U.S. Department of State
U.S. Department of State
Other State Department Archive SitesU.S. Department of State
U.S. Department of State
U.S. Department of State
U.S. Department of State
U.S. Department of State
U.S. Department of State
U.S. Department of State
Home Issues & Press Travel & Business Countries Youth & Education Careers About State Video

Oxford Leverhulme Programme on the Changing Character of War

John B. Bellinger, Legal Advisor
Lecture at the University of Oxford
Oxford, England
December 10, 2007

As Delivered

As many of you may be aware, I have been engaged over the last three years in extensive bilateral and multilateral efforts to discuss developing a common legal approach regarding combating transnational terrorism. As part of this effort, I have traveled to a dozen countries, engaged in seven rounds of discussions with the legal advisers of the 27 EU countries, and held additional discussions with the legal advisers of the member states of the Council of Europe. I have also participated in numerous panels and roundtable discussions on the matter with legal experts in this area.

Some of my work has been retrospective, in which I have tried to explain to our allies the actions we took after the September 11th attacks. I have acknowledged that one of the mistakes the United States made after 9/11 was not discussing with our allies the reasoning and legal basis behind the steps we took to combat al Qaida. A little more than a year ago, I gave a speech at the London School of Economics in which I gave a comprehensive public explanation of our legal views and policy decisions with respect to the detention and treatment of terrorists, as these have evolved in the United States since September 11th. In that speech I explained the legal basis for various decisions the United States took after 9/11, and then asked critics to consider what realistic alternatives existed to our approach, and whether those alternatives are legally mandated or are simply among the acceptable alternative options available to policy makers. In retrospect, we might well have handled some matters differently as a matter of policy, but that does not mean our approach was flawed as a matter of law. The bottom line, as an increasing number of legal experts now acknowledge, is that the legal framework for conflicts with transnational terrorists like al Qaida is not clear.

Rather than continue to look back, however, tonight I would like to focus more prospectively on whether international humanitarian law in general, and the Geneva Conventions in particular, provide a satisfactory set of rules for contemporary conflicts. I am not advocating that we discard existing rules, which serve a critical role in dealing with the situations for which we developed them. Nor am I straining to find gaps in the existing legal framework in order to place detained persons in a legal black hole. The gaps are real, and recognizing this fact does not mitigate the obligation of States to comply with international law, nor does it justify placing persons beyond the protection of the law. My key point tonight is that the Geneva Conventions were designed for traditional armed conflicts between States and their uniformed military forces, and do not provide all the answers for detention of persons in conflicts between a State and a transnational terrorist group.

Common Article 2 of the Conventions restricts the scope of applicability of most of the Conventions’ provisions to conflicts between High Contracting Parties. But as we are seeing throughout the world, contemporary conflicts often do not have more than one High Contracting Party to the Conventions involved. Some of these conflicts occur within the boundaries of a country, like Sri Lanka’s conflict with the Tamil Tigers. But more and more the conflicts cross national boundaries, like Israel and Hezbollah or the ongoing conflict between the U.S., its allies and al Qaida. In cases such as these, we are left in a situation where Common Article 3, and depending on a State’s treaty obligations and the nature of the non-state actor, Additional Protocol II, provide the only treaty-based rules governing detention of unprivileged combatants.

I must note here that it was not always clear to our government that Common Article 3 applied as a treaty-law matter to a conflict between a State and non-state actors that transcended national boundaries. While the U.S. Supreme Court decision in Hamdan v. Rumsfeld held that the conflict with al Qaida, as one not between States, is a non-international conflict covered by Common Article 3, I think many international legal scholars would question that conclusion. Textually the provision is limited to armed conflict “not of an international character” occurring “in the territory of one of the High Contracting Parties,” suggesting the scope of the provision is limited to conflicts occurring in the territory of a single state. Indeed, other states, such as Israel, have concluded that conflicts with terrorist organizations outside the State’s borders are international armed conflicts not falling within the scope of Common Article 3. I make these points not to re-litigate the Hamdan case, or to disregard the view of many that Common Article 3 is customary international law, but rather to note that in some cases, not even Common Article 3 may apply as a treaty-law matter to conflicts with transnational terrorist groups.

But even assuming that Common Article 3 does cover contemporary transnational conflicts of this sort, I think it is striking just how little guidance Common Article 3 in fact provides. The one area where Common Article 3 does provide good detail is with respect to the treatment of detainees once in custody. Treatment protections include the prohibition against torture and cruel, humiliating and degrading treatment, and a requirement that those criminally tried in relation to the conflict be provided judgment by “a regularly constituted court affording all the judicial guarantees considered essential by civilized peoples.” Depending on a State’s treaty-law obligations these treatment protections can be supplemented in certain circumstances by Additional Protocol II. Many would also argue that Article 75 of Additional Protocol I provides other relevant protections as customary international law applicable in non-international armed conflict.

But quite clearly, the meaning of particular treatment protections may be subject to different interpretations. Common Article 3 was not designed with the precision of a criminal statute. Indeed, the International Criminal Tribunal for the Former Yugoslavia acquitted a defendant of violation of Common Article 3’s prohibition on “violence to life and person” because the term lacked a sufficiently precise definition under international law. The U.S. has also wrestled with how to implement this article in our criminal law, especially since the Hamdan court ruled it governs our operations in the conflict with al Qaida. For example, “outrages upon personal diginity” is defined in Pictet’s Commentary on CA3 states as capturing only those acts that “world public opinion finds particularly revolting.” But reasonable people can and do differ about what behavior that phrase captures. It was this concern that led the Administration and Congress to agree in the Military Commissions Act to amend the War Crimes Act to clarify which specific violations of Common Article 3 are criminally sanctionable.

More important, though, Common Article 3 does not address at all four central questions that I believe must be answered with respect to conflicts with non-state groups. I want to discuss each of those questions this evening. First, who may a State detain in a conflict with a global non-state actor? Second, what processes must a State provide detainees to determine whether they can be detained? Third, when are hostilities over in armed conflict with a non-state group? And fourth, what legal obligations do States have in connection with repatriating detainees at the end of the conflict?

The Gaps Are Not Already Filled

A response I have frequently heard to these questions is that we are looking in the wrong place for their answers. Critics respond that other treaties or customary international law fill these gaps. It is not clear, however, that they do.

First, some argue that 1977’s Additional Protocol II of the Geneva Conventions was designed to address the limited scope of Common Article 3 by providing additional rules for non-international armed conflict. President Reagan submitted Additional Protocol II to the Senate seeking advice and consent to ratification in 1987, but the Senate has not acted on the treaty to date, meaning its provisions do not bind the United States as a treaty law matter. But even for States that have become party to AP II, such as the United Kingdom, the Protocol does not provide a satisfactory answer to the questions I just posed. While AP II expands on the treatment protections provided in Common Article 3, it has a more limited scope of application defined in Article 1, and its provisions do not squarely address any of my four questions.

Second, some have suggested that customary international law can be used to fill gaps in treaty law. As I just explained, the conclusion that Article 75 of Additional Protocol I is customary international law applicable in all armed conflicts would add to the treatment protections in non-international armed conflict provided by Common Article 3. But as a general matter, States need to be careful to adhere to proper methodology before describing particular provisions of treaty law as custom. Many commentators assert customary international law as they would like it to be, rather than as it actually is. The U.S. Government sent a letter to the ICRC President Dr. Kellenberger noting concerns with the methodology employed by the ICRC IHL Customary International Law Study in deeming treaty provisions customary international law. Although it may seem attractive as a policy matter to import rules developed in international armed conflict to other situations, we must be careful not to describe rules as custom when there is an insufficient basis to do so. Providing unprivileged combatants the same or greater protections and rights as those provided prisoners of war risks rewarding illegal actions, ultimately placing innocent civilians at greater risk.

Third, human rights groups and some European states argue that human rights law fills the gap wherever IHL is insufficiently specific to address a particular situation. It is important to remember here that States have different obligations under different treaties. U.S. obligations under the International Covenant on Civil and Political Rights only apply in U.S. territory, while Europeans States are parties to human rights instruments with protections that extend outside national borders. So when we talk about human rights law, we need to be sure we are taking into account different national circumstances.

But even where States do have human rights obligations, it is fair to ask proponents of this approach what particular human rights provisions they would apply to activities arising in the conduct of armed conflict, and how they would apply them in practice. For example, Article 9 of the ICCPR requires States to provide anyone detained the right to bring their case before a judge without delay to determine the legality of the detention. Would it be practical to expect States detaining tens of thousands of unprivileged combatants in a non-international armed conflict to bring them before a judge without delay? This is not something States must do even for prisoners of war under the Third Geneva Convention. If the answer is that the State should derogate from Article 9 if the exigencies of a civil war so demand, then what contribution has human rights law made to answering questions regarding the procedures owed combatants in non-international armed conflict? Some rights deemed non-derogable by the ICCPR, such as the right to life, would be clearly displaced by more specific law of war rules that govern as the lex specialis.

In the end, I think the gaps I have identified in the rules regarding detention of combatants in non-international armed conflict are real, and that simply labeling international armed conflict rules custom in non-international armed conflict or importing human rights law does not satisfactorily resolve these difficulties. Through the course of my dialogue, more and more Europeans have been willing to acknowledge that the existing rules were not designed for, and are in fact not well-suited for, the threat posed by transnational terrorism. For example, earlier this year the Foreign Affairs Committee of the UK House of Commons wrote that the Geneva Conventions dealt inadequately with the problems posed by transnational terrorism, and called on the British government to work with other States and the ICRC on updating these Conventions for modern problems. Although I think it is premature to talk about negotiating a new international instrument, I am pleased to see that more people are beginning to think about whether the challenges terrorism poses to the law of war requires more than just calling for more robust implementation of existing rules.

Detention Scope and Procedures

Having established that the issues I have identified with Common Article 3 are not easily resolved by resort to other treaties or customary international law, I want to explore each of the four major unaddressed issues in turn. These issues are not an exhaustive list of areas where further dialogue and legal development are needed, but are perhaps the most important issues I have faced as Legal Adviser. The first question is how States should define the category of persons that can be detained in non-international armed conflict. With respect to combatants, traditional international armed conflict has a relatively easy answer: a State detains enemy forces, who usually wear uniforms, are in clear command and control structures, and conduct their operations in accordance with the laws of war. But in the contemporary conflicts we are discussing tonight, determining the legal contours of the category of “combatant” can be extremely difficult.

Clearly, Taliban militants captured on the battlefield in Afghanistan, as many of those at Guantanamo were, would fall within the scope of persons that can be detained for the duration of hostilities. So too would an al Qaida terrorist in Iraq with a strapped-on suicide vest headed to a civilian area to detonate. But what about the person who made the explosive-laden vest? The financier whose money laundering for al Qaida made the suicide operation possible? The religious leader who knowingly inspired the suicide bomber to embark on his mission? This issue has been a difficult one for the United States with regards to al Qaida, and has been a source of tension with our European allies, some of whom are concerned that our definition of combatant is over-inclusive. But where exactly to draw the line here is unclear. Although it may seem reasonable to say that only those like the suicide bomber or vest maker should be detained as combatants, it may be the financier’s broad operations that in fact pose the greatest threat to a State.

Of course, the law of war envisions that a State will detain both combatants and civilians during armed conflict. The laws of war have long permitted the detention of supporters of hostile forces during armed conflict, including civilians connected to armies such as laborers, messengers, guides, scouts, and civilians transporting military supplies and equipment in proximity to the battlefield. Article 42 of the Fourth Geneva Convention clearly contemplates security internment of protected persons, “where the security of the Detaining Power makes it absolutely necessary.” The Israeli Supreme Court in the Public Committee against Torture case concluded that combatants not in regular armies or militias that meet the requirements of Article 4(A)(2) of the Third Convention were in fact civilians, who lost their comprehensive protections against attacks, “for such time as they take a direct part in hostilities.”

It’s worth noting here that the term “direct part in hostilities” in Article 51, paragraph 3, of Additional Protocol I, has been a difficult phrase to define. For years, a group of forty law-of-war experts have grappled with this issue in a series of expert meetings co-organized by the ICRC and the TMC Asser Institute. Although the experts’ work is not finished, I am aware that it delves into these difficult questions, and I look forward to reviewing the report. More centrally though, query what the relevant differences are between categorizing some as unprivileged combatants (e.g., al Qaida) and other civilians who may be the object of direct attack but only for such time as they take a direct part in hostilities. In each case, a State can detain these persons for the duration of the conflict, and must treat individuals involved in a non-international armed conflict consistently with Common Article 3

This question of whom a State may detain relates to the second major question I want to discuss: what procedures must a State use before deciding someone may be detained in non-international armed conflict. In international armed conflict, normally no process is used to determine whether or not soldiers from the opposing army may be detained. Such detained combatants, usually prisoners of war, who are not criminally charged are not entitled to counsel or judicial review. After 9/11, we took the view that Taliban and al Qaida militants we picked up on the battlefield were subject to detention under the law of war. As with traditional conflicts, these combatants were not provided lawyers nor afforded judicial review of the legality of their detention. But while this practice may make sense with respect to clearly identifiable soldiers, how should a State decide whether to detain non-state actors who often lack identifiable indicia of being a combatant? Is it sufficient to treat them as the law of war treats traditional combatants, or does something about their non-traditional status make further process necessary?

The U.S. Supreme Court clearly felt uncomfortable with applying the traditional rules to these unprivileged combatants. In its Hamdi decision in 2004, the Court ruled that US citizens picked up on the battlefield and detained in the United States are entitled to an administrative review process to determine whether they are in fact combatants. And in the companion Rasul decision, the Court extended statutory habeas corpus rights to alien detainees held at Guantanamo. The issue in last week’s Boumediene argument was whether the right to common law habeas corpus protected by the Suspension Clause extends to the Guantanamo detainees. Ultimately, the United States appears to have arrived at a place where it is unquestioned, as a general matter, that administrative review of combatant status, and often subsequent judicial review of the legality of detention, accompanies extended detention in non-traditional conflicts.

It may be that we may have arrived at rules not that different from the rules set out for internment of civilian Protected Persons in Article 43 of the Fourth Geneva Convention. That article states in part, “Any protected person who has been interned or placed in assigned residence shall be entitled to have such action reconsidered as soon as possible by an appropriate court or administrative board designated by the Detaining Power for that purpose.” We are following this procedure in Iraq, where we adhere to the Fourth Convention as a policy matter with civilian security internees. But we also meet this standard with combatants who are detained at Bagram in Afghanistan or Guantanamo. While I continue to question whether it makes sense to classify al Qaida members as civilians as opposed to unprivileged combatants, as the Israeli Supreme Court and others have suggested, the added procedural protections afforded interned or detained civilians may provide a model for appropriate rules for the detention of unprivileged combatants.

End of the Conflict?

Along with these two questions surrounding initiation of detention, Common Article 3 and other applicable IHL do not provide clear answers to two questions regarding termination of detention in contemporary conflicts. Even if one acknowledges that al Qaida militants may be lawfully detained as unprivileged enemy combatants, when must detained persons be released? Again, traditional IHL principles provide a simple answer: upon the cessation of active hostilities. In traditional conflicts it is obvious why this is the case. Could anyone imagine Allied forces during World War II releasing before the end of the conflict German soldiers who could return to the fight? And in the U.S. conflict in Viet Nam, captured U.S. military personnel were held by the North Vietnamese for up to nine years without any idea as to when they might be released or repatriated. At the same time this answer seems deeply unsatisfactory to some in the current conflict with al Qaida. Critics ask fair questions when they query how the United States will identify the end of hostilities. Although it would have been difficult for those living in Blitz London to identify when hostilities would have ended, at least there was a sense of what an end to the conflict might look like. It is highly unlikely this conflict will end with the signing of a formal surrender document on a battleship.

But what are the consequences of the conclusion that it will be difficult to identify when the conflict may end? Does this mean we should just release everyone we are holding now? This option is unpalatable given that many of the people we would release would immediately return to the fight. The Defense Department believes that more than 30 released Guantanamo detainees have already returned to the fight. Presumably, releasing the more dangerous individuals still detained at Guantanamo would result in an even greater number of recidivists. Or could it mean that we must after some period of time States must release the detainees or subject them to trial? I have in the past given lengthy explanations of the difficulties Western legal systems have faced in criminally prosecuting terrorists-- from the challenges posed by extraterritorial and retroactive legislation to difficulties in collecting admissible evidence in battlefield and intelligence settings.

The better answer may be to conceptualize the end of the conflict differently, possibly looking to principles found in the Fourth Geneva Convention. Article 43 of the Fourth Convention contemplates twice-yearly reviews of security internment of protected persons by a court or administrative board. In situations where the end of the conflict is as uncertain as it is with our conflict with al Qaida, administrative reviews could be used to determine whether the conflict has ended as to a particular detainee. Two leading legal experts, Curt Bradley and Jack Goldsmith, have written on this point, arguing that the unique characteristics of the war on terrorism require an individualized determination on end of the conflict. They suggested that such a determination could take into account the detainee's past conduct, level of authority within al Qaida, statements and actions during confinement, age and health, and psychological profile.

At Guantanamo, we have implemented annual Administrative Review Boards, or ARBs, in which a panel of military officers considers whether an individual detainee can be released or transferred in a manner that would not threaten the security of the United States or its allies. In a sense, this is an assessment of whether or not the conflict has can be viewed as having been ended with respect to the detainee in question. Perhaps we should consider what changes to the ARB process might be warranted to pursue this concept further.

The fourth and final question I want to address this evening is what should be done with detainees we no longer have a reason to hold in these non-traditional conflicts. Common Article 3 does not answer this question. The Third Geneva Convention offers a simple answer with respect to Prisoners of War. Article 118 states, “Prisoners of war shall be released and repatriated without delay after the cessation of active hostilities.” Traditional state practice has been to return these detainees to their States of nationality. But although this traditional rule has been easier to apply in conflicts involving a limited number of States, it becomes far more challenging to apply when there are nationals of many States involved in the conflict. At Guantanamo, for example, we have detained nationals from more than forty countries. This has raised numerous practical problems. Rather than negotiate one bulk repatriation, as envisioned in Article 118, we have been forced to negotiate separate agreements with every country whose nationals we detain in the conflict. Needless to say, this has delayed the repatriation process significantly.

This problem grows in magnitude when the detainees we wish to repatriate express fears of mistreatment or persecution upon return. Although this is not a new problem, Article 118 is conspicuously silent on what States should do when those they wish to return do not wish to go back due to their concerns about treatment upon return. In World War II many thousands of Soviet nationals who had taken up arms for Germany, and who expressed fears of returning to the Soviet Union were forcibly repatriated by the US and UK in compliance with the 1945 Yalta Agreement. Christine Shields Delessert’s good book on this topic details the brutal treatment these prisoners received after being returned to Soviet custody, including relocation to forced labor camps in Siberia and in some cases execution. In Korea and again in the first Gulf War, Allied forces used a different approach with prisoners not wishing to be repatriated, eschewing forcible repatriation in favor of third-country resettlement. In the current conflict with al Qaida, the United States has looked to human rights law as a non-binding guide for determining when to repatriate prisoners to third countries, establishing the firm policy not to turn over detainees where it is more likely than not they will be tortured. This policy, central as it is to Western values, has meant that dozens of detainees who cannot be repatriated, such as the Uighurs to China, have remained at Guantanamo for years after we have wished to transfer them. This is an area where the U.S. has asked for assistance from its European partners and other allies to assist in the humanitarian resettlement of these individuals.

I would suggest that this problem is only likely to grow. In the conflict with al Qaida, for example, the majority of detainees are nationals of countries with poor human rights records. The problem is even more acute than in traditional armed conflict, because these governments are often harshest towards the very group of citizens that are being detained— people considered to be terrorists. This is less true when those being repatriated are a State’s own soldiers. Exacerbating the problem is the lack of available third countries to resettle those detainees expressing credible fears. Unlike in previous conflicts when those detained may have had no ideological disagreement with the detaining power beyond the current conflict, and who may be expected to live peaceful lives once re-settled, terrorists such as those at Guantanamo have the training and ideological desire to pose a continuing threat once resettled. Not surprisingly, third countries, including the United States, have not been willing to accept this risk.

Ultimately, I would posit that the solution is going to require a greater pragmatism in approaching this question. Although groups like Human Rights Watch have argued against the use of diplomatic assurances as the basis for repatriations, I would posit that such groups need to think about what alternative tools exist to manage humane treatment concerns in States that mistreat their citizens. Not only can assurances be effective when properly obtained and monitored, but taking a principled stand against assurances results in detainees being marooned in detention facilities years after they might otherwise have been released. For those detainees who come from countries where even assurances do not sufficiently mitigate the risk of mistreatment, the West is going to need to consider what realistic options exist to allow for third-country resettlement.


As we move forward then, I hope I have demonstrated that Common Article 3 and other applicable international legal rules do not answer important questions related to both the initiation and termination of detention in armed conflict with transnational terrorist groups. While there may be a range of reasonable policy answers, none are dictated by international law. I hope that the scholarly debate in this area will move beyond assertions that all that is needed is better implementation of existing law, and instead work will begin in earnest on addressing the difficult challenges I have identified. It is very easy for all of us to agree that the fight against transnational terrorism must be conducted in accordance with the rule of law, but it is much harder to say what the law exactly is, and how it should be applied in this context. As I continue my dialogue with other governments, I will continue to encourage them to work towards a common approach in dealing with these issues. I look forward to a good discussion this evening and for the rest of the conference of the way forward on these issues. Thank you.

  Back to top

U.S. Department of State
USA.govU.S. Department of StateUpdates  |   Frequent Questions  |   Contact Us  |   Email this Page  |   Subject Index  |   Search
The Office of Electronic Information, Bureau of Public Affairs, manages this site as a portal for information from the U.S. State Department. External links to other Internet sites should not be construed as an endorsement of the views or privacy policies contained therein.
About state.gov  |   Privacy Notice  |   FOIA  |   Copyright Information  |   Other U.S. Government Information

Published by the U.S. Department of State Website at http://www.state.gov maintained by the Bureau of Public Affairs.