Virginia Journal of International Law
Fall, 2005
46 Va. J. Int'l L. 197
The Politics of the Geneva Conventions: Avoiding Formalist Traps
Rosa Brooks
It always feels good to do what you're good at, and, as Sam Estreicher observed during an earlier session, lawyers are good at formalism. When we talk about the Geneva Conventions, it's particularly tempting to retreat into formalism, because emotions so easily start running high: after all, if we leave the tidy formalist world, we're into a messy argument about good and evil, right and wrong, terror and torture, cruelty and necessity. Few lawyers are good at that sort of conversation.
Nonetheless, I am not making news when I say that formalism has limits, as well as virtues, and these limits are quite quickly reached when the subject is the Geneva Conventions. Let me say a bit about what those limits are - and what it would take to somehow move beyond them.
The Geneva Conventions were drafted in 1949, in another world. The world of the Geneva Conventions' "framers" is still familiar to all of us, though increasingly it is familiar from movies and books rather from the evening news or, still less, our own lived experience. The world in which the Conventions were drafted was a world of states: powerful states, weak states, predatory states, law-abiding states, but states all the same. Soldiers wore uniforms designed by their states, carried weapons issued by their states, obeyed orders given by their commanders, and fought against the armies of other states.
Well - most of the time, anyway. It's true that even then, there were actors and conflicts that didn't fit the mold. There were partisans who wore no uniforms and answered to no recognized authority, and guerillas and resistance fighters who straddled the line between civilian and combatant. But although it is sometimes hard to make students see this, lawmaking is an imaginative enterprise: lawmakers look at the existing world, project onto it an image of a better, tidier world, and then try to develop contingency plans for dealing with various imaginable forms of untidiness. In this sense lawmaking is inevitably backward looking, because none of us is very good at correctly predicting future changes.
So the diplomats who negotiated the Geneva Conventions took the raw materials already at hand, from the Hague Conventions and from international custom, and coupled these with their own searing sense of what had gone wrong in the world war just ended. In a sense, the Geneva Conventions read like an attempt to revisit the Second World War, without the mess, confusion, cruelty, and slaughter of civilians. This is not surprising.
Inevitably, the Geneva Conventions were "out of date" from the moment they entered into force; they laid out rules for a world more orderly than the world they had inherited, and hoped that by doing so, they would encourage life to imitate art.
Up to a point, it worked. The Conventions have been normatively important. They have led powerful states to integrate Convention rules into their own domestic law, and they have provided an important tool for shaming parties to conflicts into behaving better than they might otherwise.
But the Conventions were always aspirational, and since their entry into force, human ingenuity has devised new and different ways to fight and kill. We now fear the terrorist's bomb, anthrax in the mail system, sabotage of critical infrastructure, or a lethal virus released deliberately as much as we fear an invasion by a powerful state.
The threats we face today are not necessarily "worse" than the threats we used to face. It is important to emphasize this. There is no satisfying way to quantify the risks we face today and compare them to the risks we faced three or four decades ago. Yes, a "dirty bomb" in New York could be catastrophic, potentially killing thousands and making a major city uninhabitable. Is this "worse," though, than the Cold War risk of nuclear war? Worse than the risk of ethnic slaughter, exemplified by the Holocaust?
Still, the increased threat of terrorism, though perhaps not a "worse" threat than any prior threat, is certainly a "different" threat in crucial ways. Guerillas and terrorists have always existed, and never fit neatly into the Geneva Conventions framework, but they operated on the margins until globalization scattered the tools of mass destruction around the world. The hijacked airplane, the simple materials to make a hundred or a million IEDs, the cell-phone detonators, the viruses in test-tubes: these are new. Globalization has turned the marginal, nuisance threat of terrorism into a threat that even powerful states must take seriously.
Even in powerful states, intelligence services, militaries and laws all evolved to handle "traditional" conflicts and traditional threats from belligerent foreign states. Yet terrorists - like other non-state actors - are, by definition, not party to the Geneva Convention. They play by a different set of rules - if indeed there is any set of rules they follow.
As a result, the formal framework of the Geneva Conventions does not fit the struggle against terrorism well. Too many of its threshold distinctions are premised on the continued existence of a rapidly vanishing world.
The Geneva Conventions take it for granted, for instance, that we can draw meaningful spatial distinctions between zones of conflict and zones of peace, but this breaks down when the enemy is a geographically diffuse terrorist network, neither confined to one state nor interested in controlling territory. If al Qaeda has a secret cell in Yemen - or in Germany, or in the United States - from which it plans and trains for terrorist attacks, is Yemen (or Germany, or the United States) in a conflict zone? The Geneva Conventions offer no way to answer this question.
Temporal boundaries between war and peace, as well as spatial boundaries, are challenged by the rise of non-state actors. The Geneva Conventions assume a world in which diplomacy and negotiations can bring an end to a conflict, but with loosely organized terrorist organizations, there is often no one with whom one could negotiate, and no one with the authority to bring about a peace. Attacks may be constant, or intermittent, and the Geneva Conventions don't offer helpful standards for determining when a conflict begins or ends in the absence of agreements between parties to the conflict.
Most troublingly, given recent events, the Geneva Conventions don't offer satisfying answers to the question of which people are entitled to benefit from its protections. On the one hand, the text is clear that the Conventions apply only to "High Contracting Parties," which, by definition, are states. n1 On the other hand, we have Common Article 3, n2 which may - or may not - be taken to apply to all parties to all conflicts not otherwise covered, regardless of nationality and state allegiance. And even when we're confident that the Conventions apply to a conflict within a territory, what do we do about combatants who wear no uniforms, or don't appear to work within a traditional military command structure? A strict reading of the Geneva Conventions suggests that the Bush Administration is not unjustified in its claim that neither al Qaeda nor Taliban detainees captured in Afghanistan are entitled to prisoner of war status.
I have written extensively about these issues elsewhere, so I won't run through additional examples here. n3 The point is that an entirely formalist read of the Geneva Conventions leads to the conclusion that the Conventions just aren't applicable, for the most part, to the "war on terror."
To human rights advocates (of whom I am one), this is an unpalatable conclusion, for it appears to suggest that there are virtually no rules of international law governing how the war on terror is waged. It suggests that the United States is legally entitled to offer few or no protections - procedural or substantive - to those it suspects of being terrorists, and in fact can treat suspected terrorists in a manner that would be clearly illegal in both a domestic criminal context and in a traditional armed conflict context.
Most rights advocates adopt one of two strategies in reaction to this dilemma. The first strategy consists of agreeing that a formalist read of the Geneva Conventions makes them hard to apply to terrorism, but shrugging this off on the grounds that terrorism is not a form of armed conflict at all, and is merely criminal activity. Obviously, if terrorism is simply a form of crime, and not a form of armed conflict, it is subject not to the Geneva Conventions but to domestic criminal and constitutional law, which offer relatively robust protections for suspects.
The second strategy employed by many rights advocates consists of what one could call "modified formalism": it consists of accepting that terrorism is a form of armed conflict, but arguing that the Geneva Conventions should be interpreted less like a statute than a constitution. That is, a treaty should be interpreted according to its "spirit" and according to the "intent" of its framers, which in this case was to protect fundamental human rights during armed conflicts. A modified version of this argument is that some of the substantive aspects of the Geneva [*201] Conventions are customary law, and that customary law imposes additional obligations on detaining powers beyond those outlined in the Conventions. Both variants of this theory can then be used to justify demanding POW-like protections even for those detainees who most clearly lack any entitlement to them on a strict formalist reading of the Conventions.
Neither approach overcomes the problems I have discussed. The first has a head-in-the-sand quality, since international terrorism is different from ordinary crime in significant and obvious ways, and it seems clear that traditional criminal investigations and trials are not an adequate means of combating terrorists who seek to cause mass death and who operate in many different countries. For the most part, serious rights advocates have abandoned this line of argument, and acknowledged that at least some of the time, if not all of the time, the activities of terrorist groups look more like armed conflict than crime.
The second approach is better, but it is also problematic. First, it raises unresolved questions about treaty interpretation. There is no clear legal basis for insisting that the Geneva Conventions be read according to their "spirit" rather than their letter, and in any case it is far from clear that the framers of the Convention would have chosen to accord terrorists most of the rights given to POWs, had they foreseen international terrorism of the sort we now face. There is also little consensus about which aspects of the Geneva Conventions reflect (or have become) binding customary international law. Certainly core provisions, such as those that prohibit the intentional targeting of civilians, reflect customary international law - but more subtle questions of due process are less easily resolved by reference to custom.
This means, among other things, that Attorney General Alberto Gonzales was not wrong when, during his days as White House Counsel, he advised President Bush that some provisions of the Geneva Conventions seem "quaint" or even "obsolete" when applied to terrorists. n4
It is difficult - almost impossible - to advance this proposition without generating enormous opposition from most rights advocates, but I think that the opposition is knee-jerk and misplaced. Acknowledging that the Bush Administration's read of the Geneva Conventions is not implausible does not require agreement with the Administration's policies. It is entirely possible to accept that the Geneva Conventions don't apply to terrorist suspects, but still consider the Administration's detention and interrogation policies both morally bankrupt and strategically foolish.
The Geneva Conventions should not be sacralized. They are important, relevant, and often useful. But if we resist Bush Administration policies by treating the Geneva Conventions as sacred cows, we do the long-term cause of promoting human rights a disservice. The Conventions are not perfect - how could they be? They are the product of a time and a place, of divisive negotiations between human beings who sought to represent the interests of their states.
Indeed, before the "war on terror" rallied rights advocates around the Conventions, some in the human rights community were themselves attacking the Convention framework for its inadequacies, arguing that the Geneva Conventions actually legitimize (and - by implication - can worsen) the violence of states. These are important critiques, and we need to clear enough intellectual space to have them.
The only way to do that is to stop treating the Geneva Conventions as our sole or most crucial point of reference. To be sure, the Geneva Conventions are the law we've got, so it is useful to ask what the Geneva Conventions require, what they permit, and what issues they just don't address; it's also useful to identify areas of consensus and discord about how and when they are applicable. But if we value the rule of law, this needs to be done in an honest and disinterested way. This means facing up to it when the Conventions are silent or seem archaic, and acknowledging it when the questions we face are primarily policy questions, not legal questions.
In practice, this does not happen much. In the debate about post-9/11 U.S. practices, few of the participants are willing to move beyond formalism or faux formalism of one sort or another. As I suggested at the beginning, perhaps this is because we find the policy discussions too difficult and emotional. But it is dangerous to avoid them.
When we refuse to admit the limits of formalism, we inevitably have to start stretching rules and ordinary meanings. And if rights advocates adopt a rigid formalism, it becomes difficult to focus clearly on important differences between various Administration arguments, which in turn makes it hard to develop effective rejoinders.
Take three different examples.
First, consider the Bush Administration's decision to deny POW status to Taliban detainees. As noted above, a strict formalist read of the Geneva Conventions makes the Administration's decision seem perfectly plausible, though hardly inevitable. It would be possible to argue that the Administration was just wrong, however, either on the facts or on the law; one could argue, for instance, that most Taliban soldiers wore distinctive emblems and otherwise met Geneva requirements, or that at any rate some did, and therefore individualized status hearings were required.
Second, consider the Administration's arguments relating to torture, contained in the Bybee Memo of August 1, 2002, which concluded that under federal law, "torture" must involve only the sort of pain associated with organ failure and death. n5 The response here would be that Administration lawyers were not simply mistaken about the conclusions warranted by statute, treaty, and case law, but that they were engaging in illegitimate and arguably unethical forms of legal argumentation, ignoring and selectively misreading various relevant texts in order to reach a predetermined conclusion.
Third, consider the arguments, also in the Bybee memo, claiming that the President has the inherent constitutional power to override conflicting federal law when he deems it necessary. n6 These arguments are legal in their form - they rely on the President's commander in chief powers - but their implications go beyond law, insofar as they assert, in some fundamental sense, that political power is simply beyond law's reach.
In each of these three cases, Administration arguments are couched in the language of formalism. Everyone wants the law on their side, so this is understandable. But the kinds of arguments are fundamentally different.
Think of law - of the enterprise of legal interpretation - as a game, like basketball or tennis. The game of legal interpretation has rules, some written, some customary, some bright-line, some ambiguous. But although it is difficult to say just what elements make tennis tennis, we all know that there is a difference between playing tennis in a way that pushes the envelope between the permissible and the impermissible, cheating, and leaving the game. Thus: calling a ball "in" when it just touches the outside of the baseline is skirting the edge of the permissible, but it is clearly within the rule. Calling a ball "in" when one knows it to have landed outside the baseline is cheating, but it is still playing tennis: it is just cheating at tennis. Pausing to beat up one's opponent when he objects to cheating is no longer tennis, however; the resort to force destroys the game entirely.
Map this onto the three different Administration arguments I mentioned above. The Bush administration claim that Taliban detainees are not entitled to POW status or individualized hearings on the question may be wrongheaded in a strategic or moral sense. But if we think of law as a game, the Administration is clearly playing by the rules here, though perhaps pushing boundaries a bit.
The claim about torture relies on cheating, insofar as it depends on selective and misleading citation and odd logical leaps. Nonetheless, cheating, however reprehensible, is a way to play the game; by definition, if you are cheating at a game you are still accepting most aspects of the game itself.
But the claim about inherent executive powers is of a different order. Though couched in formalist terms, it's a game-ending move, the rough equivalent of a threat: "Play by my rules or I'll crush you."
Although law is "gamelike" in many respects, it is, of course, crucially different from tennis. The "rules" of law and legal interpretation are not there for the entertainment of the players; they're not merely self-referential. Law is supposed to bear some relation to facts on the ground, and law enables coercive action to be taken in ways that alter the facts on the ground. If we create a legal system in which cheating is widespread - or, worse, if we overlook game-ending moves and treat them as legitimate modifications of the game - then it isn't merely rules that get bent, but the rule of law altogether.
This is what makes it so important to move beyond the sacralization of the Geneva Conventions to a more particularized discussion of just what morality and strategy require. As long as we insist that the Geneva Conventions and related law and custom set the terms of the debate about post-9/11 U.S. actions, we create strong incentives for players to push the envelope, cheat, and even go outside the law altogether in one way or another. This dangerously weakens the rule of law.
The Geneva Conventions - what they permit, what they require, and what they do not reach - should thus be the beginning of the discussion, not the end. If we're serious about both rights and security, we should make an effort to get back to basics: what sort of world do we want to live in? Knowing that terrorism will persist, and that technological development will continue to give tools of mass destruction to non-state actors, what principles ought to govern how terrorist suspects are treated, and how states respond to real or perceived threats from terrorists?
This requires thinking both about morality and about strategy. And the process, if taken seriously, might help break through impasses and allow us to identify areas of agreement that are obscured when the Geneva Conventions dictate the terms of debate.
Imagine, for instance, that we live in the "all security, all the time" world, in which collective physical security always outweighs individual rights. There is no free expression, no freedom of movement, no right to due process. None of us would want to inhabit so totalitarian a state - and most of us would wonder, in any case, if "perfect" security is even a realistic goal. We do not strive for "perfect" road safety, because we value convenience, speed, and a relatively low level of state monitoring, and because we probably could not achieve perfect safety in any case. Security from terrorist threats is not inherently different; at some point one realizes diminishing returns on further rights restrictions.
Of course, few of us would want to inhabit an "all rights, all the time" state either, one in which state authorities entirely lacked the power to adapt law and policy to meet new kinds of threats if doing so meant longer detentions or more limited due process rights.
Imagining an "all security, all the time" world versus an "all rights, all the time" world helps get us past the tendency to assume that security must trump rights or rights must trump security, and recasts the questions as one about precisely which tradeoffs are worthwhile. But even this is still a bit misleading, since conceptualizing the debate in terms of a tradeoff between rights and security obscures the possibility - indeed, the strong likelihood - that "security" and "rights" are causally linked.
That is: it is quite possible that some rights-restricting U.S. actions in the war on terror have actually increased the terrorist threat against the United States. Our policy of open-ended detentions at Guantanamo, for instance, justified by the Bush Administration on security grounds, has alienated even many of our allies. Accept, for the sake of the argument, that at least most, if not all, of the Guantanamo detainees are dangerous terrorists who, if released, will continue their efforts to attack the United States in some way. A level-headed policy maker needs to take this prospect seriously, and critics of Guantanamo Bay should not assume that all detainees are innocent or harmless. But a level-headed policy maker should also take into account the externalities of Guantanamo. These are difficult to measure, but might include some or all of the following: greater difficulties in persuading "friendly" states to share intelligence information, greater difficulties in persuading ordinary people in Iraq, Afghanistan and elsewhere to trust U.S. forces and share information; and greater difficulty discouraging some from actively supporting al Qaeda and other anti-U.S. organizations.
All these externalities create new security risks that must be evaluated alongside the risks of releasing potentially dangerous terrorists. Note, too, that the "moral" and "strategic" arguments prove to be intertwined. Regardless of how U.S. policy makers view the morality of open-ended detentions, they may have serious costs to the United States as long as many others view them as immoral.
Of course, one can, and should, turn this around as well. High levels of physical insecurity make the enjoyment of other human rights difficult or impossible, and terrorism is itself a human rights violation of enormous magnitude. If a rigid insistence on procedural due process rights for terror suspects led to a massive increase in catastrophic terrorist attacks, we would not be better off. The often-repeated claim that it is our very openness that makes us vulnerable is not entirely frivolous.
All this implies, I think, that we need to evaluate the various kinds of threats and potential responses with some specificity. Certain restrictions on rights - limited in duration - might well be justifiable in the face of an imminent catastrophic threat (nuclear weapons, for instance). But the phrase "the war on terrorism" lumps together, by implication, a wide range of different threats, most of which would not justify serious restrictions on rights. The captured Taliban soldier, for instance, however resolutely anti-American, will probably do the United States no great harm if released - even if he promptly rejoins the Taliban. Osama bin Laden, if captured, would present a different story. Similarly, if we are concerned about biological attacks on the United States, it may be that massive investments in our dysfunctional public health system are a more effective response to this threat than weakening norms against torture through the use of various "coercive" interrogation techniques.
It's beyond the scope of this short essay to discuss concrete means of dealing with the variety of security threats we now face. My point is that we need to walk back from the panicky, post-9/11 sense that anything goes in the name of fighting terrorism. We need to develop responses to terrorism that are nuanced and proportionate, and acknowledge both that the varying forms of terrorism do not pose equal threats and that every response has costs of its own.
That will be a monumental undertaking, but not a hopeless undertaking. The work of several of the participants in this conference makes valuable contributions to the project I have outlined. To some extent, we're facing empirical questions: how severe are the various threats, and what can be done about each? Lawyers may be good at formalism, but increasingly, many of us are also good at making cost-benefit analyses, and a bit more of this would be useful here. So too would be looking to the experience of other nations that have addressed terrorist threats through domestic legislation. While other nations have at times restricted certain rights to combat terrorism, none have found it necessary to turn to torture or indefinite detention. It seems likely that we could learn from a careful study of the effectiveness of different domestic legal regimes.
Paul Stephan opened this conference by asking, "Who owns the Geneva Conventions? Who gets to say what is in them, and should be in them?" My answer, on some level, is: why not us? Why not begin the process of imaging a new law of armed conflict, and new domestic laws, right here and right now?
It's easy to respond that this is too hard, and will take too long, and opens up too many cans of worms. But this is a cop-out. It may be a hard, long, and divisive process, but we certainly won't be able to develop a new consensus if we don't try.
FOOTNOTES:
n1. E.g., Geneva Convention Relative to the Treatment of Prisoners of War, Aug. 12, 1949, art. 1, 6 U.S.T. 3316, 3318, 75 U.N.T.S. 135, 136 [hereinafter GPW].
n2. E.g., id. art. 2, 6 U.S.T. at 3318-20, 75 U.N.T.S. at 136-38.
n3. Rosa Ehrenreich Brooks, War Everywhere: Rights, National Security Law, and the Law of Armed Conflict in the Age of Terror, 153 U. Pa. L. Rev. 675 (2004).
n4. Memorandum from Alberto R. Gonzales, White House Counsel, to the President (Jan. 25, 2002).
n5. Memorandum from Jay S. Bybee, Assistant Attorney General, to Alberto Gonzales, Counsel to the President (Aug. 1, 2002).
n6. Id.