Torture has torn through additional restraints since I first tried to get a grip on it (Shue 1978). Then I began by apologizing for raising an issue that I thought most Americans considered closed. The worst torturers seemed to be in such places as the Shah’s Iran, Marcos’s Philippines, South Korea and Latin American dictatorships, including Chile, Argentina, Brazil and Guatemala.[i] Courageous US Congress members like Don Fraser of Minnesota and Vietnam war veteran Tom Harkin of Iowa led an outcry against torture in the early 1970s, a protest that provoked my own interest -and was later embraced by the presidency of Jimmy Carter after 1976. Many of us thought that” torture was practised in the Third World but condemned by Americans and Europeans, reminiscent of the author of the 1911 entry on “Torture” for the Encyclopaedia Britannica, who claimed “[t]he whole subject is now one of only historical interest as far as Europe is concerned” (Waldron 2005b: 1683-84; 2010: 187). How far, and how sadly, wrong can one be!

It emerged that a number of the Latin American torturers had been trained at Fort Benning, Georgia, in the School of the Americas in techniques developed by the Central Intelligence Agency (CIA) (Gill 2004; McCoy 2006: 60-107; Rejali 2007). Unknown to me and most other US citizens whose taxes were funding it in 1978, CIA had already for decades been funding a massive academic research programme designed to create a new paradigm of torture, which is now in operation and has to a considerable degree corrupted the US military (Wol­fendale 2007, Forsythe 2011). The most penetrating account of this secretly funded research is Wisconsin historian Alfred W. McCoy’s A Question of Torture: CIA Interrogation, front the Cold war to the War on Terror: “From 1950 to 1962, the CIA became involved in torture through a massive mind-control effort, with psychological warfare and secret research into human con­sciousness that reached a cost of a billion dollars annually – a veritable Manhattan Project of the mind” (McCoy 2006: 7).

In April 1953 the CIA organized its various research grants “under the umbrella of a unified project, MKUltra”: (ibid.: 28). Much of the research on human consciousness involved human experimentation:

“[f]rom 1953 to 1963, MKUltra and allied projects dispensed $25 million for human experiments by 185 nongovernmental researchers at eighty institutions, including forty-four universities and twelve hospitals” (ibid.: 29). Direct and indirect CIA contracts went to universities including, for instance, McGill, where pioneering research had been done by the great psychologist Donald 0. Hebb (ibid.: 32); also Yale, with Irving L. Janis and Stanley Milgram (ibid.: 32-33); and Cornell Medical Center’s Lawrence Hinkle and Harold Wolff (ibid.: 41, 45).[ii] Further research was done during the war in Vietnam as part of the Phoenix programme (and others) where, in McCoy’s ironic words, there was “a limitless supply of human subjects” (ibid.: 65).

In 1963, the CIA distilled its findings in its seminal Kubark Counterintelligence Interroga­tion handbook. For the next forty years, the Kubark manual would define the agency’s interrogation methods and training programs throughout the Third World. Synthe­sizing the behavioral research done by contract academics, the manual spelled out a revolutionary two-phase form of torture that relied on sensory deprivation and “self-inflicted” pain for an effect that, for the first time in the two millenia of this cruel science, was more psychological than physical. (Ibid.: 50)[iii] 

The “Manhattan Project of the mind” paid off for the CIA with “the first real revolution in the cruel science of pain in more than three centuries” (ibid.: 8). McCoy thinks the revolutionary method is “best described as ‘no-touch torture” (ibid.: 7); 1 will refer to it as the CIA paradigm of psychological torture or, sometimes, the American way of torture, funded as it was by receipts from US taxpayers provided to CIA by a somnolent Congress. As McCoy indicated, the CIA paradigm of psychological torture has two central elements: sensory deprivation and “self-inflicted” pain.

First, “sensory deprivation became the conceptual core of the agency’s paradigm” (ibid.: 32). What to an ordinary layperson are the amazingly powerful effects of sensory deprivation were summarized in the Kubark Manual, relying on human experimentation done at Harvard, as follows:

Drs. Wexler, Mendelson, Leiderman, and Solomon conducted a somewhat similar experiment on seventeen paid volunteers … The results confirmed earlier findings that (1) the deprivation of sensory stimuli induces stress; (2) the stress becomes unbearable for most subjects; (3) the subject has a growing need for physical and social stimuli; and (4) some subjects progressively lose touch with reality, focus inwardly, and pro­duce delusions, hallucinations, and other pathological effects. (CIA 1963: 89) 

Kubark goes on to quote another researcher as having said: “Three studies suggest that the more well-adjusted or ‘normal’ the subject is, the more he is affected by deprivation of sensory stimuli” (ibid.). And this section of the manual concludes: 

The deprivation of stimuli induces regression by depriving the subject’s mind of contact with an outer world and thus forcing it in upon itself. At the same time, the calculated provision of stimuli during interrogation tends to make the regressed subject view the interrogator as a father-figure. The result, normally, is a strengthening of the subject’s tendencies toward compliance.(Ibid.: 90)

The CIA’s hope that sensory deprivation will produce regression to the state of a compliant infant appears frequently in the manual: “all coercive techniques are designed to induce regression” (ibid.: 83).[iv]

“Self-induced” pain is of course not actually self-induced, but the interrogators contrive to make the person being tortured feel that he is causing his own pain, thereby greatly enhancing the unbearability of the stress mentioned above by the Harvard experimenters.[v] For example, a victim is told to stand up against a wall in a certain specific position and not move, lest something else very distressing be done to her. After a relatively short period, this rigid stance becomes quite painful. On the one hand, the subject is likely to feel that the pain is her own fault because she is obeying the order not to move. But if, on the other hand, she moves and the other distressing thing is done to her, she is also likely to feel that that distress is her own fault because she disobeyed the order not to move. Obviously she cannot win but is in fact likely nevertheless to feel that she is responsible for whichever bad outcome she suffers, although of course the torturers are responsible for having created the inescapable Catch-22. According to McCoy, the discovery of how profoundly stressful “self-inflicted” pain is “emerged from a series of studies by two respected neurologists at Cornell Medical Center, Lawrence Hinkle and Harold Wolff, who founded the Human Ecology Society in 1953-54, soon making it the most important CIA conduit for legitimate scientific research” (McCoy 2006: 45).

Secretary of Defense Donald Rumsfeld demonstrated the depth of his ignorance about the very CIA techniques he was authorizing for use by the Bush Administration on its prisoners at Guantanamo and elsewhere with his notorious mocking marginal comment on one of the torture-authorizing memos from the Pentagon’s chief lawyer, Waal-11J. Haynes, II: “However, I stand for 8-10 hours a day. Why is standing limited to 4 hours? D. R.” (Haynes 2002: 236 [photocopy] and 237 [transcription]). Presumably Rumsfeld shifts position at will when he chooses to stand. Here is one experienced insider’s interpretation of the Rumsfeld scribble: “It said, ‘Carte blanche, guys’. That’s what started them down the slope. You’ll have My Lais then” (Mayer 2008: 223).

If discussion of torture is to have any semblance of a connection with torture as actually practised, it is essential to remember at least that torture is not some rare and exceptional onetime ad hoc emergency measure, but a carefully researched and tested practice adopted as a matter of policy for a general category of suspects whose psychological equilibrium is if possible, shattered (Constitution Project 2013, Singh 2013). As indicated by the subject line of the above-mentioned Haynes memorandum that Rumsfeld annotated, torture is conceived by its proponents as consisting of “counter-resistance techniques”, where the resistance to be overcome is constituted by what one would normally consider to be a person’s identity, self or personality, embodying her commitments and values. One’s identity as a person is the “resistance” that must be broken! You are the resistance. Whether, once broken, you will ever mend is an open question. Space does not permit also taking up the damaging effects on torturers here, but we have good reason to think that “the toxic dividends of torture are shared by victims and victimizers” (Phillips 2010: xi).


When one reflects on what torture typically consists of specially trained government agents 11,1cling a person captive and assaulting the structure of his personality by the most effective means known to modern psychology — one is strongly inclined to think that torture ought indeed to be absolutely prohibited, precisely as it has been by international law (Waldron 2005b: 1688-91; 2010: 191-94). The first step in the contemporary justification for torture is to attack the absoluteness of the moral and legal prohibitions with “a single mesmerizing example: the ticking time bomb” (Luban 2005: 1440; 2006: 44). The ticking-bomb scenario has appeared in many guises, at least as far back as Bentham’s frequently quoted bomb-less version (Ginbar 2010: 379-86); for the origins of the modern “imaginary of torture” with the ticking bomb, Rejali emphasizes novels and films about the Franco-Algerian War, especially fiction by Jean Larteguy (Rejali 2007: 545-46), commenting “too often fantasy sells better than reality” (ibid.: 547).

The gist of the scenario is that “we” manage to capture someone who is known to have planted in some unknown location a ticking bomb that will surely cause a catastrophe if not defused, who refuses to disclose the location, but whom “we” quickly and successfully torture, only as much as necessary and in time for the information gained to enable us to keep the bomb from exploding. Different variants deal with the various implausibilities here in alternative ways with various degrees of success, and toying with different versions seems to provide endless fascination for analytic philosophers who think hypothetical examples can teach us more than we can learn from studying what actually happens and what is likely to happen (Shue 2009).

David Luban analyses the general function of the reasoning focused on ticking bombs, whatever the details of the hypotheticals. Someone who begins by believing that torture ought to be subject to the kind of absolute prohibition found in international law is made to confront what is presented as an undoubted case of torture that is, all things considered, justified. Yes, she is told, torturing anyone is not just wrong but terribly wrong; however, the preventable catastrophe would be an even more terrible wrong, and so – surely! – it would be irrational to allow a much more terrible wrong to occur rather than being willing to commit a less terrible one oneself in order to bring about a rescue by torture. The defender of the absolute prohibition may grant the exception and thus qualify the absolutism of her original position. “Now that the prohibitionist has admitted that her moral principles can be breached,” Luban observes, “all that is left is haggling about the price” (Luban 2005: 1440; 2006: 44): that is, haggling about how catastrophic the preventable catastrophe needs to be in order to make the torture the better choice, all things considered.

All that one really needs to see, on the contrary, is that the ticking-bomb scenario is not an example of torture as it is practised on planet Earth. First, it might be that one can imagine a possible world in which, although torture is almost never used, it somehow can virtually instantly be employed successfully to deal exclusively with rare emergencies. But, in Luban’s words, “the real world is a world of policies, guidelines, and directives. It is a world of practices, not of ad hoc emergency measures” (Luban 2005: 1445; 2006: 47). The scenario is at best irrelevant in focusing on one-off or rare instances rather than on routine bureaucratic policies and general directives.


in a world of uncertainty and imperfect knowledge, the ticking-bomb scenario should not form the point of reference … The real debate is not between one guilty man’s pain and hundreds of innocent lives. It is the debate between the certainty of anguish and the mere possibility of learning something vital and saving lives. (Luba,: 2005: 1444; 2006: 46-47)

The scenario is positively misleading in transferring intuitions about certain success to cases of possible success. So, as Bob Brecher has put it in one of the most thorough and persuasive critiques of the use of ticking-bomb scenarios:

your position therefore has to be that torture is justified by even the possibility of catastrophe – not [only] by its certainty … [What we are actually being invited to accept is that interrogational torture is morally justifiable because it might – and, if my arguments so far are right, only_just might – avoid the catastrophe. (Brecher 2007: 29, 35)

Third, once mere possibility seems to be enough, torture easily becomes a “general fishing expedition”, merely hoping to turn up something useful sooner or later (Luban 2005: 1443; 2006: 46):

The torture at Abu Ghraib had nothing to do with “ticking bomb” terrorism. It was intended to “soften up” detainees so that U.S. military intelligence could get infor­mation from them about likely attacks by Iraqi insurgents against American occupiers. The important point is that the use of torture is not an area in which human motives are trustworthy. (Waldron 2005b: 1717; 2010: 221)

But for those eager to believe in short cuts as a substitute for patient investigative work, the scenario may now seem to provide a rationale for practices that are much more different from it than similar to it.

“Ticking-bomb stories are built on a set of assumptions that amount to intellectual fraud Ticking-bomb stories depict torture as an emergency exception, but use intuitions based on the exceptional case to justify institutionalized practice and procedures of torture” (Luban 2005: 1427; 2006: 36). That these ticking-bomb scenarios have harmful effects is not speculation. The Schle­singer Report, provoked by Abu Ghraib, noted: “[f]or the U.S., most cases for permitting harsh treatment of detainees on moral grounds begin with variants of the ‘ticking time bomb’ scenario” (Independent Panel to Review Department of Defense Detention Operations 2004: 974). One repeatedly finds politicians, journalists, and talk-show participants moving from premises based on the ticking-bomb scenario to conclusions about, for example, treatment of prisoners at the U.S. prison in Guantanamo, even though the cases are worlds apart (Sands 2008: 117, 135, 158, 170, 191, 200-210). One cannot entirely prevent others from misinterpreting what one writes, but in an area like torture one should take every possible measure to discourage it.

Empirical work on state torture bureaucracies provides explanations of why the practice of torture has the features it has (Bufacchi & Arrigo 2006), not the features built into ticking-bomb scenarios. Fora empirical reasons — psychological, political, sociological and bureaucratic — the carefiilly limited use of torture is impossible: “history does not present us with a government that used torture selectively and judiciously … One can imagine rare torture, but one cannot institutionalize rare torture … It is an optimistic thought with no social embodiment” (Shue 2006: 234, 238). In Brecher’s words, “the institutionalization of the profession of torturer is a necessary condition of the example’s even getting off the ground” (Brecher 2007: 24).

However, a hard-nosed consequentialist willing to bite any necessary bullet can respond by abandoning such fantasies of isolated instances of successful torture in rare emergencies and embracing torture bureaucracies for “permanent” emergencies. “The general point is simple: any finite costs to torture can be outweighed by sufficient expected benefits. The worse the anticipated evil, the inure horrible the things we can do to ward it off’ (Luban 2009: 201). Now, one might well judge that if this is the direction in which consequentialism is taking us, it i., nine to ball out. In a recent essay that breaks new ground, David Luban suggests two different ways to think about such an exit, one turning on the limits of moral theories in general and one turning on a category of the morally “unthinkable” (ibid.: 202-6).

In my 1978 essay “Torture” I had proposed the methodological principle that “artificial cases make had ethics” (Shoe 1978: 141). Rightly observing that this dictum was left rather enig­matic, Luban notes that in so far as I elaborated it at all, the point seemed to be that it is falla­cious to reason from extraordinary cases to ordinary cases with different features from the extraordinary ones (Luban 2009: 196-98). We both still think that this is an important point. It has been the guiding principle here so far and is the reason I have emphasized that real-world torture is done by entrenched (secret) professional bureaucracies with a far lower threshold for using it than any almost certain, imminent catastrophe. Luban formulates this first reading of my methodological principle as:

By focusing on improbable artificial cases, theorists misdirect readers’ attention from genuine issues in the real world to specious issues. They illicitly change the subject from important and authentic questions about the limits of legitimate interrogation in non-TBS [ticking-bomb scenario] cases to intuition-mongering about a tendentious hypothetical. (Ibid.: 197)

I agree that this still matters very much.

However, employing extremely generous principles of interpretation, Luban also finds the seeds of three other readings buried in “artificial cases make bad ethics”. Ignoring the second reading here, I shall briefly consider the other two closely related readings, which probe considerably deeper than I had done at the time and which draw on subsequent insights by Bernard Williams. The next reading to which we turn says:

Ordinary practices of moral rationality fail in cases where all courses of action are monstrous. The artificial cases ethicists cook up to control for monstrosity by isolating the right- and wrong-making characteristics of action are misleading. That is precisely because they cover over the monstrousness with a veneer of rationality. (Ibid.: 203)

Luban is thinking of a philosopher’s conundrum like “would you be justified in torturing one person in order to find out a secret location where ten other persons were being tortured so that you could rescue them? How could you not?” But Williams had proposed: “there are certain situations so monstrous that the idea that the processes of moral rationality could yield an answer in them is insane”, and “to spend time thinking about what one would decide if one were in such a situation is also insane, if not frivolous” (Williams 1973: 92).

Luban offers an explanation of why moral rationality might be limited in the way Williams suggested:

Moral systems … arise by generalizing and abstracting from prototypical cases in which they make intuitive sense and yield intuitively satisfying answers … But completeness claims are illusory This should not surprise us: the origin of moral systems suggests that they are good only over certain domains, those in the neighborhoods of their prototypical cases.(Luban, 2009: 203, 204)

If one finds it disgusting even to consider torturing a person in order possibly to rescue ten others from torture, but one is told that it would be morally irrational not to do so, one has no particular reason to trust the judgement about moral rationality to be one’s guide rather than trusting one’s strong feeling of disgust. This bizarre case may well fall completely outside the range of cases in which the moral theory is a reliable guide.

Williams had also suggested that “the unthinkable was itself a moral category” (Williams 1973: 92). Insightful analysts have suggested that the other assault to which torture is most similar is rape (Sussman 2005: 4-5). A column in the New York Times had said:

A clear sign of progress in Western society is that one does not need to argue against rape: it is “dogmatically” clear to everyone that rape is wrong. If someone were to advocate the legitimacy of rape, he would appear so ridiculous as to disqualify himself from any further consideration. (Zizek 2007)

Luban comments:

Suppose that the only way Jack Bauer [of Fox’s torture-romanticizing 24] can prevent ten women from being raped is to rape one woman, You will never see that plot-line on television, for obvious reasons: the audience, which is meant to root for Jack Bauer [when he tortures], would find Jack the rapist viscerally revolting. That’s the mark of the unthinkable. Conversely, if we insist on arguing the costs and benefits of rape with an unblinking accountant’s eye, as if it is just one option among others, we run the risk of normalizing it, moving it out of the category of the unthinkable. (Luban 2009: 205)

So, Luban proposes, a final reading of my principle is: “Artificial cases make bad ethics because their very artificiality makes the unthinkable thinkable” (ibid.: 206).

Luban’s two points, about limits of moral rationality and about unthinkability, are closely connected, if not two facets of a single one. If our moral theories generalize from cases where they can be brought into equilibrium with our intuitions, then when they confront a case with radically different features from the original cases and suggest a response that seems intuitively wrong, we may do better to trust our intuition instead of the response that follows from the theory. if the action seems disgusting or revolting and the only reason to do it is that a permission flows out of a moral theory, we may simply have crossed outside the range of reliability of that theory. Perhaps we could construct a better theory that adequately accounts for the horror that torture instinctively arouses. Meanwhile, perhaps we should “stop thinking” in the following very specific sense: avoid following the guidance of a theory the reliability of which is dubious in the range that includes the case at hand. We have good intuitive grounds in our revulsion and disgust to believe that a moral theory that suggests one rape to prevent ten, or one torture to prevent ten, has blundered beyond its competence. Leave the rape and the torture morally “unthinkable” until such time as we have a moral theory that we can be confident is robust enough reliably to guide any thinking we do in this disorienting territory.



Space here does not permit discussing the features of the contemporary American way of torture, which are the empirical basis for the arguments, in any detail. David Luban and I elaborated a number of features in the course of showing that McCoy’s “no-touch torture” can easily slip through the crude net of badly drafted US statutes that purport to prohibit torture, revealing a desperately urgent need for re-writing US law on torture if anyone is to be protected against the new American way (Luban & Shue 2012). We emphasized the mercilessness of all torture:

part of the special wrongfulness of torture lies, then, in the limitlessness of the extent to which the victim is at the mercy of the torturer, who never relents until he himself, for his own reasons, chooses to end the terror which this implants in the victim. The victim can attempt to end the torture by trying to give the torturer what the victim thinks the torturer wants, but the torturer decides entirely for himself what he wants at any given time and whether he believes he has it all. The victim may well guess wrong about what the torturer wants, and often the victim does not have what the torturer wants in any case, especially of course if the victim is not who the torturer thinks he is or has not done what the torturer suspects he has done. All power remains with the torturer, who may move the goal posts as often and as far as he wishes. The victim is utterly at his mercy. Unlike even war, torture has no natural end. It ends when the torturer chooses to end it. (Ibid.: 859)

Psychological torture in accord with the CIA paradigm, in particular, undermines the structure of the personality — it literally breaks apart the self, unhinging its parts from each other. The victim is reduced to a quivering bundle of fears, driven to try to please, that is, to try to fulfil the wishes of others, with few wishes of her own, except release from the awful psychological stresses that are being systematically and relentlessly imposed by all-powerful others. This goes far beyond what slavery involved and gives new meaning to being at the mercy of someone else. (Ibid.: 856)

Writing decades later about the psychological effects of his crude old-fashioned physical torture by the SS, much less devilishly fine-tuned than the state-of-the-art CIA paradigm, Jean Amery wrote:

Whoever was tortured, stays tortured   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 … Whoever has succumbed to torture can no longer feel at home in the world … Trust in the world, which already collapsed in part at the first blow, but in the end, under torture, fully, will not be regained. That one’s fellow man was experienced as the antiman remains in the tortured person as accumulated horror. It blocks the view into a world in which the principle of hope rules. (Amery [1966] 1980: 34, 40)[vi]

Lest anyone think that Amery’s loss of what he movingly calls “trust in the world” is a melodramatic exaggeration by a powerful writer, a typical report by a group of empirical researchers finds:

Although the psychological impact of torture on the individual is determined by his/her personality and personal history, in most cases the survivor undergoes irrevocable change. Torture often unleashes a traumatic neurosis with symptoms of recurrent dreams of the traumatic event, anxiety, fear, crying, panic and feelings of helplessness.(Kordott et al. 119921 2007: 444)


The best very brief characterization of the American way of torture is law professor Seth : Kreimer’s phrase, “government occupation of the self’ (Kreimer 2003-4: 299).[vii] The CIA

paradigm for torture constitutes the most intrusive imaginable assertion of state tyranny over the individual: the elimination of personal autonomy by means of intentionally undermining the  structure of that individual’s personality and producing a psychological regression to a less differentiated, compliant servant to the torturer’s master. The new twist in contemporary justifications for the use of torture is that they are forward-looking: the torture is part of the intelligence gathering that will keep us safe from, this time around, terrorists. Torture’s “sole purpose is preventing future harms … It becomes possible to think of torture as a last resort of men and women who are profoundly reluctant to torture … Torture to gather intelligence and save lives seems almost heroic” (Luban 2005: 1436; 2006: 42).

Luban observes, however, that such a “protective” state practice is an extreme instance of the kind of political tyranny that Western political philosophers, for centuries, and the global human rights movement, especially since the Second World War, have strenuously condemned: “torture is tyranny in microcosm, at its highest level of intensity” (Luban 2005: 1438; 2006: 43). And Luban’s latest, penetrating, “communicative” formulation is: “Torture of someone in the torturer’s custody or physical control is the assertion of unlimited power over absolute helplessness; communicated through the infliction of severe pain or suffering on the victim that the is meant to understand as the display of the torturer’s limitless power and the victim’s absolute helplessness” (Luban 2014: 128). 

Legal archetypes

Jeremy Waldron has presented an entirely independent case for the absolute prohibition against torture already found in international law, including the International Covenant on Civil and Political Rights, Articles 4 and 7 and the Convention Against Torture and Other Cruel, inhuman or Degrading Treatment, both of which apply to absolutely everyone, and the Geneva Conventions, which apply to the persons they protect, including prisoners of war. Waldron notes that “[n]o one denies that law has to be forceful and final … But forcefulness can take many forms … [L]aw can be forceful without compromising the dignity of those whom it constrains and punishes” (Waldron 2005b: 1727; 2010: 233-34). What he calls “an important underlying policy of law” is:

Law is not brutal in its operation. Law is not savage. Law does not rule through abject fear and terror, or by breaking the will of those whom it confronts. If law is forceful or coercive, it gets its way by nonbrutal methods which respect rather than mutilate the dignity and agency of those who are its subjects.

(Waldron 20056: 1726; 2010: 232)

The legal prohibition against torture, besides its importance in its own right, serves as an archetype of the policy that legal force is not brutal or savage. The prohibition against torture has kind of iconic significance as a symbolic anchor of the intransgressible requirement that law respect dignity by avoiding brutality. Legal archetypes do foreground work as rules or precedents, but in doing that work they sum up the spirit of a whole body of law that goes beyond what they might be thought to require on their own terms. The idea of an archetype, then, is the idea of a rule or positive law provision that operates not just on its own account …but…. also operates in a way  that expresses or epitomizes the spirit of a whole structured area of doctrine, and does so vividly, effectively, publicly …(Waldron 2005b: 1723; 2010: 228)

The archetype is a functioning symbol. Waldron explains the working of the archetype as follows:

[T]he archetype idea is the reverse of a slippery slope argument … Starting at the bottom of the so-called slippery slope, I am arguing that if we mess with the prohibition on torture, we may find it harder to defend some arguably less important requirements that … are perched above torture on the slippery slope. 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 … The confidence we have in them depends partly on analogies we have constructed between them and torture.(Waldron 2005b: 1735; 2010: 243)

If Waldron is correct in his partly empirical hypothesis about the torture prohibition’s anchoring role in legal argument, it is not surprising that the upsurge in the use of torture by American authorities in the twenty-first century has been accompanied by other governmental assaults — all in the name of national security against terrorists — on fundamental civil liberties, including Congressional legislation authorizing indefinite detention without trial, thereby assaulting even the ancient principle of habeas corpus. The maintenance of the prohibition on torture may in fact have been the vital protection of a crucial anchor for the general rule of law, which is now shakier in America than it was in 2000 (McKeown 2009, R. Gordon 2014).

The arguments made by Waldron and Luban in 2005 are strongly complementary. Luban, as we saw earlier, focused in 2005 on the evil of state tyranny; Waldron focused at the same time on the evil of state brutality. Contemporary state-of-the-art torture, as we saw at the beginning, is a form of psychological brutality in which the state invades and at least temporarily occupies the mind, removing the “resistance” consisting of the person’s personality structure. A government that uses torture is both tyrannical and brutal: the CIA paradigm uses brutal assault on the individual’s identity to exercise its tyranny over her. Torture is Kreimer’s “government occupation of the self”. An absolute prohibition on torture is a wall against brutal state tyranny, a wall that, as Waldron emphasizes with the idea of a legal archetype, is also part of the larger structure of the rule of law. The rule of law is protection against tyranny and brutality.

Moral archetypes

David Rodin has suggested that the absolute moral prohibition on torture may play an archetypical role within our system of moral norms somewhat analogous to the archetypical role of the absolute legal prohibition (Rodin forthcoming). If, Rodin observes, one thinks of the Quinean image of a web of belief such that beliefs near the centre of the web can be changed only if substantial portions of surrounding beliefs are also changed, the absolute prohibition on torture serves as one of those key central beliefs. My version of this would be that, as Waldron says of the legal case, the absolute moral prohibition is vital in its own right but is also a symbol of the fundamental point that morality demands limits. A person who will stop at nothing is a person without morality. That no one may ever torture is both an instantiation of a firm moral limit and a radiant emblem of intransgressible moral limitation.

In my account of basic rights, I noted that the enjoyment of every right can potentially be thwarted by any of a multiplicity of “standard threats” (Shue [1980] 1996b: 29-34). A basic right in my sense directly protects one vital interest but in doing so it also blocks a standard threat to other vital interests. For example, a right against CIA-style psychological torture protects the integrity of one’s personal identity, which is intrinsically valuable in the case of every person, but the right against torture also blocks a standard threat against many other rights, such as freedom of assembly and freedom of speech. If I know that if I go to the square and say what I think, I will be whisked away by government agents and tortured in order to find out who my friends are and what I read, the threatened assault against my psychological stability is also a coercive threat against my free speech. A right that protects me against tor-hire, partially protects my free speech as well as my mental balance. In this respect this right helps to anchor other rights.

Rodin (forthcoming) makes the significant additional point, however, that one can embrace an absolute prohibition against torture even if one does not believe that the right not to be tortured is absolute: the absolute prohibition does not presuppose an absolute right. Suppose, for example, that just as many people believe that the right not to be killed is not absolute and that one can forfeit it in at least some cases by killing someone else, the right not to be tortured aught also be thought not absolute and one could forfeit it by, say, torturing others. Suppose, that is, that one believed that torturing a torturer is not a violation of any right. Because they have forfeited that right, we are assuming for the sake of argument, we would not wrong them by torturing them (especially, perhaps, if in accord with the usual narrative of the ticking-bomb scenario, we might possibly thereby obtain the information that would enable us to destroy a secret. site where many other torturers are operating, thereby producing a net reduction in torture, etc.).

Even so, Rodin points out, one might still judge that although it would not wrong the guilty torturer, it would still be wrong — perhaps disastrously wrong — to engage in the torture because it is supremely important to maintain the moral firewall against torture. The function of the absolute moral prohibition against torture as an archetype of the fact that there are some activities in which civilized people do not engage is too important to allow a breach of the prohibition even if the degree of the wrongfulness of torturing the person in question were not reason enough in itself. One is tempted to ask: how bad does it have to be before we torture? Rodin’s point is: nothing compels us to open any discussion of “how bad it needs to be.” We can perfectly well simply say: no torture. As Waldron put it “there are some scales one really should not to be on.”(Waldron 2005b: 1701; 2010: 205, emphasis original).


But, someone is bound to ask, is all this not too confident? Is it not conceivable that if we never torture anyone, we will sonic day pay a terrible price because we will fail to obtain the only information that would have enabled us to avoid a catastrophe: sonic monumental act of terrorism, for example? This is of course simply one more variant of the ticking-bomb scenario: this time there really is a bomb, and it is a really bad one, and we really will get the information in time, but only if we torture, and so on. Yes, if it is conceivable that if we torture enough people, we will find out something very important, then if we torture no one, we may not find it out, arid we may suffer the consequences. Of course, there are all the usual questions about ticking bombs, such as whether we are simply to assume that our intelligence agencies — which did not foresee even monumental developments like the collapse of the Soviet Union and the rise of the Arab Spring, for example — are so good that they immediately have the person who knows most, or, if they have to blunder their way through a sequence of torture victims, each of whom leads them to the next, they will have time to reach the end of the trail. But 1 want to underline two quite different considerations.

First, we have no empirical basis on which to believe that interrogational torture is the most effective form of interrogation (Kleinman 2006: 130 and n.93). If we torture no one, we may not find out something important — unless we have a better method than torture for finding it out! Space here does not allow a thorough discussion of issues about effectiveness, and I obviously do not believe that torture would be justified if it were effective. But it surely could be justified only if it were the most effective alternative, given how wrong everyone on all sides admits torture is. A priori it is difficult to understand how the CIA paradigm in particular can be a good method of gathering accurate information. The person whose personality structure is undermined and who is made to regress to an infantile state of wanting to please certainly will be inclined to give the interrogator what the torture victim thinks the interrogator wants. But is this a reliable method of quickly gathering accurate information? It partly depends on how often a person tortured simply does not have the information that he thinks the interrogator wants; in those cases the victim is likely to manufacture something in order to try to please the interrogator. Amery, under SS torture, tried hard to comply but could not because, like any member of a moderately well-run underground, he had been allowed to know only the aliases of his colleagues:

What they wanted to hear from me in Breendonk [Prison], 1 simply did not know myself. If instead of the aliases I had been able to name the real names, perhaps, or probably, a calamity would have occurred, and I would be standing here now as the weakling I most likely am, and as the traitor I potentially already was. Yet it was not at all that I opposed them with the heroically maintained silence that befits a real man in such a situation … 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.(Amery 119661 1980: 36)

And even if torture does obtain correct information at some stage, it may well be less efficient than other approaches to interrogation. This is certainly the view of a number of experienced American interrogators. For example, Matthew Alexander (a pseudonym) was a US Air Force interrogator who was instrumental in locating the rural safe house in which Abu Musab al Zarqawi, the leader of al-Qaeda in Iraq, was hiding so that he could be killed by an air strike on 7 June 2006 (Alexander & Bruning 2008). The choice of verb in the title of his account, How to Break A Terrorist, is unfortunate because it does not mean what “break” means in the CIA torture paradigm: there is no assault on the personality structure of the person being interrogated. “Break is the jargon we use to signify getting a prisoner to open up a little — like cracking an egg” (ibid.: 108). Alexander has only contempt for the American torturers with whom he had to serve in Iraq, mainly because of their incompetence, which may involve trying to tear down the prisoner’s self-respect (ibid.: 185). Alexander gives his prisoners the opposite treatment: “Still, I have given him hope, and hope is the most powerful weapon” (ibid.: 257).

So, is Alexander interrogating with kindness? Far from it: the hope he gives is false hope. He pretends, and tricks, and lies to his prisoners, treating them in ways that in almost any other circumstances would be clearly immoral:

The best interrogators are outstanding actors. Once they hit that booth, their personalities are transformed. They … allow a doppelganger to emerge. What doppelganger is most likely to elicit information from a detainee changes from prisoner to prisoner. Sometimes I must have a wife or children so I can swap stories with the prisoner, though I have neither.(Ibid.; 91)

The interrogation that Alexander practises does not make a pretty picture. He coercively manipulates his prisoners. They are not treated as ends: they are used instrumentally to obtain information. Some people will certainly feel that it is morally wrong ever to treat people like this.

But it is not torture. Alexander’s technique is unrelenting until he obtains the information he wants, but it is not mercilessly cruel and destructive. No severe pain or suffering, physical or Mental, is inflicted on prisoners. The struggle is a battle of wits. The prisoner’s values and beliefs are not respected: he is, if he can be, tricked into betraying them. He is treated as an enemy and outwitted if possible. But the soundness of neither his body nor his mind is undermined. He is not shamed and humiliated like the men at Guantanamo who were forced to wear women’s Aindenvear and made to learn dog commands like “stay” and “bark” in order to mock their values and undermine their dignity. Alexander’s prisoner may well ultimately regret that his interrogator outsmarted him and obtained the wanted information, but his regret will be possible because his mind will have remained sound, unlike the minds of those assaulted by the CIA torture method. “We don’t have to become our enemies to defeat them” (ibid.: 284).

Of course, the testimony of a single interrogator does not settle which interrogation technique is more effective, torture or something else. I claim only to have illustrated through anecdotal evidence that a serious, smart and successful American interrogator, who personally has conducted three hundred interrogations and supervised more than a thousand, and who was awarded the Bronze Star for service in Iraq, believes that alternatives to torture are more effective. On the other hand, supporters of torture also have only anecdotes at best and fictional accounts (written and film/TV) to support their claims. All the fabulously expensive CIA research established that people can be broken and made to regress, not that they can be made to reveal timely information, even when they know something relevant.

The second point is this. Suppose Alexander is mistaken and was just lucky with Abu Musab al Zarqawi and others, but we have believed him and returned to the legally required refusal to use torture. Consequently, we fail to find out about a danger that we cannot otherwise prevent, and some of us die as a result. I still say: “What kind of world would these children live in if I added torture to the terrorism while imagining that I was saving their lives? Conceivably they would live, not die, but they would live in a world in which even the strong had abandoned restraint and stink further into barbarism … Let’s risk it … and live in civilized countries” (Shue 2003: 91).

Henry Shue, Senior Research Fellow at the Centre for International Studies [CIS] of the Department of Politics and International Relations,University of Oxford;  Professor Emeritus of Politics and International Relations, and Senior Research Fellow Emeritus at Merton is best-known for his book, Basic Rights, (Princeton 1980; 2nd edition, 1996); for his articles, ‘Torture’ (1978) and ‘Subsistence Emissions and Luxury Emissions’ (1993); and for pioneering the sub-field of International Normative Theory, which he taught as an optional subject in the MPhil in International Relations from 2002 until his retirement from teaching. He studied at Merton from 1961-1964 as a Rhodes Scholar. In 1976 he was a co-founder of the Institute for Philosophy and Public Policy at the University of Maryland and later became the inaugural Wyn and William Y Hutchinson Professor of Ethics & Public Life at Cornell University. His research has focused on the role of human rights, especially economic rights, in international affairs and, more generally, on institutions to protect the vulnerable.


[i] Even then it was clear that the hands of the US Government were not entirely clean; for example, the Shah had been restored to power by CIA and was receiving massive US security assistance, as was the South Korean dictator viewed as a bulwark against North Korean communism.

[ii] One Milgram experiment has of course become infamous: see Milgram (1974).

[iii] The Klibark Manual is now available (but was not then, of course: it was classified as “Secret”) at www. 1,wti,edu/–nsarchiv/NSAEBB/NSAEBB122 (accessed April 2014).

[iv] Also see CIA (1963: 41), which observes “regression is basically a loss of autonomy”, and ibid.: 50. Cornell’s Hinkle is quoted as assuring CIA: most people who are exposed to coercive procedures will talk and usually reveal sonic information that they might not have revealed otherwise” (ibid.: 83).

[v] David Sussman is therefore absolutely correct to observe that “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” (Sussman 2005: 22). This not only is not accidental, it is firm policy, in the case of the CIA paradigm. Sussman’s philosophical analysis is one of the most faithful to the actual practices.

[vi] Having physically survived until 1978, Amery committed suicide.

[vii] In his note 73, Kreimer quotes the Supreme Court decision in Stanley v Georgia, 394 U.S. 557 (1969), at 565: “Our whole constitutional heritage rebels at the thought of giving government the power to control men’s minds”.