victim’s world, a unique object of fascination from which the victim cannot hope to free herself. The torturer thereby makes himself into a kind of perverted God and forces his victim into a grotesque parody of love and adoration. Améry writes, “I also have not forgotten that there were moments when I felt a kind of wretched admiration for the agonizing sovereignty they exercised over me. For is not the one who can reduce a person so entirely to a body and a whimpering prey of death God or, at least, a demigod?”³⁶ Like love or religious devotion, such an attitude can develop an emotional hold that persists beyond the circumstances that initially created it, as the phenomenon of “traumatic bonding” or “Stockholm syndrome” attests. Judith Herman writes, “The sense that the perpetrator is still present, even after liberation, signifies a major alternation in the victim’s relational world. The enforced relationship during captivity, which of necessity monopolizes the victim’s attention, becomes part of the victim’s inner life and continues to engross her attention after release.”³⁷ Such attention is dominated by fear, but other forms of fascination and concern may be engendered as well. Conroy recounts several disturbing incidents in which victims, at their own initiative, have become the friends, comrades, and in one case even the lover of their former torturers.³⁸

Torture thus turns out to be something like sexual seduction, accomplished through fear and pain rather than through erotic desire. After all, such desire can be almost as insistent and arresting as physical pain, and as able to “saturate” our bodies as thoroughly as fear or disgust. We can find ourselves to be simultaneously passive before and profoundly engaged by our sexual desires, which we experience as something like both an effect we suffer and an attitude we adopt. In seduction, we find ourselves becoming fixated on the wants, intentions, and body of another (and the physical possibilities of that body), through the nonvoluntary arousal and frustration of our own desires. Successful seduction leads not just to appetite, but to a kind of fascination with another as an embodied agent that makes one acutely aware of one’s own physical exposure before that other. In torture, as in sex, the body is touched in ways that make the most personal and intense feelings manifest themselves publicly and involuntarily (e.g., in erection, lubrication, sweat, shivering, urination, defecation, and centrally, spontaneous cries).³⁹ Of course, even an unwelcome attempt at seduction is not wrong in anything like the way torture is. Normally, the person seduced voluntarily enters into the interaction in question, and is able to voluntarily withdraw from or otherwise reshape the dynamics of the encounter. Just as the seducer works on the desire of his target, she may work on his, or end the interaction at her discretion. We talk about “playing games” in such erotic contexts, a description that never suggests itself in the case of real torture.

Where such possibilities of exit, resistance, and reply are absent, however, torture may well be realized through sexual desire rather than through physical pain. In fact, on the analysis I have offered it should be possible to torture someone through many different sorts of desires whether or not they are sexual in nature. Any suitably intense and relentless craving, whether for food, drugs, sleep, or just quiet, could be the medium through which a suitably constrained and dependent person might be tortured (such torture need not involve touching the victim’s body, so long as his physical environment is appropriately controlled). Our sexual desires and responses are just particularly apt avenues of torture. This aptitude derives not just from the strength and attention-fixing powers of these responses, or the degree to which such feelings are potential sources of shame. In addition, the sexual is an area in which we have one of our basic experiences of ourselves and others as simultaneously free and embodied beings. Sexual interest provides the materials for profound forms of intimacy, self-revelation, and trust: as such, it is readily perverted into a mode of self-estrangement and self-betrayal.⁴⁰ Rape and the threat of rape are common means of torture,⁴¹ as are prolonged nakedness, forcing couples to engage in sex in view of their tormentors, forcing a victim to watch a family member being raped, or forced sexual contact with animals (the forced masturbation of Abu Ghraib seems to be a recent innovation of the U.S. military). In such cases, I suggest, victims are forced to experience the question of whether they are in some way aroused by or welcoming of these violations of themselves and their loved ones,⁴² of whether despite their most sincere commitments they are ultimately in league with their torturer. Victims of both rape and torture often obsess over the question of whether they resisted enough or whether they let themselves “give in” too readily. In both cases, the victim must confront the question of whether her will really manifests her sincerest convictions, or whether it more readily expresses the will of another even as that other displays nothing but hatred or contempt for her.

VI

In “The Genesis of Shame,” J. David Velleman argues that shame is primarily a response to an injury to one’s public standing as a “self-presenting creature.”⁴³ To be able to effectively communicate and cooperate, a rational agent must have the ability, and must be recognized by others as having the ability, to choose which of his feelings, desires, and emotions to present to others. Velleman observes that these abilities are called into question by such “bodily insubordination” as erection or blushing that reveals our feelings despite our best attempts to keep them to ourselves. We feel shame when we seem unable to keep from publicizing what we wish to keep private and hence seem unable to control the persona we present to others. For Velleman, shame is only properly occasioned by one’s own inability to properly maintain one’s privacy, not by the invasion of it, for “[W]hen people forcibly violate your privacy, no doubt is cast on your capacity for self-presentation.”⁴⁴

In torture, the victim suffers a violation that does cast doubt on something similar to his capacity for self-presentation, but deeper and more fundamental. This experience resembles but is worse than the sorts of shaming that Velleman discusses. Insofar as the victim experiences some part of himself to be in collusion with his tormentor, he confronts not just a loss of control over the way he presents himself to others. Rather, doubt is cast on his ability to have cares and commitments that are more immediately and authentically his own than those of another agent. Whatever its ultimate goal, torture aims to make its victim make himself into something that moral philosophy tells us should be impossible: a natural slave, a truly heteronomous will. The victim retains enough freedom and rationality to think of himself as accountable, while he nevertheless finds himself, despite all he can do, to be expressing the will of another, the will of a hated and feared enemy.⁴⁵

Even if the victim does not break, he will still characteristically discover within himself a host of traitorous temptations. His problem is not just that his body is insubordinate, as when an erection reveals his desire in a way completely independent of his will, but that it is treacherous. This treachery is to be found not in the wayward physiological responses of his body, but in those feelings and desires in which he finds his will to be already incipiently invested. Even if the torturers seek only information, they nevertheless try to make their victim experience himself as a moral abomination, as a free and accountable agent whose freedom nevertheless truly belongs to the will of somebody else. The victim finds himself to be not only losing control of his persona. Rather, he also finds himself to be actively giving up control of his person, insofar as his personality is bound up with his ability to immediately define and know his concerns and commitments through his own sincere avowal of them.

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. The violence of war or police action may injure or insult an agent’s capacities for rational and moral self-governance, but such violence need not make the victim an accomplice in his own violation. Torture, in contrast, involves not just the insults and injuries to be found in other kinds of violence, but a wrong that, by exploiting the victim’s own participation, might best be called a humiliation.

VII

So far I have considered torture in contexts in which the victim is completely at the mercy of another person, a person who operates (or at least appears to operate) with no independent constraints on how he may treat his victim. I have argued that in such cases, what morally distinguishes torture from other forms of great violence is in part how the victim is put into a position where she is psychologically and rationally compelled to anticipate and identify with her tormentor’s attitude toward her, thereby being forced to experience her affects, emotions, and imaginings as colluding against her. On this account, torture requires not just the complete helplessness and vulnerability of the victim, but a kind of open-ended freedom (real or perceived) on the part of the torturer.

Much actual torture has this character, especially the kinds that we are most tempted to think might sometimes be morally justified in combating terrorism. But torture need not have precisely this structure, especially when it is used not as a means of extracting information or confessions, but as a form of criminal punishment. A legal system might well exist in which after receiving due process of law, an individual is to be punished by a specific number of lashes or electric shocks of a particular intensity, where the infliction of such punishment is properly monitored by reliable third-party authorities (and perhaps legal and medical representatives of the victim as well). Normally we would call such sort of treatment torture, even though the victim, while helpless, is not (and does not take himself to be) exposed to the caprice of any other agent. Here, the victim is not called upon to answer any questions, or to present himself in some way that might satisfy his torturers. He knows instead that regardless of whatever he says or does, and regardless of what any of his tormentors think or want, he will receive the same punishment, without any possibility of its being eased or aggravated. In medieval legal parlance, such forms of torture were known as ordeals, where there was nothing for the victim to do but suffer in some very fixed and determinate way.

On the analysis I have offered, it may appear that such ordeals do not really qualify as torture; or more precisely, that they are not morally objectionable in the distinctive way other kinds of torture are (although they may be still be highly objectionable in the broader ways that any kind of violence may be). Yet while such ordeals do differ significantly from the central cases of torture I have so far discussed, there is a special objection to such treatment that is similar to the objections to torture “proper.” Even in the case of ordeals, there is a way in which the basic structure of embodied agency is perverted or turned against itself.

What is important here is again not just the experience of pain and fear, but the fact that the pain is administered by another person, that it serves some purpose or point of another, and that the victim cannot effectively evade, retaliate, or shield himself against these assaults. Unlike other kinds of attack, here the victim must simply take it: there is no reply or counter open to him, nothing for him to do in response. In normal action, a person’s movements and feelings are expressive of his prior attitudes, desires, or intentions, which then have their significance recognized by others. But in ordeals, this expressive relation between the public and the private is reversed. Another person or institution has some interest, project, or attitude toward the victim, realized immediately in the body and bodily affects of the victim, who experiences himself as merely the ultimate passive recipient of these acts. His thoughts, his attention, his interest assume whatever form his body demands. In a sense, his body ceases to be his, to be the substance in which he expresses his own attitudes, intentions, and feelings in a way that can be meaningful for others as a form of self-expression. Since the victim cannot effectively reassert himself physically against the assault (by fighting, fleeing, or shielding himself), his body becomes the medium in which someone else realizes or expresses his agency. The victim here becomes little more than a point of pure receptivity, having the most basic forms of agency effectively at the command of another.⁴⁶ Améry observes:

The other person, opposite whom I exist physically in the world and with whom I can exist only as long as he does not touch my skin surface as border, forces his own corporeality on me with the first blow. . . . Certainly, if there is even a minimal prospect of successful resistance, a mechanism is set in motion that enables me to rectify the border violation by the other person. For my part, I can expand in urgent self-defense, objectify my own corporeality, restore the trust in my continued existence. . . .⁴⁷

In such torture by ordeal, the victim is not actively colluding against himself: rather, it is as if basic conditions of his agency have been completely assimilated by another, having become so thoroughly his enemy’s that even the idea of betrayal is out of place. Here the victim experiences his body in all its intimacy as the expressive medium of another will,⁴⁸ a will to which what is left of his personality finds itself immediately conforming. In the ordeal the victim’s will is not annihilated, as in death, but turned into just a locus of suffering, as something that is aware of itself as a body available to and saturated by the active will of another.⁴⁹ The experience has been likened to a kind of paradoxical consciousness of oneself as dead.⁵⁰ Améry writes that

[a]mazed, the tortured person experienced that in this world there can be the other as absolute sovereign. . . . Astonishment at the existence of the other, as he boundlessly asserts himself through torture, and astonishment at what one can become oneself: flesh and death. The tortured person never ceases to be amazed that all those things one may, according to inclination, call his soul, or his mind, or his consciousness, or his identity, are destroyed when there is that cracking and splintering in the shoulder joints. That life is fragile is a truism he has always known. . . . But only through torture did he learn that a living person can be transformed so thoroughly into flesh and by that, while still alive, be partly made into a prey of death.⁵¹

Like interrogational torture, ordeals involve not just an insult or injury to the victim’s agency. Through the combination of captivity, restraint, and pain, the physical and social bases of rational agency are actively turned against such agency itself. In torture, Améry writes, “one’s fellow man was experienced as the antiman.”⁵² If so, then torture by ordeal should be objectionable in ways akin to torture that seeks some sort of response from its victim, insofar as both involve some sort of perversion of the most basic human relations. Whether such objections could ever be overcome by legitimate military or punitive interests is a question that waits upon more comprehensive understandings of the morality of punishment, warfare, and self-defense.

Earlier versions of this article were presented at the University of Toronto, the University Center for Ethics and the Professions, Harvard University, and the Philamore Ethics group. I am grateful to Kate Abramson, Arthur Applbaum, Marcia Baron, Elizabeth Harman, Simon Keller, Christine Korsgaard, Rahul Kumar, Margaret Little, Dick Moran, Henry Richardson, Arthur Ripstein and Dennis Thompson for helpful comments and suggestions. I am also grateful for fellowship support from the University Center for Ethics and the Professions, Harvard University, during the preparation of this article.

See Edward Peters, Torture (New York: Basil Blackwell, 1985), p. 154.

Donald Rumsfeld insists that what the prisoners at Abu Ghraib suffered was not “technically” torture but only “abuse,” and the European Court of Human Rights has concluded that while Great Britain had subjected suspected IRA sympathizers to “inhuman and degrading treatment,” what they suffered was nevertheless “not quite torture.” In neither case, however, was it made clear just what the relevant distinction was supposed to be (see n.15 below).

“There can never be any justification for torture. It creates an escalation of violence in the internal affairs of states. It spreads like a contagious disease from country to country. It has lasting effects on the mental and physical health of the victim, and brutalizes the torturer” (Declaration of the First Amnesty International Conference for the Abolition of Torture, as quoted in A Glimpse of Hell, ed. Duncan Forrest [London: Amnesty International UK, 1996], p. viii). Of course, all the same objections might be pressed against warfare in almost any form, without making war categorically unjustifiable.

See, e.g., Alan Dershowitz, “Torture of Terrorists” in Shouting Fire (Boston: Little Brown, 2002), pp. 470–77; Bruce Hoffman, “A Nasty Business,”The Atlantic Monthly (January 2002); and

Although many in the United States do seem open to torture as a form of entertainment, as attested by the popularity of “reality television” devoted to little more than the infliction of various humiliations and cruelties upon its (sometimes unwitting) participants. To cite one notable example: an episode of “Culture Shock” offered a prize to whichever of two contestants could endure being suspended in the back-bending “harness of pain” longer. At least in one case, it was a contestant’s husband who was called upon to hold the rope that contorted her body. The contestant claims to have been permanently injured by the experience and has brought suit against the producers of “Culture Shock.” See Adam Liptak, “Growing Rowdier, TV Reality Shows Are Attracting Suits,”The New York Times (January 7, 2003), Sec. A, p. 1.

See Richard Bernstein, “Kidnapping Has Germans Debating Police Torture,”The New York Times (April 10, 2003), Sec. A, p. 3.

Even those who morally condemn capital punishment often consider it to be distinct from and less objectionable than punitive torture, as Amnesty International discovered when it decided to group capital punishment with torture and mutilation as the sort of human rights violation that Amnesty International would protest.

That is, than other forms of great violence that we sometimes countenance in war and police action. I do not mean to say that torture is absolutely unique morally: its distinctive wrongs may well also be found in rape and many kinds of spousal and child abuse. I suspect that in a comprehensive analysis, these latter would best be understood as special types of torture.

Which is not to say that when torture is wrong, it is worse that wrongful acts of maiming and killing. This I leave as an open question, to which there may not be any very general answer. My interest here is only with the prima facie objections to these acts, and what kinds of concerns may or may not properly counter such objections. Nothing I argue here entails 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.

Office of the High Commissioner for Human Rights, http://www.unhchr.ch/html/menu3/b/h_cat39.htm. The Convention entered into force in 1987.

See Metin Basoglu and Susan Mineka, “The Role of Uncontrollable and Unpredictable Stress in Post-Traumatic Stress Responses in Torture Survivors,” in Basoglu, pp. 182–217.

In some consensual sado-masochistic acts, the “victim” may similarly occupy a position of complete vulnerability before another. Yet here the victim need not be entirely helpless before his tormentor, should there be agreed upon signals or “stop words” that he can use to call the session to a halt. Such acts may differ from real torture in that they are (at least sometimes) conducted with a shared understanding that the victim retains an effective power to withdraw consent, and hence that his subjection is merely a kind of pantomime (however real the pain). Of course, in such cases the victim must trust that his tormentor will respect that authority, but unlike real cases of torture, that trust need not be unreasonable. However, such pretend-torture may turn into real torture when the tormentor makes it clear that he does not recognize the victim’s right to opt out, taking himself to be as unconstrained morally as he is physically.

In Northern Ireland, Great Britain employed what it called “interrogation in depth” against suspected IRA sympathizers. Such interrogation consisted in hooding the prisoners, keeping them in a room pervaded by the din of a large engine or fan, depriving them of food, water and sleep, and forcing them to stand for long periods pressing themselves against a wall. (Such techniques are also part of current U.S. interrogation procedures; see Mark Danner, “The Logic of Torture” in The New York Review of Books [June 24, 2004], pp. 71–72). In the British case, the European Court of Human Rights concluded that these five techniques, while constituting “inhuman and degrading treatment,” were still “not quite torture.” See John Conroy, Unspeakable Acts, Ordinary People: The Dynamics of Torture (Berkeley: University of California Press, 2000), p. 187. From what I have argued, this judgment might be defended by appeal to the fact that the isolation and sensory disorientation were only (very successful) attempts to undermine the victim’s psychological integrity. However, hunger, thirst, and the prolonged standing were intensely painful, and conducted in an atmosphere in which the prisoners continually feared for their lives and health. Such techniques are clearly forms of torture as I understand it.

For an interpretation along these lines, see William Twining, “Torture and Philosophy,”The Proceedings of the Aristotelian Society (1978, supp. vol. 52): 147.

See, e.g., Barrie Paskins, “Torture and Philosophy,”The Proceedings of the Aristotelian Society (1978, supp. vol. 52): 165–94.

See Glenn R. Randall and Ellen L. Lutz, Serving Survivors of Torture (Waldorf, Maryland: AAAS Books, 1991), pp. 28–30.

Insofar as torture involves an arbitrary or capricious use of power (or the appearance of such a use), it tends to erode any legal strictures imposed upon it. The Romans initially limited interrogational torture to slaves, and when Europe reintroduced torture in the thirteenth century it took care to exempt “children, the elderly, pregnant women, knights, barons, aristocrats, kings, professors, and, usually, the clergy.” In both cases, however, these restrictions were gradually relaxed until almost anyone could be tortured pursuant to a judicial proceeding. See Conroy, p. 30. Modern attempts to introduce torture as a limited and temporary emergency measure have generally met the same fate.

Randall and Lutz deny that there is any unique “torture syndrome.” Instead, they argue, the psychological consequences of torture are similar to those suffered by other survivors of great stress and trauma, including combat and imprisonment. Randall and Lutz, pp. 4–5.

Henry Shue, “Torture,”Philosophy & Public Affairs 7 (1978): 124–43.

Scarry, pp. 45–51.

For a related discussion to which I am considerably indebted, see Christine Korsgaard, The Sources of Normativity (Cambridge: Cambridge University Press, 1996), pp. 138–40, 145–60.

“There were times when he feebly tried to compromise, when he said to himself: I will confess, but not yet. I must hold out till the pain becomes unbearable. Three more kicks, two more kicks, and then I will tell them what they want.” George Orwell, 1984 (New York: Plume Books, 1983), p. 215.

Scarry, p. 47.

Susan J. Brison, Aftermath: Violence and the Remaking of a Self (Princeton, N.J.: Princeton University Press, 2002), p. 44.

Torturers also frequently heighten this experience of collusion and betrayal by making victims witness or otherwise come to believe that their loved ones are being tortured, with the suggestion that this torture will cease if the victim complies. See Randall and Lutz, pp. 109–10; Basoglu and Mineka, pp. 205–06. Here, the victim is forced into a kind of complicity with his torturer on two levels. Overtly, his actions determine whether his family or his comrades (or whoever else he might incriminate) are tortured, or the degree or manner in which they suffer. More deeply, the victim’s feelings of love for his family have been mobilized by the torturer, and he has been made to regret the fact that he cares for them. The victim’s love has been made to condemn itself (a result that is also to be found in hostage-taking and many kinds of terrorism). My thanks to Christine Korsgaard for pressing me on this point.

Such as the hooded prisoner in the notorious Abu Ghraib photograph, balancing on a box and holding his wired arms apart in fear of being electrocuted if he fell.

The U.S. government admits to subjecting Khalid Sheik Mohammed to this technique, in a variation given the more sporting name of “water-boarding.”

Conroy, p. 169. Of course, another reason why such techniques are so common is that they are cheap and easy to perform and tend to leave no incriminating marks on their victims.

Brison tells of “the heightened lucidity that had led me to memorize my assailant’s face during the attack, when my life had depended on reading every gesture, hearing every noise, taking everything down, storing it all away” (Brison, p. 109). See also Jacobo Timerman, Prisoner without a Name, Cell without a Number (New York: Knopf, 1981), p. 35.

Timerman, p. 35.

“I talked. I accused myself of invented absurd political crimes, and even now I don’t know at all how they could have occurred to me, dangling bundle that I was” (Améry, p. 133).

It is not surprising that victims of torture typically suffer a kind of profound alienation from their own affective and emotional lives. Torture typically produces alexithymia, “the inability to be aware of and to tolerate basic feeling states” (Henry Krystal, “The Paradigm of Adult Catastrophic Trauma and Infantile Trauma,” as quoted in Randall and Lutz, p. 42). Someone suffering alexithymia “cannot use [his] emotions as guidelines for action, is unable to tolerate feelings that do occur and cannot recognize these feelings, has a decreased ability to conceive abstract thoughts, and has a decreased ability to feel pleasure” (Randall and Lutz, p. 42).

Conroy, p. 171.

Cf. Elie Wiesel’s account of witnessing his father being beaten by a concentration camp guard: “I had watched the whole scene without moving. I kept quiet. In fact I was thinking of how to get farther away so that I would not be hit myself. What is more, any anger I felt at that moment was directed, not at the [guard], but against my father. I was angry with him, for not knowing how to avoid Idek’s outbreak. That is what concentration camp life had made of me” (Night, as quoted in Judith Herman, Trauma and Recovery [New York: Basic Books, 1991], p. 84).

Améry, p. 133. Timerman quotes one of his tormentors remarking, “Only God gives and takes life. But God is busy elsewhere, and we’re the ones who must undertake this task in Argentina” (Timerman, p. 31).

Herman, p. 91.

Conroy, p. 175.

“De Sade claimed that the object of sexual desire was to evoke involuntary responses from one’s partner, especially audible ones. The infliction of pain is no doubt the most efficient way to accomplish this, but it requires a certain abrogation of one’s own exposed spontaneity” (Thomas Nagel, “Sexual Perversion” in Mortal Questions [Cambridge: Cambridge University Press, 1979], p. 50).

Scarry similarly considers torture and sex both to involve some profound experience of self-revelation, although in torture such self-revelation serves only to destroy the possibilities of human intimacy. Here pain and fear are used to bring “all the solitude of absolute privacy with none of its safety, all the self-exposure of the utterly public with none of its possibility for camaraderie or shared experience” (Scarry, p. 53).

See Inge Lunde and Jorgen Ortmann, “Sexual Torture and the Treatment of its Consequences” in Basoglu, pp. 310–11.

See Lunde and Ortmann, p. 314.

J. David Velleman, “The Genesis of Shame,”Philosophy & Public Affairs 30, (2001): 37.

Velleman, p. 38.

Améry writes of the torturer “expand[ing] into the body of his fellow man” (p. 132). Cf. Brison, “This working through, or remastering of, the traumatic memory involves going from being the medium of someone else’s (the torturer’s) speech to being the subject of one’s own” (Brison, p. 56).

Améry, p. 126: “The boundaries of my body are also the boundaries of my self. My skin surface shields me against the external world. If I am to have trust, I must feel on it only what I want to feel.”

Améry, p. 126.

This aspect of punitive torture may be shared to varying degrees with other problematic forms of corporal punishment such as mutilation, castration, and death (even when such punishments can be inflicted without causing physical pain). Cf. Jeffrey H. Reiman, “Justice, Civilization, and the Death Penalty: Answering van den Haag,”Philosophy & Public Affairs 14 (1985): 115–48. Reiman argues that one of the principal objections to all forms of corporal punishment is the “urgency” and all-absorbing nature of physical pain or, in the case of execution, of the fear of death. However, it is unclear why the fear of a (painless) death is different in kind than the fear of a lengthy prison sentence, or the “reflective pain” of serving one out. Nor does Reiman consider potentially painless forms of corporal punishment, such as mutilation or branding, that seem objectionable just as immediate assaults on the body.

Cf. Herman, p. 93: “Whatever new identity she develops in freedom must include the memory of her enslaved self. Her image of her body must include a body that can be controlled and violated. Her image of herself in relation to others must include a person who can lose and be lost to others.”

See Brison, pp. 38–66.

Améry, p. 136, my emphasis.

Ibid., p. 136.