Torture Interrogation of Terrorists: A Theory of Exceptions (With Notes, Cautions, and Warnings)


by Major (USAF) William D. Casebeer, Ph.D.






Since September 11, there has been a groundswell of support among the general population (and even among some professional ethicists) regarding the need to permit torture interrogation of terrorists and suspected terrorists in certain cases.  The arguments in favor of torture interrogation are usually consequential; they rely on the horrific outcomes of failing to torture certain kinds of terrorists (such as those that possess information about the planned use of weapons of mass destruction) for their moral force.  The arguments against torture interrogation are sometimes consequential, but most often deontological; they usually rely on some conception of inviolable human rights and the duties that correspond to those rights for their persuasiveness.

Conflicts between consequential and rights-based moral theories are not new.  Here, I use insights from previous attempts to meld utilitarian and deontic moral approaches into a coherent system to formulate a “theory of exceptions” that tells us just when it is morally permissible to engage in torture interrogation; such a theory will provide us with principled exceptions to absolute rights.[1]  It closely resembles Michael Walzer’s doctrine of “Supreme Emergency” in terms of structure and has four critical upshots.  First, there are vanishingly few circumstances wherein torture interrogation of terrorists would actually be justified; while they exist in principle, meeting the requirements of the exceptions theory proves very hard to do in practice.  Second, owing to the extraordinary prima facie pressure exerted by the basic human right not to be used as a mere means, most exceptions (while morally permissible) nonetheless leave those involved in the decision to torture with moral “dirty hands.”  Third, owing to the epistemic difficulty of removing one’s self from the exigencies of the circumstance, exceptions require dramatic oversight, and such oversight might have to be international in nature to be effective.  Finally, most consequential justifications for the permissibility of torture neglect to consider the institutional and character-based harm that we do to ourselves when we actually attempt to build a system for torture interrogation that the utilitarian would find praiseworthy.  Perversely, consequential justifications for torture interrogation require well-trained torturers who know where and when to apply pain, but establishing the institutions required in order to sustain such well-honed practice is fraught with perils that the utilitarian would condemn, all things considered.

I conclude that torture interrogation is permissible in tightly constrained circumstances, and that we should not rule it out tout court; nevertheless, in practice it will be practically impossible to justify any particular decision to act on an exception to the prohibition of torture interrogation of terrorists.  More, our decisions to act will require an oversight apparatus that is not presently in place and that may not be possible to create.  Grappling with these issues may very well force us to articulate yet more clearly just what the common bond of humanity requires of us, morally speaking.




First, I’ll briefly motivate an examination of torture interrogation before discussing its definition and previewing my methodology.  I then examine (at a high level of granularity) utilitarian, deontic and virtue theoretic aspects of the practice.  I reconcile the viewpoints offered by the three major moral theories in a doctrine that parallels “supreme emergency.”  After laying out torture techniques along a spectrum of objectionability, I conclude that while, in theory, the requirements can be met, meeting them in practice is very difficult, and that we should keep in mind sundry “notes, cautions, and warnings” before condoning torture interrogation in real world circumstances.

A word on terminology: when you first learn to fly an Air Force airplane, you are given a document called a “Dash One.”  This document contains all the procedures you need to study in order to safely fly the aircraft.  It also contains: notes—important information to keep in mind as you operate the airplane; cautions—telling you what equipment could be damaged by following inappropriate procedure; and warnings—discussing an item that you ignore at the peril of serious injury or death.  This paper constitutes a “Dash One” for torture interrogation, with notes, cautions and warnings stringent enough such that we should think twice before we declare the practice airworthy at all.


Motivating the Issue


            Since the tragedy of September 11, calls for torture interrogation of terrorists have come from unlikely places.  According to one online poll conducted by, 65% of those taking the poll approve of torture interrogation (49% in all circumstances, 7% to prevent acts of terror and gain information about other terrorists, and 9% just to prevent upcoming terrorist action; n = 1821 as of September 2003).[2]  When otherwise staunch defenders of civil liberties, such as Alan Dershowitz, argue that torture might be permissible and that courts should be allowed to issue torture warrants, we should take note.[3]  There are indications that these sentiments resonate with government officials.  For example, Cofer Black, the former head of the State Department’s Counterterrorism Center, says “There was a before 9/11, and there was an after 9/11…After 9/11 the gloves come off.”[4]  A recent Washington Post article quotes a government official supervising the transfer of Camp X-Ray (Guantanamo Bay, Cuba) detainees as saying “If you don’t violate someone’s human rights some of the time, you probably aren’t doing your job…”.[5]  The cover story for the January 9, 2003 issue of The Economist was about torture interrogation, as was that of the October 2003 issue of The Atlantic Monthly.  Given the positions of public intellectuals, government officials, and public interest in the issue, it is time for a serious reexamination of the moral issues surrounding torture interrogation.


Preliminaries and Methodology


            What do we mean by torture interrogation?  According to two major international treaties governing the practice, torture interrogation has a certain nature (severe physical or mental pain and suffering inflicted upon its victim), is accomplished by certain perpetrators (public officials), and has a certain aim (obtaining information or confessions).[6]  In this paper, I limit myself to consideration of whether a government official should intentionally inflict pain upon a terrorist or suspected terrorist in order to gain information.  I will not consider torture as a form of punishment, revenge, or character development—my focus is on torture interrogation.

            In order to answer the question of whether there can be circumstances in which such a practice is justified, I draw upon three traditional tools of moral analysis (the “big three” moral theories of utilitarianism, deontology, and virtue theory)—I throw everything and the kitchen sink at this problem.  This, then, is an exercise in moral casuistry (in the non-pejorative sense): “a resolving of specific cases of conscience, duty, or conduct through interpretation of ethical principles…”.[7]  My analysis will not necessarily satisfy partisans of any of the major moral theories; my hope is that it is a reasonable attempt to accommodate our considered moral judgments about the permissibility of torture.


Substantive Issues


            Utilitarian approaches to torture consider it in light of John Stuart Mill’s “Greatest Happiness Principle,” which states that right actions are those which produce the greatest amount of happiness (or prevent the greatest amount of unhappiness) for all sentient creatures, where by happiness Mill means the presence of pleasure or the absence of pain.[8]  Famously, utilitarians come in two flavors.  Act utilitarians think that the pertinent level of analysis for consideration of consequences is the individual action: with each action that you take, maximize happiness.  Rule utilitarians, on the other hand, think that act utility is short-sighted; instead, we should find that set of rules or laws that, if followed, would produce the greatest amount of happiness, and follow them.  Rule utility is the more popular form of utilitarianism among ethicists of that bent.[9]

            An act utilitarian would license torture in certain circumstances.  When an act of torture would, all things considered, maximize happiness or minimize unhappiness, then it should be done.  The rule utilitarian would be more circumspect, as, on some readings, respect for “the rules” amounts to a de facto respect for the rights of individuals (“rights as rules”; for example, your right to free speech is a reflection of the fact that in my hypothetical rule utilitarian rule set, I would have a rule that said “allow people to speak their mind, for all will be better off if we do so”); even all but the most hardcore of rule utilitarians, however, would admit that the rules have exceptions and that they can conflict (as Mill himself believed).[10]

            Some of the utilitarian considerations that would enter the picture include the short and long term harm done to the prisoner, and the benefits of obtaining the information the prisoner is thought to hold (modulated by the likelihood of actually obtaining it).  The scenario that is normally offered to support a utilitarian justification for torture interrogation is the (infamous) “ticking bomb” scenario: a terrorist has planted a bomb beneath a football stadium.  There’s not enough time to search the stadium to find it.  If we torture the terrorist, he will tell us where the bomb is and we can defuse it before it explodes and kills thousands of innocent people.  In popular treatments, however, the utilitarian analysis usually stops after considering the three consequences I discussed at the beginning of this paragraph.

            This is an incomplete analysis, though, as we should also consider the costs and benefits of establishing the institutions necessary to enable effective torture.  This includes the training and equipping of a professional torture force (sloppy torturers are not as effective at getting suspects to divulge the information necessary to realize the benefits of the interrogation), accomplishing the basic scientific research necessary to support effective torture practice, and oversight and review (so that the torturers really do torture only in the justifiable cases).  For a thorough discussion of these issues, see Jean Maria Arrigo’s paper “A Consequentialist Argument against Torture Interrogation of Terrorists,” posted on the Joint Services Conference on Professional Ethics website.[11]  In addition, there are character development issues that even the utilitarian should be concerned about: what of those involved in the professional torture interrogation unit?  Will they lead healthy and complete human lives?  What will the long term psychological consequences of being a torturer do to their chances at achieving some measure of human happiness?

            Even when taking these consequences into account, though, it is conceivable that a utilitarian, particularly an act utilitarian, would permit torture interrogation in certain instances.  I’ll discuss this in more depth in a moment.

            The second major moral theory, deontology, does not focus on consequences to ascertain the morality of an action; rather, it focuses on the nature of an action to judge its worth: does the action’s very structure respect the rights of human being?  Do we have a duty not to engage in torture interrogation, and do individuals have a right not to be tortured?  Deontological approaches are concerned with respecting the existence of those things that make morality possible at all—the transcendental preconditions for there to be morality—namely, the existence of human free will and reasonability.  Our agency gives us a dignity, Immanuel Kant would say, that is beyond price.  Kant captures the demands of duty in one formulation of his “categorical imperative” (that thing which commands free and reasonable creatures unconditionally, at all times and places): “Never treat humanity, either in your own person or in the person of others, merely as a means, but always also as an end.”[12]  We avoid treating other humans as mere means by getting their consent, either implicit or explicit, but in the best of cases explicitly, to involve them in our plan, project or practice.  Questions of consent are thus paramount.

            Kant’s theory of justice allows you to treat people as they are, with their own maxims and principles, asking to be treated.  In this way, Kant would say that the death of a combatant in a war, so long as such a combatant was willingly involved in the process required to wage war, is not necessarily a violation of the demands of the categorical imperative.  But the consent we give in such ways is not unbounded; when I cease to be a combatant, I am no longer consenting to give up my life for the cause which I am defending.

            The rights-based focus of a deontological perspective means that those looking to license torture must obtain the consent of those being tortured.  Needless to say, it is implausible to think that terrorists have given such consent.  They would not give it explicitly, and the implicit consent also seems to evaporate once they have been captured, as the terrorists are no longer combatants at that point.[13]

            Setting aside issues of justice (recall that we are torturing in order to obtain information, not to punish), then, the tone of the Kantian corpus is difficult to reconcile with the practice of torture interrogation.  Deontic approaches give us a prima facie argument against torture interrogation: it is a violation of human rights.

            The third major approach to morality—virtue theory—is usually left out of discussions about torture interrogation.[14]  In part, this may be because it is sometimes difficult to wring out public policy related guidance to a particular moral issue using the virtue theoretic critical infrastructure.  Virtue theorists such as Aristotle or Plato focus on eudaimonia, variously translated as flourishing or proper functioning.[15]  What kind of person must I be if I am to live a fruitful human life wherein I fulfill the human telos or end—my function?  Virtues are those states of being that enable us to be excellent human beings, while the vices are those states which cause us to be dysfunctional human beings.  What habits must I have if I am to flourish, and what kind of person, in the ultimate sense, do I want to become?  While virtue theory might not have as a direct entailment that virtuous persons would never be involved in torture interrogation, that is at least a plausible first pass of its indirect entailment.  Virtuous people are probably not involved in the intentional and coerced causation of pain and suffering in other human beings, at least not in ordinary circumstances.  And if they are, they probably face a “dirty hands” problem (that is, even if torture interrogation is justified in certain exceptional circumstances, torturers may nonetheless be “damaged goods,” so to speak, from a moral psychological point of view) in much the same way that (some would argue) those involved in warfare where innocents are killed may have dirty hands.

            Virtue theoretic considerations speak most strongly against the maintenance of a professional torture force.  To be the kind of person whose techne just is being skilled at the application of pain to others for the purpose of forcing information from them is not a prescription for a complete and fulfilled human life.  This is problematic, as the utilitarian would demand that the torturer be well-trained, skilled in the “profession,” and precise in the application of pain—the utilitarian would not want excess pain to be generated, nor would torture that fails to cause the victim to reveal the information be praiseworthy.  But these very skills require time and experience to develop, and it is exactly this aspect of the practice of torture interrogation to which the virtue theorist would have the strongest objection!

            To summarize our “first draft” analysis, then: if there are reasons to torture, they will almost certainly be act utilitarian in nature.  Act utilitarian considerations are important, but they should probably be “side constrained” (as Robert Nozick would argue) by deontic or rule utilitarian considerations.  Nonetheless, what of exceptional circumstances?  Is it really the case that I must respect a right though the heavens may fall?  Other rights-based theorists have attempted to accommodate this intuitive objection from the utilitarian camp.  In the case of war, Michael Walzer argues that while the innocent noncombatant has a right not to be killed, in certain exceptional circumstances we can justifiably violate that right.  Walzer calls those exceptional circumstances cases of “supreme emergency.”  Can we articulate a parallel set of standards for torture interrogation?


Supreme Emergency for Torture Interrogation


            In his book Just and Unjust Wars, Walzer argues that when a danger threatening a nation-state is of a certain nature, we can justifiably, but with regret, violate the rights of innocent people not to be killed in order to prevent it.  The danger threatened must not merely be the loss of my life, or even that of a few hundred combatants or innocents; rather, it must be a grave danger that threatens the very existence of a nation-state, a political community, or a culture.  The danger must be “supreme,” and not merely run-of-the-mill.  Moreover, this danger must be imminent, both spatially and temporally; if the danger is distant, then the likelihood that we can avert it without violating someone’s rights increases dramatically.  The “imminence” and “nature of danger” tests are framed against a backdrop of last resort, of course; if there is a way we can prevent the danger from occurring without violating human rights, then we are morally obligated to do so.  In other words, respect rights until the heavens really are about to fall; but when they are about to fall, you may justifiably, but with regret, intentionally kill an innocent person or otherwise violate someone’s rights.  This bit of casuistry is designed to nod respectfully in the act utilitarian direction in those most extreme of circumstances.[16]

            We can formulate a parallel doctrine for torture interrogation.  It may help to think of a spectrum of torture interventions.  At one end, we have as anchor torture interrogation methods that leave no lasting physical or mental damage and only minimally disable the human agency that deontic approaches remind us to respect.[17]  At the other endpoint are methods that are catastrophic, leaving lasting physical or mental damage, and causing such severe pain that human dignity is crushed.  Methods at the “no lasting damage” end might include such approaches as putting a hood over a suspect, bombarding them with noise, depriving them of sleep or of food (only minimally), or forcing them to stand before interrogation.  These methods, for example, were used by Great Britain in the 1970’s when questioning suspected Irish Republican Army members, and were called “the five techniques.”[18]  The threat of the use of force, or a lie told to encourage the divulgence of information, might also fall towards the “no lasting harm to agency” end of the spectrum.  Of course, even these methods, when applied continuously over time, can become extremely cruel, at which point they would move towards the other end of the spectrum.  Methods which fall on that end from the get-go include such things as catastrophic damage to the brain or to the sexual organs, the breaking of bones, or similar tactics that cause prolonged and intense suffering. 

The “easy cases,” then, will be those where the minimally threatening methods are used so as to prevent a catastrophic danger.  Extremely mild “torture” interrogation might be morally required in those circumstances.  But if the consequence to be prevented is mild, even they would be morally prohibited.  On the extreme end of the spectrum, if the consequence prevented is mild, the tactics are again obviously morally prohibited.  The truly hard case regards the use of torture interrogation techniques that drift away from the no harm done end of the spectrum but which nonetheless seem to be required to prevent a “supreme emergency” style catastrophic danger from materializing.  In those cases, I would argue, torture interrogation may be a morally justifiable tactic.  Figures one and two make this clearer.



Figure One: Easy Cases on the Torture Spectrum



Figure Two: The Hard Case on the Torture Spectrum




The Difficulty of Meeting the Requirements in Practice


But note that meeting the requirements similar to those for “supreme emergency” is exceedingly difficult in practice.  While we can envisage such scenarios arising theoretically, practically speaking we will almost never be confronted with them.  First, note that usually it is only a toy ticking-bomb style scenario that leads to a favorable cost benefit analysis; when the oft-neglected institutional and character-based harms are included, it becomes more difficult to see when the practice could be warranted.  Moreover, it will hardly ever be the case that a single act of terrorism will threaten the very existence of a nation-state or political community; the nature of the danger prong will be difficult to meet.  The imminence of the danger requirement will probably only be met in radically underspecified thought experiments like the ticking bomb scenario (indeed, the very intelligence that will enable us to know we are facing an imminent danger will also likely serve to give us means to discover the source of the danger without having to resort to torture interrogation).  Finally, we will rarely be in a case of last resort.  Other methods of intelligence collection will usually be available, and the efficacy of torture is doubtful in many circumstances (keep in mind that we are allowed to violate rights only if that will be an efficacious means to prevent the danger).


Notes, Cautions, and Warnings for a Theory of Exceptions


As we grapple with this attempt to articulate a theory of exceptions telling us when torture interrogation is permissible, we would do well to keep in mind the notes, cautions and warnings implicit in our substantive analysis.  Note that even when the exceptional circumstance is present, we have a “dirty hands” problem—we have violated human rights, and this makes even the exceptional circumstances problematic, morally speaking.  Be cautioned that in order to prevent self-deception about whether we are in supreme emergency from occurring, and in order to jump the epistemic hurdles presented by these cases, torture interrogation will require oversight; such oversight will at the very least require the intervention of our judiciary and perhaps the Congress, and it may even have to be international in order to be truly effective.  Finally, keep in mind this warning: if we are not careful, our attempt to train and equip for these exceptional circumstances can create a culture that permits the worst degradations to dignity know to humanity to be licensed by our political institutions and carried out by our fellow human beings.  We should take a deep breath before creating such a culture.




Our attempt at reasonable casuistry about the problem of torture interrogation has brought us full circle.  I have argued that it is possible to formulate a theory of exceptions for torture interrogation.  This theory can be patterned after another attempt to reach a reasonable compromise between utilitarian and deontic demands: Walzer’s doctrine of supreme emergency.  However, the requirements for this test are stringent enough that while severe forms of torture interrogation might be permissible in certain circumstances, those circumstances will almost never present themselves in actual practice.  For all practical purposes, the requirements will never actually be met.  Our common bonds of humanity demand certain things of us, and while those bonds can be broken in certain extremes, there are practical and epistemic hurdles to be cleared before we can do so justifiably.  Our “Dash One” for torture interrogation gives us reason to doubt whether our torture interrogation airplane can get enough altitude to clear those hurdles in all but theoretical circumstances.



Works Cited  Last accessed September 18, 2003.


Aristotle.  384-322 BC/1999.  Nicomachean Ethics, Second Edition.  Trans. Terence

Irwin.  Hackett Publishing Company, Inc.: Indianapolis.


Arrigo, Jean Maria.  2003.  “A Consequentialist Argument Against Torture

Interrogation.”  Joint Service Conference on Professional Ethics, 2003.  Last

accessed September 18, 2003.


Bowden, Mark.  2003.  “The Dark Art of Interrogation: A Survey of the Landscape of

Persuasion.”  The Atlantic Monthly, Vol. 292, No. 3, October, 2003.


Dershowitz, Alan.  2002.  Why Terrorism Works.  Yale University Press: New Haven.


Duner, Bertil.  1998.  An End to Torture: Strategies for its Eradication.  Zed Books: New



Forrest, Duncan (ed).  1996.  A Glimpse of Hell: Reports on Torture Worldwide.  New

York University Press: New York.


Fotion, Nicholas.  2003.  “Keynote Address” to the 2003 Joint Service Conference on

Professional Ethics.  Author’s Notes.


Hooker, Brad.  2003.  Ideal Code, Real World:  A Rule-Consequentialist Theory of

Morality.  Oxford University Press: New York.


Kant, Immanuel.  1785/1993.  Grounding for the Metaphysics of Morals.  Trans. James

W. Ellington.  Hackett Publishing Company, Inc.: Indianapolis.


Mill, John Stuart.  1863/1993.  On Liberty’ and ‘Utilitarianism’.  Bantam Books: New



Murphy, Jeffrie G.  1973.  “The Killing of the Innocent.”  The Monist, Vol. 57-64.


Plato.  427-347 BC/1968.  The Republic.  Trans. Allan Bloom.  Basic Books: New York.


Priest, Dana and Gellman, Barton.  2002.  U.S. Decries Abuse but Defends

Interrogations.”  Washington Post, December 26, 2002.  Page A01.


Shue, Henry.  1978.  “Torture.”  Philosophy and Public Affairs, Vol. 7, No. 2.


Walzer, Michael.  1992.  Just and Unjust Wars: A Moral Argument with Historical

Illustrations, Second Edition.  Basic Books: New York.


Webster’s Ninth new Collegiate Dictionary.  1987.  Merriam-Webster Inc., Publishers:

Springield, MA.

[1] Nicholas Fotion calls this attempt to articulate just when we are permitted to do otherwise morally impermissible things a “theory of exceptions” (see his keynote address to the 2003 Joint Service Conference on Professional Ethics ).  I use his language here.

[3] See his book Why Terrorism Works (2002).

[4] Washington Post, December 26, 2002.

[5] Ibid.

[6] See the UN Declaration of Human Rights of 1948, or the UN Convention Against Torture of 1984, both of which the United States has signed (for a much more thorough review of the relevant international treaties and legal documents, see Kellberg (1998)).  In this essay, I am concerned more so with the morality of torture interrogation, less so with its legality.

[7] Webster’s Ninth New Collegiate Dictionary (1987), p. 213.  I hope I am not engaging in casuistry in the second sense of the term: “specious argument” (Ibid).

[8] See chapter two of his essay “Utilitarianism” (1863/1993).

[9] See Brad Hooker’s Ideal Code, Real World: A Rule-Consequentialist Theory of Morality (2003).  Exemplars for each school include Peter Singer, an act utilitarian, and Richard Brandt, a rule utilitarian.

[10] See the last two pages of chapter two of Mill (1863/1993).

[12] See Kant’s Grounding for the Metaphysics of Morals (translated by James Ellington, 1785/1993).

[13] Arguably, in any case.  Much more philosophic work needs to be done here, and it may very well be that certain kinds of terrorists do not cease to be in the logical and causal chain required to do harm to another when they are captured.  See Murphy’s seminal article on the combatant/noncombatant distinction for more discussion of this idea (Murphy, 1973).

[14] See, for example, Henry Shue’s excellent article “Torture” in Philosophy and Public Affairs (1978), which does not discuss in any detail the virtue theoretic aspects of the practice.

[15] Consult, for instance, Aristotle’s Nichomachean Ethics (384 – 322 BC/1999), or Plato’s Republic (427-347 BC/1968).

[16] For a thorough discussion of supreme emergency, with enlightening historical examples, see chapter 16 of Just and Unjust Wars (1992).

[17] The hard core Kantian may insist that agency is an all or nothing affair.  I don’t intend to engage in heavy deontic exegesis here; I rely on the common sense intuition that some rights violations are much less serious than others, and that my free will can be disrespected in minimally disruptive ways and in maximally disruptive ways.

[18] See Gudjonsson (as printed in Forrest (1996)).  The Forrest volume is a useful reference for facts about many aspects of torture interrogation.