Physician Involvement in Hostile Interrogations

It should be recognized that there are a host of moral issues that confront physicians in times of war, though I will herein focus only on the issue of their involvement in hostile interrogation.⁵

To this end, consider what allegations have actually been made against physicians: The International Committee of the Red Cross (ICRC) has reported that the medical staff at Guantanamo Bay has shared patient records with military personnel who planned interrogations, and the ICRC has called these actions “a flagrant violation of medical ethics.”⁶ In response, the Pentagon has claimed that its detention operations are “safe, humane, and professional” and that “the allegation that detainee medical files were used to harm detainees is false.”⁷ Gregg Bloche and Jonathan Marks further allege that documents and interviews have shown not only that medical personnel shared confidential documents with potential interrogators, but that “physicians assisted in the design of interrogation strategies, including sleep deprivation and other coercive methods tailored to detainees' medical conditions. Medical personnel also coached interrogators on questioning technique.”⁸ Whether these allegations turn out to be true is, of course, an empirical question beyond the scope of this paper. However, their truth is irrelevant to the moral debate that investigates whether such acts (actual or counterfactual) are morally permissible.

I propose to proceed in two stages. First, I consider moral arguments against physician involvement in hostile interrogations and show why these arguments are problematic. Second, I advance positive arguments against the thesis that medically trained interrogators are physicians. Just for precision, let us designate the controversial activities as the following: physicians' sharing of medical records (especially with would be interrogators), physicians' development of interrogation strategies (especially strategies that make special use of medical knowledge and/or medical susceptibilities of detainees), and physicians' direct participation in hostile interrogations (whether in an oversight or advisory role or else physical participation in the interrogations). What arguments could be offered against these practices?

The most obvious argument would be that these practices run contrary to the moral nature of medicine and the moral obligations of physicians. Since the advent of medical ethics, harkening back to Hippocrates, medicine has been argued to be an inherently moral enterprise. Correspondingly, physicians have been ascribed various moral responsibilities; the most notable of these have been beneficence, nonmalfeasance, confidentiality, honor, and loyalty.⁹

Certainly we think that physicians should help patients (beneficence) and that they should not harm them (nonmalfeasance). Confidentiality is also an important moral good insofar as it is necessary for optimal medical care: If patients cannot trust that their physicians will maintain confidences, patients will be less likely to disclose relevant medical information and physicians will not be able to provide as effective treatment. Physicians have always been among society's more esteemed members, and they (arguably) therefore have a greater obligation to serve as role models and to carry themselves honorably. Finally, we expect physicians to be loyal to their patients and to serve the interests of those patients, uncompromised by extraneous influences or conflicts of interest. Physicians who turn over medical records to potential interrogators would arguably violate all five of the aforementioned moral ideals; thus such an act would be morally problematic. Physicians who devise hostile interrogation strategies would at least be violating the principle of nonmalfeasance, and active participation in these interrogations would be similarly morally problematic.

First, consider the extent to which endorsement of these core medical values actually precludes physician participation in torture, though these results will be readily transferable to the less controversial hostile interrogations. The American Medical Association's Code of Medical Ethics is quite unequivocal on this question:

Certainly these statements inveigh against participation in the three aforementioned controversial practices. However, I find the injunction against physician presence during torture to be peculiar and problematic. If torture is morally permissible (or even if it is merely occurrent), then the traditional values of medical ethics mandate physician presence during the interrogations. This contention has been vehemently criticized (Kenneth Kipnis and Matthew Wynia, personal communication), but it directly follows from the principle of beneficence: insofar as this principle motivates concern for the welfare of the detainee, then physician oversight is morally obligatory to ensure the safety of the interrogatee. Obviously the well-being of the interrogatee is jeopardized during torture (and, less extremely, in hostile interrogations), and there is a chance that, absent physician intervention, she or he might die, suffer irreversible damage, and so forth. For example, imagine that, during an interrogation, an interrogatee were to go into cardiac arrest and that, pursuant to AMA opinions, no physician were present. The interrogatee might easily suffer a preventable death, and this would be entirely unacceptable.¹¹

The rub, of course, is the principle of nonmalfeasance: Physicians must not revive interrogatees just so that the interrogatees can be tortured even more, ultimately being made the worse off for the physician intervention. A couple of comments are appropriate here. First, at least in some cases, torture would not be worse than death. Therefore, resuscitating someone merely so she or he can face more torture does not necessarily violate the principle of nonmalfeasance if, absent resuscitation, she or he would have been even worse off (e.g., dead). And, again, the principle of beneficence would seem to require these interventions.

Presumably the argument against this claim would effectively be epistemic: Physicians would not know whether resuscitation would lead to further hostile interrogations and therefore whether interrogatees would be made worse off from physician intervention. However, this is not a moral argument and would be impotent against the claim that physicians should resuscitate (and, a fortiori, be present during the interrogation) if resuscitation were in the best interest of the interrogatee. But even contra the epistemic worry, the interrogatees, at least in some cases, would be better off even if their resuscitation led to more torture (or, less controversially, to more hostile interrogating) simply because their lives would usually be ones worth living so long as the interrogations would eventually cease and they could resume a quasi-normal (if detained) life. Regardless, the principle of beneficence requires at least minimal physician participation in hostile interrogations, namely, in those cases where physician intervention would be in the medical interest of the interrogatee.

Are Medically Trained Interrogators Physicians

The previous section investigated how traditional medical values would constrain physician involvement in hostile interrogations, though I argued that these values actually mandate at least minimal participation. However, it has so far been assumed that traditional medical duties or responsibilities apply to medically trained interrogators, and this assumption is contestable. Ultimately, my conclusion will be that medically trained interrogators are not physicians, and therefore are exempt from whatever medical duties or responsibilities might otherwise be incumbent upon them. To motivate this discussion, consider actual remarks made by David Tornberg, Deputy Assistant Secretary of Defense for Health Affairs: Tornberg argues that medically trained interrogators who helped to plan interrogations are not acting qua physicians (i.e., have not entered into a patient–physician relationship) and are therefore not bound by confidentiality, beneficence, nonmalfeasance, and so forth.¹²

Tornberg further contends that a medical degree is not a “sacramental vow,” but rather a certification of technical merit. Some military physicians and Pentagon officials have claimed that their medically trained personnel act as combatants, not physicians, when they put their medical knowledge to use for military ends.¹³

As a possible defense of this position (which they ultimately reject), Bloche and Marks propose to consider civilian parallels, wherein the “Hippocratic ideal of undivided loyalty to patients fails to capture the breadth of the profession's social role.”¹⁴

The general problem, that of dual loyalties of the physician, warrants far more discussion than this paper can afford, but the simple point is that there are at least some cases in contemporary society where physicians have duties or responsibilities beyond those merely to their patients; examples include forensic psychiatry, occupational health, and public health.¹⁵ We certainly could look at the hostile interrogation debate as one of dual loyalties. The idea here would be to say that medically trained interrogators have duties or responsibilities to the interrogatees as well as to something else (e.g., the military chain of command, national security) and that the latter duties trump the former. Although I think that this approach would be profitable and that these extramedical invocations could successfully countervail medical duties or responsibilities,¹⁶ I wish to defend the more extreme claim that there are no medical duties or responsibilities that the medically trained interrogator has to the interrogatee, or at least no “special” duties or responsibilities that present themselves merely in virtue of the interrogator's medical knowledge and that could not be accommodated by general moral approaches (e.g., consequentialist or deontological). In other words, no tension results from a physician's dual loyalties (i.e., to the interrogatee and elsewhere) because the medically trained interrogator is not a physician at all (nor acts qua physician during the interrogation).

Whether the medically trained interrogator has special duties or responsibilities to the interrogatee depends on two central questions. First, has the interrogator entered into a patient–physician relationship with the interrogatee? Second, even if she or he has not, would there be other arguments (i.e., ones not based on invocation of nonexistent patient–physician relationship) that would inveigh against the use of medical knowledge in a way that could bring harm to the interrogatee? One quick way to get around the issue of the patient–physician relationship is to note that physicians have the prerogative to refuse to enter this relationship. According to the American Medical Association's Code of Medical Ethics: “Physicians are free to choose whom they will serve. The physician should, however, respond to the best of his or her ability in cases of emergency where first aid treatment is essential.”¹⁷

The first statement here is quite unequivocal, though the latter qualification is somewhat confusing. I read it as a nonbinding suggestion because, otherwise, it would contradict the first statement (which then should have been written as “physicians are free to choose whom they will serve so long as …”). (The invocation of “essential” is also confusing: essential for what?) Elsewhere, the Code of Medical Ethics says: “[I]t may be ethically permissible for physicians to decline a potential patient when … [a] specific treatment sought by an individual is incompatible with the physician's personal, religious, or moral beliefs.”¹⁸ Although the language is not entirely congruent with the issue we are trying to consider, this statement affords physicians the license to decline patients on moral grounds because the physician might think that any treatment would be inappropriate for the interrogatee. The physician might think this because she or he could ascribe to a moral view that would license the hostile interrogations.

Although the Code of Medical Ethics supports a physician's prerogative not to enter into a specific patient–physician relationship, this prerogative could also be defended by simple moral philosophy. Because this relationship is a deontic one (i.e., one defined by duties or responsibilities), its formation would presumably be grounded in deontological ethics. Taking Kant as the standard torchbearer of this enterprise, we could observe that it would be a violation of the physician's autonomous will to force him or her to enter into a relationship if he or she does not want to; such a forcing would be morally impermissible. So, whether we invoke the Code of Medical Ethics or Kant, arguments can be made for the physician's right to decide whether to enter into a particular patient–physician relationship (the patient would obviously also have to consent in order to establish the relationship). Therefore, we have at least one argument supporting the claim that medical interrogators do not have medical duties to interrogatees: So long as the physicians do not wish to assume the relationship, the duties will not apply. If we limit our considerations to medically trained interrogators who willingly assume their hostile roles (as opposed to merely assuming them once commanded to), we might reasonably infer that these interrogators do not wish to assume a patient–physician relationship with the interrogatee and therefore are exonerated from the associative moral duties.¹⁹

These previous paragraphs have tried to establish that physicians can simply choose not to enter into a patient–physician relationship and that, absent this constitution, they are exempted from the associative moral duties. However, these comments are not overwhelmingly convincing, though they do merit consideration. For one thing, the citations from the AMA Code of Ethics are not even arguments at all, they are merely statements (and, furthermore, statements that might not be interpreted in the way that was suggested). Because the Code of Ethics only issues statements, we have no reasons (as might be offered by premises and a purported inferential structure) to accept them aside from the fact that the American Medical Association endorses them. This is, to my mind, the devastating weakness of the Code of Ethics for philosophical purposes, because philosophy requires argumentation and not simply stipulation by fiat. And any deontological argument which prioritizes individual autonomy (viz., the autonomy not to enter into a patient–physician relationship) will be impotent against anyone who does not ascribe to such a moral theory. So, although these ideas should be taken seriously, more argumentation would be useful.

Therefore, I propose to directly argue against the view that medical knowledge confers moral duties, including the moral duty to establish a patient–physician relationship and absorb the associative moral burdens. Or, to explicitly reference the current administration's position, I want to defend the notion that a medical degree is merely a certification of technical merit and not a “sacramental vow.”²⁰

The conclusion, already mentioned, is that medically trained interrogators have no medical duties or responsibilities to interrogatees, and this is a stronger claim than the one that would hold that such duties exist but are countervailed by other duties (e.g., toward the public good). Three arguments will be offered in favor of this position: the logical argument, the metaphysical argument, and the argument from analogy (with other professions).

First, consider the logical argument. People who claim that medically trained interrogators are violating medical duties often say something like: the medically trained interrogator has medical knowledge, therefore she or he has certain moral duties. (Sometimes this is expressed as: medically trained interrogators have certain moral duties in virtue of their medical knowledge—this equivalent expression just inverts the premise and conclusion). At the risk of being tedious, this argument has one premise (viz., that the medically trained interrogator has medical knowledge) and one conclusion (viz., that she or he has certain moral duties). As currently formulated, the argument is at least formally invalid (i.e., we might represent the argument as P, therefore Q, which is formally invalid). Furthermore, many philosophers endorse a fact-value divide such that we cannot validly move from descriptive premises to a normative conclusion; this principle is not above debate, but something always looks suspicious about attempts to circumvent it.²¹

Accepting this principle, the argument is again going to be invalid because the premise is descriptive and the conclusion is normative.

The most obvious way to try to repair the argument's doubly alleged invalidity would be to say that it is enthymatic: If we can restore the suppressed premise, then we can render the argument valid. What is the suppressed premise? One likely candidate is: If the medically trained interrogator has medical knowledge, then she or he has certain moral duties. A coupling of this premise with the original premise will yield the conclusion by simple application of modus ponens, thus making the revised argument valid. But is it sound? I think not, which leads to my second argument: the metaphysical argument.

So consider this new premise: “If the medically trained interrogator has medical knowledge, then she or he has certain moral duties.” For the argument to be sound, this premise would have to be true. But it is not; this premise (and ones like it) has dubious metaphysical commitments. More formally, such a premise holds that because A knows P, A is morally required to φ.²²

Or, in other words, the knowledge of P is sufficient to obligate an agent to φ. But this just seems obviously wrong. It cannot be the case that mere knowledge of some (nonmoral) proposition (or set of propositions) can obligate someone to do something. Rather, normative principles are the sorts of things that create moral obligation. For example, if A is obligated to φ, the reason has to be that φ maximizes happiness, that its negation cannot be willed to be universal law, and so forth; obligation can follow from these sorts of normative claims. But how could obligation possibly follow from propositional knowledge alone? Certainly such knowledge could be necessary for moral obligation: Insofar as we endorse the “ought implies can” principle, we cannot be obligated to that which we are unable. And, insofar as knowledge could be a necessary precondition for ability (e.g., A cannot save the drowning child unless she or he knows where the child is), moral obligation would not exist absent knowledge of the appropriate propositions. But the necessity of knowledge for moral obligation is irrelevant here. Rather, the proponent of medical knowledge's ability to create moral duties must defend the sufficiency of knowledge for moral obligation, and this is metaphysically problematic because knowledge alone cannot create moral obligation. Medical knowledge alone is not sufficient to create moral obligations absent some moral principle that would yield those obligations. And remember that, at least for the sake of argument, we are supposing that some moral principle could license the hostile interrogations themselves if not physician participation in them. Although this argument decisively refutes the opposing view, there is also room to consider a third argument: the argument from analogy.

The argument from analogy proceeds by looking at how moral obligations can be said to work in other professions; maybe we can learn something about medicine (i.e., the controversial one) by looking at less controversial professions. For example, take engineering ethics, although similar examples could be illustrated by appealing to other professions. The analogous question would be: Do engineers have moral obligations merely in virtue of their technical knowledge? For example, do chemical engineers, qua chemical engineers, have duties not to construct chemical weapons? I think not: It would only be impermissible to construct these weapons if such constructions were morally impermissible as dictated by some plausible normative principle. In other words, there is nothing intrinsic about their technical knowledge that would morally prohibit them from doing something. Rather, the moral wrongness of any application of their technical knowledge must reside in some incriminating moral principle. If this is true and if this engineering case is analogous to the medical case, then it would be implausible to suggest that medically trained interrogators have moral duties or responsibilities in virtue of their medical knowledge.

Is the case analogous? Or, in other words, is there anything that makes medicine special as a profession? First off, I am against things being special (i.e., exceptions to general principles) because heterogeneous conceptions of value always seem less parsimonious than homogenous ones (and rarely have compensating and/or countervailing advantages). Someone might try to say that medicine is special on the grounds that it aims, most fundamentally, at healing, and that healing is an inherently moral project. Engineering, she might continue, is not inherently moral because it merely aims at, let us say, building things. So, even if the critic were to share my intuition in the engineering case, she might argue that it is irrelevant in the medical case. Medicine does not necessarily (i.e., conceptually) aim at healing: The field of medicine (as any other field) is merely constituted by an accumulation of facts, and facts do not do anything, much less heal people.²³

Rather, medical knowledge (as with all knowledge) can be applied to any ends, whether healing or harming. If those with medical training should heal, it is because pleasure is better than pain, because people consent to healing and not harming, and so on, not because there is any intrinsic feature of medical knowledge such that it should be applied to healing. So, contra the critic, the analogy stands.

One objection to the aforementioned three arguments might be that they are against the wrong thesis: They have been considering whether medical knowledge alone gives rise to medical duties, and we might instead think that this is the wrong question to be asking. To wit, another question might be whether it is a physician's role (as opposed to merely his knowledge) that gives rise to the medical duties. Although the structure of my answer to this question is already in place, some further remarks are warranted. The idea of role-differentiated morality is fairly intuitive: Many of us think that duties or responsibilities are sometimes dictated by the roles that moral agents occupy. For example, we might think that parents have a stronger duty to provide for their own children than do complete strangers, and that law enforcement is more justified in using lethal force than the general public would be. Although these ideas could be discussed in more length, they are so intuitively plausible that I will not pursue their justification.

The application of these concepts to our current investigation is straightforward, and the question now becomes: Is there something about the role of the medical interrogator that could ground medical duties or responsibilities? Presumably, this challenge assumes that this question is logically distinct from the one I have been asking, which is whether medical knowledge alone grounds medical duties or responsibilities. However, even accepting the thesis of role-differentiated morality (which I do), I do not think that the interrogator's role is relevant to our inquiry.

What is the role of the interrogator? Certainly it is not that of a physician: The interrogator's primary task is to facilitate the acquisition of information, not to heal. So, as a matter of empirical fact, the role the interrogator actually plays is not that of physician, but rather that of interrogator (hence the job title). The relevant objection to this assertion is that we are not concerned with empirical fact, but rather with the normative realm. So now we must ask whether the interrogator should assume the role of physician and therefore the associative moral duties. And, again, I think that the answer is no.

Remember that we are assuming, at least for the sake of argument, that hostile interrogations are morally justified. Therefore, absent any other considerations, the interrogator's role is justified a fortiori (as a necessary element of those interrogations). What about other considerations? We might propose that the medically trained interrogator is obligated to assume some special role that would ground medical duties or responsibilities, but I contested this claim in the second section. We can now exhaust logical space by postulating a third option, that the medically trained interrogator is somehow thrust into the role of healer, even if she or he would rather not assume this role, and therefore bound by the duties and responsibilities thereof. But why would this be true? If it were true, it would have to be in virtue of his or her medical training, because that is the only feature that could serve as a differentia between him or her and anyone else. Whether the duties and responsibilities derive directly from the interrogator's medical training or are else mediated through some role is irrelevant, because either position would be committed to the sufficiency of medical training for these duties and responsibilities.²⁴

And this, of course, is the claim contested earlier.

Finally, a few remarks on the notion of professionalism and oaths are in order, as these notions have been invoked in criticisms against my position.²⁵

First, the medically trained interrogator is acting outside of the profession of medicine and thus exempt from whatever standards might therein apply. If professional societies do/should play some role in safeguarding the profession, then they might well take issue with admitting medically trained interrogators into their ranks. For example, we might imagine that these people return from their military assignments and want to practice medicine: Should they be licensed by local boards? Maybe we think that their ability to heal has been compromised or else that we simply do not want such people as part of the profession. I do not really have views on these issues, other than to say that they are irrelevant to my current inquiry (albeit interesting in their own right). In addition, one might wonder whether medically trained interrogators are violating some oath (e.g., Hippocratic) they might have at one time taken. Here two comments are important. First, we could easily imagine medically trained interrogators who never took such an oath and therefore, ex hypothesi, would not be breaking one. If the only argument against medically trained interrogators is merely a contingent avowal of some oath, then this is not a particularly powerful objection. For example, we could just identify prospective medically trained interrogators earlier and make sure they never uttered the words. Then what would be the objection? But the second comment might be more substantial: Oaths only govern behavior within a profession, and medically trained interrogators lie outside of professional medicine.²⁶ To be sure, there are some duties that would “follow” medically trained interrogators as they leave the profession (e.g., the duty to maintain confidences established while part of the profession). However, those duties can only attach to particular relationships forged while in the profession and cannot provide blanket edicts that universally apply after departure from the profession. So, whether the medically trained interrogator never enters the profession of medicine or else leaves to go interrogate, arguments predicated upon professionalism and oaths are impotent in rendering moral indictments against him or her.