He also said in his peroration: 'Freedom and fear, justice and cruelty, have always been at war, and we know that God is not neutral between them'.⁶ It is evident that the scope for traditional neutrality was implicitly understood by the Security Council, and explicitly proclaimed by the US, to be very limited in this anti-terrorist war. Naturally, some states, including Iran, proclaimed that they were with 'neither Bush nor bin Laden'; and not all states were willing to assist the US-led military action directly. However, this war confirmed the lesson of many recent episodes, including the 1991 Gulf War and the 1999 Kosovo War, that when armed conflict by a coalition is combined with the application of general UN sanctions against the adversary state, the scope for traditional (i.e. impartial) neutrality is indeed limited - especially so when, as in the case of al-Qaeda, the adversary operates in numerous states, which are required by the UN to take a range of measures against it.

The general indignation caused by terrorist attacks can also affect the implementation of jus in bello when fighting terrorism is the jus ad bellum. Because the terrorists started the war, it is sometimes argued, they are responsible for all the death and destruction that ensues. Such a view, implying that the peculiar circumstances involved in the jus ad bellum might override certain considerations of jus in bello in the war that follows, has no basis in the law. There was evidence of such thinking in some statements made in the US in connection with Afghanistan. In early December, discussing civilian casualties, US Secretary of Defense Donald Rumsfeld said: 'We did not start this war. So understand, responsibility for every single casualty in this war, whether they're innocent Afghans or innocent Americans, rests at the feet of the al-Qaeda and the Taliban'.⁷

Another possible connection between jus ad bellum and jus in bello relates to the principle of 'proportionality'. This is a long-established principle that sets out criteria for limiting the use of force. One of its meanings relates to the proportionality of a military action compared to a grievance. It involves a complex balance of considerations, and it would be incorrect to interpret this principle to imply a right of tit-for-tat retaliation. It would be legally unjustified for a military response to a terrorist act to have the objective of killing the same number of people and there was no suggestion or indication that this was a coalition objective.

The other main meaning of proportionality relates to the actual conduct of ongoing hostilities. As a US Army manual succinctly interprets it, 'the loss of life and damage to property incidental to attacks must not be excessive in relation to the concrete and direct military advantage expected to be gained'.⁸ This meaning of proportionality is an important underlying principle of jus in bello, and is not directly linked to jus ad bellum. However, this meaning of the principle is often difficult to apply in armed conflict. It may, but does not necessarily, limit the use of force to the same level or amount of force as that employed by an adversary. It exists alongside the principle of military necessity, which is defined in the US Army manual as one that 'justifies those measures not forbidden by international law which are indispensable for securing the complete submission of the enemy as soon as possible'.⁹ The principle of proportionality is therefore in tension, but not necessarily in conflict, with the current US military doctrine which favours the overwhelming use of force in order to achieve decisive victory quickly and at minimum cost in terms of US casualties.¹⁰

Anti-terrorist military operations

Anti-terrorist military operations, including those resulting from the events of 11 September, can have fundamental characteristics that are far removed from those of inter-state armed conflicts as principally envisaged in the laws of war. This is because of six factors relating to the nature of the opposition:

• Neither all terrorist activities, nor all anti-terrorist military operations, even when they have some international dimension, necessarily constitute armed conflict between states. Terrorist movements themselves generally have a non-state character. Military operations between a state and such a movement, even if they involve the state's armed forces acting outside its own territory, are not necessarily such as to bring them within the scope of application of the full range of provisions regarding international armed conflict in the 1949 Geneva Conventions and the 1977 Geneva Protocol I.¹¹

• Anti-terrorist operations may assume the form of actions by a government against forces operating within its own territory; or, more rarely, may be actions by opposition forces against a government perceived to be committing or supporting terrorist acts. In both these cases, the conflict may have more the character of non-international armed conflict (that is, civil war) as distinct from international war. Fewer laws-of war rules have been formally applicable to civil as distinct from international war, although the situation is now changing in some respects.

• In many cases, the attributes and actions of a terrorist movement may not come within the field of application even of the modest body of rules relating to non-international armed conflict. Common Article 3 of the 1949 Geneva Conventions is the core of these rules, but says little about the scope of application. The principal subsequent agreement on non-international armed conflict, the 1977 Geneva Protocol II, is based on the assumption that there is a conflict between a state's armed forces and organised armed groups which, under responsible command, exercise control over a part of its territory, and carry out sustained and concerted military operations. The protocol expressly does not apply to situations of internal disturbance and tension, such as riots, and isolated and sporadic acts of violence.¹²

• Since terrorist forces often have little regard for internationally agreed rules of restraint, the resolve of the anti-terrorist forces to observe them may also be weakened, given the low expectation of reciprocity and the tendency of some part of the public under attack to overlook any breaches by their own forces.

• A basic principle of the laws of war is that attacks should be directed against the adversary's military forces, rather than against civilians. This principle, violated in terrorist attacks specifically directed against civilians, can be difficult to apply in anti-terrorist operations, because the terrorist movement may not be composed of defined military forces that are clearly distinguished from civilians.

• Some captured personnel who are members of a terrorist organization may not meet the criteria for PoW status as set out in the 1949 Geneva Convention III. In particular, such personnel may fail to pass the tests of 'belonging to a Party to the conflict', 'being commanded by a person responsible for his subordinates', 'wearing a fixed distinctive sign', and 'conducting their operations in accordance with laws and customs of war'. However, even if they are not entitled to PoW status, such persons should still be treated humanely. (The question of prisoners is discussed in greater detail below.)

These six factors reflect the same underlying difficulty governments have in applying the laws of war to civil wars, namely, that the opponent tends to be viewed as a criminal, without the right to engage in combat operations. This factor above all explains why, despite the progress of recent decades, many governments are anxious about applying the full range of rules applicable in international armed conflict to operations against rebels and terrorists.

For at least 25 years, the US has expressed a concern, shared to some degree by certain other states, about the whole principle of thinking about terrorism in a laws-of-war framework. To refer to such a framework, which recognises rights and duties, might seem to imply a degree of moral acceptance of the right of any particular group to resort to acts of violence, at least against military targets.¹³ Successive US administrations have objected to certain revisions to the laws of war on the grounds that they might actually favour guerrilla fighters and terrorists, affording them a status that the US believes they do not deserve. When, on 29 January 1987, President Reagan explained why he was not recommending Senate approval of 1977 Geneva Protocol I additional to the 1949 Geneva Conventions, he mentioned that granting combatant status to certain irregular forces 'would endanger civilians among whom terrorists and other irregulars attempt to conceal themselves'. In addition, he indicated a concern that the provisions would endanger US soldiers when he stated in very general terms that 'the Joint Chiefs of Staff have also concluded that a number of the provisions of the protocol are militarily unacceptable'. He argued that US repudiation of the protocol was an important move against 'the intense efforts of terrorist organisations and their supporters to promote the legitimacy of their aims and practices'.¹⁴ Whether all this was based on a fair interpretation of 1977 Protocol I is the subject of impassioned debate which is beyond the scope of this survey. The key point is the US concern - which has not changed fundamentally in the years since 1987 - that the laws of war might be misused by some in order to give an unwarranted degree of recognition to terrorists. This concern has been evident in the Afghanistan crisis.

While the application of the law may be particularly difficult in anti-terrorist operations, it is not unimportant. Indeed, some failures to observe legal restraints in past campaigns have been instructive. In military operations with the purpose of stopping terrorist activities, there has been a tendency for counter-terrorist forces to violate basic legal restraints. There have been many instances in which prisoners were subjected to mistreatment or torture. In some cases, excesses by the government or by intervening forces may have contributed to the growth of a terrorist campaign against it. External states supporting the government have sometimes contributed to such excesses. Applying pressure on a government or army to change its approach to anti-terrorism, to bring it more into line with the laws of war and human-rights law, can be a difficult task.

One example of an anti-terrorist military campaign, the 1982-2000 Israeli presence in Lebanon, shows the importance of legal restraints in anti-terrorist operations, and the hazards that can attend a failure to observe them. This episode has certain similarities to the case of Afghanistan in 2001-02, as well as some obvious differences.

Israel's June 1982 invasion of Lebanon was explicitly in response to 'constant terrorist provocations', including, since July 1981, '150 acts of terrorism instigated by the PLO, originating in Lebanon, against Israelis and Jews in Israel and elsewhere: in Athens, Vienna, Paris and London'. Israel said that if Lebanon was unwilling or unable to prevent the harbouring, training and financing of terrorists, it must face the risk of counter-measures.¹⁵ The invasion led to the attacks on the inhabitants of Sabra and Shatila refugee camps outside Beirut in September 1982 by Israel's local co-belligerents, the Lebanese Phalangists. At the lowest estimates, several hundred Palestinians in the camps, including many women and children, were killed. This event aroused strong opposition internationally, and also in Israel. The Israeli authorities established a Commission of Inquiry, which concluded that, while the Phalangist forces were directly responsible for the slaughter, Israel bore indirect responsibility.¹⁶ During the whole period of Israeli military involvement in Lebanon, the treatment of alleged terrorist detainees also caused controversy. Israel was opposed to giving them PoW status on the grounds that as terrorists they were not entitled to it. The detainees were held in very poor conditions in notorious camps, including al-Khiam (run by the Israeli-created South Lebanese Army) and al-Ansar (run by the Israel Defence Forces).¹⁷ The Israeli military presence in Lebanon received extensive criticism internationally and in Israel, and it cost many lives among the Israel Defence Forces as well as their adversaries and in the civilian population. It ended with a unilateral Israeli withdrawal in May 2000.

Most anti-terrorist operations are largely internal matters, conducted by governments within their own territories, often within certain legal and prudential limits. Within functioning states, terrorist campaigns have often been defeated through slow and patient police work (sometimes with military assistance) rather than major military campaigns; for example, the actions against the Red Army Faction in Germany and the Red Brigades in Italy in the 1970s. The British military and police operation against 'Communist Terrorists' in Malaya after 1948 is a good example (in a colonial context) of a long-drawn-out and patient anti-terrorist campaign that was eventually successful.

In other contexts, too, Western armed forces, engaging with adversaries showing at best limited respect for ethical and legal restraints, have themselves managed to observe basic rules of the laws of war. This was the case in the 1991 Gulf War, in which Iraq mistreated prisoners, despoiled the environment and had to be warned in brutally clear terms not to engage in chemical or biological attacks and terrorist operations. The US-led Gulf coalition sought to observe the law not because of any guarantee of reciprocity, but because such conduct was important to the maintenance of internal discipline, and of domestic and international support. Similar conclusions were drawn from the 1999 Kosovo War. Reciprocity with one adversary in one particular conflict is not the only basis for observing the laws of war.

The US armed forces have indicated their intention to observe the rules governing international armed conflicts, even in situations which may differ in certain respects from the classical model of an inter-state war. A principle codified in the Standing Rules of Engagement issued by the US Joint Chiefs of Staff on 15 January 2000 spells this out:

US forces will comply with the Law of War during military operations involving armed conflict, no matter how the conflict may be characterized under international law, and will comply with its principles and spirit during all other operations.¹⁸

The development by US and allied forces of techniques of bombing that are more accurate than in previous eras has increased the technical possibilities of air power being employed discriminately, and therefore in a manner that is compatible with laws-of-war rules about targeting. This is a momentous development in the history of war, yet its effects, especially as regards operations against terrorists, should not be exaggerated, as it cannot guarantee no deaths of innocents. Precision-guided weapons are generally better at hitting fixed objects, such as buildings, than moving objects which can be concealed, such as tanks. Civilian deaths will still occur, whether because certain dual-use targets are attacked, because of the proximity of military targets to civilians, or because of faulty intelligence and human or mechanical errors. In addition, malevolence and callousness can still lead to attacks on the wrong places or people. A further problem with the new type of US bombing campaign is that, in the eyes of third parties, it can easily look as if the US puts a lower value on the lives of Iraqis or Serbs or Afghans than it does on its own almost-invulnerable aircrews: a perception which can feed those hostile views of the US that help to provide a background in which terrorism can flourish.

In an anti-terrorist war, as in other wars, there can be strong prudential considerations that militate in favour of observing the laws of war. These include securing public and international support; ensuring that terrorists are not given the propaganda gift of atrocities by their adversaries; and maintaining discipline and high professional standards in the counter-terrorist forces. Such considerations may carry great weight even in conflicts, or particular episodes within them, which differ from what is envisaged in the formal provisions regarding scope of application of relevant treaties These considerations in favour of observing the law may be important irrespective of whether there is reciprocity in observance of the law by all the parties to a particular war. However, it is not realistic to expect that the result of the application of such rules will be a sanitised form of war in which civilian suffering and death is eliminated.

Afghanistan

In wars in Afghanistan over the centuries, conduct has differed markedly from that permitted by the laws of war. These wars always had a civil war dimension, traditionally subject to fewer rules in the laws of war; and guerrilla warfare, already endemic in Afghanistan in the nineteenth century, notoriously blurs the distinction between soldier and civilian which is at the heart of the laws of war. Some local customs, for example regarding the killing of prisoners and looting, are directly contrary to long-established principles of the law. Other customs are different from what is envisaged by the law, but are not necessarily a violation of it: for example, the practice of soldiers from the defeated side willingly joining their adversary rather than being taken prisoner. In some cases, conduct has been consistent with international norms: for example, the International Committee of the Red Cross (ICRC) had access to some prisoners during the Soviet intervention. Overall, however, compliance has been limited.

The war in Afghanistan - principally between the Taliban and Northern Alliance forces - had been going on for many years before the events of 11 September 2001. Both parties were called upon to comply with their obligations under international humanitarian law. UN Security Council Resolution 1193 of 28 August 1998, passed unanimously, reaffirmed inter alia

that all parties to the conflict are bound to comply with their obligations under international humanitarian law and in particular the Geneva Conventions of 12 August 1949, and that persons who commit or order the commission of grave breaches of the Conventions are individually responsible in respect of such breaches.

The reference to grave breaches would appear to suggest that the Security Council viewed all the rules of the 1949 Geneva Conventions as applicable, and not just common Article 3, which deals with civil war. The clear terms of this resolution are a reminder that, three years before it became directly involved, the US as well as other powers did view the laws of war as applicable to the Afghan conflict.

Following the events of 11 September 2001, when it was evident that an armed conflict between the coalition and the Taliban was likely, the ICRC, consistent with its general practice, sent confidential messages to certain governments reminding them of their obligations under international humanitarian law. Unfortunately, in the first of what would be many clashes between humanitarian bodies and national governments in this crisis, the ICRC messages touched on the issue of nuclear weapons in a way that invited antagonism and rejection: a replacement message had to be sent.¹⁹ The ICRC subsequently issued some public statements on the application of the laws of war in this crisis, reminding all the parties involved - the Taliban, the Northern Alliance and the US-led coalition - of their obligations to respect the law, and stating that the ICRC was continuing a wide range of activities inside Afghanistan.²⁰

Like the period of Soviet intervention of 1979-89, and indeed wars in many countries in the period since 1945, the war in Afghanistan from 7 October 2001 to the present can perhaps be best characterised as 'internationalised civil war'. This is not a formal legal category, but an indication that the rules pertaining to both international and civil wars may be applicable in different aspects and phases of the conflict.²¹

On the technical legal question as to which of the main laws of war treaties have been formally binding on the belligerents in the hostilities in Afghanistan since October 2001, the 1907 Hague Convention IV on land warfare applies because of its status as customary law, binding on all states whether or not parties to the treaty. In addition, Afghanistan, and also the main members of the international coalition, are parties to the following agreements:

• the 1925 Geneva Protocol on gas and bacteriological warfare;

• the 1948 Genocide Convention;

• the four 1949 Geneva Conventions.

While some of the states involved are parties to certain additional agreements, the above-named treaties provide the basic treaty framework for considering the application of the law in this particular armed conflict. In addition, rules of customary international law apply, including certain provisions of 1977 Geneva Protocol I that are accepted as having that status.

As regards civil-war aspects of the Afghan war, some but not all of the provisions of the agreements listed above apply. The 1907 Hague Land War Convention's Article 2 indicates that the convention and its annexed regulations apply only to wars between states. The 1925 Geneva Protocol is not formally applicable to civil wars.²² The 1948 Genocide Convention is considered to apply to non-international as well as international armed conflict. In the 1949 Geneva Conventions, common Article 3 lists certain minimum provisions for humane treatment of those taking no active part in hostilities that are to be applied in non-international armed conflict.

The implementation of the laws of war posed a problem for Operation Enduring Freedom from the start. The number of Afghan forces involved in the war, many of which were under local warlords, and the lack of clear structures of authority, decision-making and military discipline among them, were all complicating factors. One of the groups involved, al-Qaeda, had committed or supported numerous criminal acts in foreign countries. Difficult practical issues facing the coalition included: the need to conduct operations discriminately; the possibility that adversary forces might mistreat or execute coalition prisoners; the possibility that some enemy personnel facing capture might be reluctant to surrender their weapons, and that they might not qualify for PoW status; maintenance of order (and avoidance of looting and revenge killings) in liberated towns; how to bring pressure on Northern Alliance forces to observe basic norms; and assistance for humanitarian relief operations.

The active role of the media in this war has ensured that these issues have been heavily publicised. Reporters are operating close to, and even in front of, the front lines, sending back reports and high-resolution pictures as events unfold. Up to the end of January 2002, more reporters had died while covering the war in Afghanistan than non-Afghan coalition military personnel.²³ The lesson of other modern wars has been confirmed: that the press plays a critical role in repeatedly raising matters germane to the laws of war.

One issue relating to the laws of war which impacted upon all phases of operations in Afghanistan is humanitarian relief. Such relief constitutes a major activity in almost all contemporary wars. The need for humanitarian relief is particularly likely to arise in anti-terrorist operations against a weak or failed state, because such states breed conditions in which terrorist movements can operate and large-scale human misery can occur. The fact of a war being against terrorists, while it may affect the mode of delivery (since land convoys may be vulnerable to seizure) does not affect the law applicable to the provision of relief. The basic obligations of the various parties to an armed conflict to assist in and protect humanitarian relief operations are embodied in 1949 Geneva Convention IV, on civilians.²⁴

Humanitarian relief issues proved critically important in Afghanistan. Announcing the start of Operation Enduring Freedom, President Bush stated: 'As we strike military targets, we will also drop food, medicine and supplies to the starving and suffering men and women and children of Afghanistan'.²⁵ US forces air-dropped considerable quantities of aid at the same time as the major bombing operations took place. There was tension between the US and humanitarian agencies, some of which were critical of the bombing campaign, doubtful of the value of air-dropped supplies and concerned about the aggravated risks and obstacles to their work that resulted from the military operations. While the effects of war and the onset of winter heightened the urgent need for aid, the collapse of the Taliban regime in early December 2001 and its replacement by the interim administration facilitated, but by no means guaranteed, the secure delivery of aid by land routes. A wide range of countries and organisations took part in the provision of aid.²⁶ Refugees started to return in significant numbers: over 3,000 per day in January, but there were also movements of ethnic Pashtun from Afghanistan to Pakistan.²⁷

Bombing

The anti-terrorist rationale of the coalition operations in Afghanistan gave a particular character to two issues on which the laws of war had a substantial bearing: bombing and the status and treatment of prisoners.

The principle that the bombing of Afghanistan should be discriminate was repeatedly stated. On 21 October, General Richard B. Myers, the Chair of the Joint Chiefs of Staff, said:

The last thing we want are any civilian casualties. So we plan every military target with great care. We try to match the weapon to the target and the goal is, one, to destroy the target, and two, is to prevent any what we call 'collateral damage' or damage to civilian structures or civilian population.²⁸

US bombing in Afghanistan aroused much international concern, particularly as regards civilian casualties and damage. There were reports of many attacks causing significant numbers of civilian casualties and damage. Two attacks in October hit an ICRC warehouse in Kabul. According to press reports, over a hundred villagers may have died in bombings on 1-2 December of Kama Ado and neighbouring villages in eastern Afghanistan, not far from the cave complex at Tora Bora. In several cases, bombings led to casualties among coalition forces: while this is not a laws-of-war issue as such, and is not uncommon in armed conflicts, it highlights the fact that bombing remains far from clinically accurate.

It is difficult to arrive at a reliable estimate of the overall number of civilian deaths caused by the bombing in Afghanistan. As in the 1991 Gulf and 1999 Kosovo wars, the Pentagon has been reluctant to issue figures. Apart from certain statements by the Taliban, the highest reported estimate is over 3,500 as of mid-December, but there are grounds for doubt.²⁹ In sharp response, Rumsfeld stated in an interview on 8 January 2002 that

there probably has never in the history of the world been a conflict that has been done as carefully, and with such measure, and care, and with such minimal collateral damage to buildings and infrastructure, and with such small numbers of unintended civilian casualties.³⁰

Even if the figure is an over-estimate, the bombing has clearly resulted in large numbers of civilian deaths and caused thousands of Afghan civilians to flee their homes.³¹ Much work is needed to put right such human and material damage as can be repaired.

In legal terms, the incidence of civilian deaths per se may not always constitute a violation, absent other factors; and there are strong reasons to believe US statements that civilian deaths in the above types of episode were unintended. Some may well have resulted from errors of various kinds, and some may have been unavoidable 'collateral damage'. One cause of civilian casualties may have been the fact that, in a legacy from the period of Soviet involvement in Afghanistan, many Taliban military assets were located in towns, where they were less vulnerable to raids from rural-based guerrillas, but where they were of course closer to civilians who risked getting hit in bombing attacks. While much of the bombing was discriminate, questions are raised about whether all appropriate measures were taken to reduce civilian casualties and damage. Even if much of the civilian death and destruction is not a violation of the law, the resulting adverse public perception risks harming the coalition cause.

The US was particularly sensitive about accusations that it had acted indiscriminately. Rumsfeld accused the Taliban and al-Qaeda leaders of both causing and faking civilian damage: 'They are using mosques for command and control, for ammunition storage, and they're not taking journalists in to show that. What they do is when there's a bomb goes down, they grab some children and some women and pretend that the bomb hit the women and the children.'³² What truth there is in all this is difficult to determine.

Did the concern over civilian casualties undermine the US bombing effort? Its eventual success against the Taliban would suggest not, but there were indications that the concern had serious effects. It was reported that the US had deliberately slowed the pace of the campaign, and increased the risk to the people executing it, because of legal restraints and moral values. It was also stated that war planners frequently chose not to hit particular targets, even if they were militarily important, and pilots allegedly complained of lost opportunities. Yet the planners could not give their reasoning for ruling out certain targets, as it would give the adversary 'a recipe book for not being bombed'. The issue of civilian casualties also became ammunition for inter-service battles, particularly for Army arguments in favour of 'boots on the ground'.³³

One issue raised by the bombing, and which involved the risk of immediate and possible future civilian casualties, was the use of cluster bombs. These are air-dropped canisters containing numerous separate bomblets which disperse over a given area. The bomblets, which are meant to explode on impact or to self-deactivate after a specific period, can cause particularly severe problems if they fail to do so. There have been objections to their use, principally on the ground that they have a tendency, like anti-personnel land-mines, to kill people long after the conflict is over. Reports from Kosovo and elsewhere have confirmed the general seriousness of the problem.³⁴ On the other hand, some evidence from the Afghanistan campaign suggests that cluster bombs were an effective weapon. As the law stands, there has been no firm agreement to outlaw cluster bombs, and while they cannot be said to be illegal per se, their use does raise questions regarding their compatibility with fundamental principles of the laws of war. They are certain to be the subject of further pressures to limit or stop their use, or to ensure more effective safeguards against later accidental detonations.

A second issue concerns the use of bombing in the hunt for Taliban and al-Qaeda personnel, following the fall of the Taliban regime in early December 2001. In the preceding phase, bombing had been used primarily in support of Northern Alliance frontal operations aimed at capturing the main Taliban-held cities. Once this was achieved, a good deal of the bombing was directed against remnant Taliban and al-Qaeda forces and their leaders. Several incidents were reported in the press in which those killed were neither. The reports drew attention to the difficulty of distinguishing between civilians and these forces. They also raised the question, of broader significance in anti-terrorist wars: to what extent is bombing an appropriate form of enforcement once a state is, to a greater or lesser degree, under the control of a government that is opposed to the terrorists? At that point, to what extent can the focus be transferred to other forms of police and military action that may be less likely than bombing to cause civilian casualties?

One long-standing prohibition in warfare is the rule against use of gas and bacteriological methods of warfare. The US repeatedly expressed concern that al-Qaeda might be preparing to use such methods in terrorist attacks. In addition, there were situations in which there could have been pressures for the US to use gas. When, in 1975, the US ratified the 1925 Geneva Protocol, it indicated that it considered that certain uses of riot-control agents in armed conflict did not violate the protocol.³⁵ In early December 2001, Rumsfeld was asked at a press conference if the US might use gas in the hunt for Taliban and al-Qaeda personnel in mountain caves. Rumsfeld's response contained no denial:

Well, I noticed that in Mazar, the way they finally got the dead-enders to come out was by flooding the tunnel. And finally they came up and surrendered, the last hard core al-Qaeda elements. And I guess one will do whatever it is necessary to do. If people will not surrender, then they've made their choice.³⁶

From late November 2001, the status and treatment of prisoners taken in the war on terror became a major international controversy. Within the Pentagon, if not necessarily at the political level, it had been recognised early on that the prisoner issue could be difficult. An unpublished document circulated by the USAF's International and Operations Law Division on 21 September 2001 outlined the dimensions of the problem: terrorists were to be treated as 'unlawful combatants'; it was 'very unlikely that a captured terrorist will be legally entitled to PoW status under the Geneva Conventions'; however, there was a 'practical US interest in application of Law of Armed Conflict principles in the context of reciprocity of treatment of captured personnel.' As regards treatment upon capture,

if a terrorist is captured, Department of Defense members must at the very least comply with the principles and spirit of the Law of Armed Conflict … A suspected terrorist captured by US military personnel will be given the protections of but not the status of a PoW.³⁷

Initially, international attention focused on one event: the killing of a large number of Taliban and al-Qaeda prisoners following the revolt at Qala-e Jhangi Fort near Mazar-e Sharif in the period 25 November -1 December 2001. Even before the prisoners were taken at Kunduz at around the time of its fall on 23-24 November, it was evident that the surrender and imprisonment of the non-Afghan forces fighting alongside the Taliban would be extremely difficult. At the same time, there was very little sign of serious preparation for handling prisoners, large numbers of whom were likely to be particularly dangerous. The precise chain of events leading to the revolt has yet to be established, but the causes appear to include the following heady mix: these were particularly fanatical soldiers, for whom the whole concept of surrender would be anathema; the arrangements for receiving, holding and processing the prisoners appear to have been ad hoc and casual; a number of prisoners had not surrendered all their weapons, and by not having laid down their arms they failed to meet the requirements for PoW status; the prisoners were held in a place where there was a large store of weapons, to which they gained access; and some reports suggest that the prisoners feared that they were about to be killed, so had nothing to lose by revolt.

When asked at a press conference whether the suppression of the prison revolt at Mazar-e Sharif had been proportionate, Rumsfeld indicated bafflement:

Now, the word 'proportion' - 'proportionate' is interesting. And I don't know that it's appropriate. And I don't know that I could define it. But it might be said - and I wouldn't say it - (laughter) - but it might be said by some that to quickly and aggressively repress a prison riot in one location might help dissuade people in other locations from engaging in prison riots and breaking out of prison and killing more people. I don't know that that's true. It might also persuade the people who are still in there with weapons, killing each other and killing other people, to stop doing it … Your question's too tough for me. I don't know what 'proportionate' would be.³⁸

The revolt at Qala-e Jhangi Fort was a desperate struggle in which not only many prisoners, but also a number of Northern Alliance troops in charge of the fort, died. US bombing, and sharp-shooting by UK special forces, played a part in the defeat of the uprising. Public discussion in the UK and elsewhere has focused on the events at the fort, including the question of whether the force used to quell the rebellion was excessive. If the situation was as desperate and threatening as reports indicated, the use of force is hardly surprising. Public discussion should more usefully focus on how prisoners should be received and dealt with. Events at the fort raised the issue of whether the US and, in particular, the Northern Alliance, had a clear policy for treating prisoners, including the foreign fighters. The real cause of the disaster was probably a failure to think the issue through before the prisoners arrived at the fort, and especially the failure to disarm all prisoners.

Other reports of maltreatment and deaths of prisoners elsewhere confirm that the overall approach of the Northern Alliance was defective. In particular, by late December there had been numerous reports of Afghan captors beating their detainees, and the ICRC was reported as expressing concern that it had been able to register only 4,000 of the 7,000 prisoners which the US said it and its Afghan allies had in custody.³⁹

The actual influence of the US and its coalition partners over the Northern Alliance's actions in such basic matters as protection of prisoners - and whether they used it - is open to question. Coalition members have expressed different views on this. In his Pentagon press briefing on 30 November, Rumsfeld indicated - in general terms, not in connection to the prisoner question - that the US does have influence with the forces with which it operated in Afghanistan:

We have a relationship with all of those elements on the ground. We have provided them food. We've provided them ammunition. We've provided air support. We've provided winter clothing. We've worked with them closely. We have troops embedded in their forces and have been assisting with overhead targeting and resupply of ammunition. It's a relationship.⁴⁰

This contrasts with an earlier statement of British Prime Minister Tony Blair, who was asked on 13 November, again in general terms, 'What sanctions do we have over the Northern Alliance?'. He replied simply, 'None'.41

The question of the status and treatment of al-Qaeda fighters taken prisoner in Afghanistan, arguably distinguishable from the status and treatment of Taliban fighters taken prisoner, involves the important but difficult issue of whether or not such combatants are considered lawful. The key factor in determining the lawfulness of a combatant, and therefore the entitlement to participate directly in hostilities, is the affiliation of the combatant to a party to the conflict.

Lawful combatants comprise the organised armed forces (including militias and volunteer corps) of a state or otherwise recognised party to a conflict. They also include members of certain other militias and volunteer corps, including those of organised resistance movements, belonging to a party to the conflict, provided that they meet certain criteria: they must be under a responsible command system; wear a fixed distinctive sign; carry arms openly; and conduct their operations in accordance with the laws of war. Members of regular armed forces who meet such criteria may well be lawful combatants even if the regime that they serve is not recognised as the lawful government of the state. Lawful combatants are entitled to PoW status and all of the rights set forth in the Geneva Convention III. Lawful combatants cannot be punished for the mere fact of having participated directly in hostilities, but they can be tried for any violations of international law, including the laws of war, they may have committed.

What is the status of those many people who are involved in hostile activities in various ways, but who fail to meet the above criteria? Since the Second World War, the problem had arisen repeatedly; one suggested term for a wide range of such combatants was 'unprivileged belligerents'.⁴² Many belligerents failing to meet one of the criteria were viewed as entitled to PoW status, but not all were.⁴³ In current US military manuals two terms with apparently identical meaning, 'unlawful combatants' and 'illegal combatants', are used to refer to those who are viewed as not being members of the armed forces of a party to the conflict and not having the right to engage in hostilities against an opposing party.⁴⁴ Such combatants can face penal sanctions for participating directly in hostilities and for other acts they may commit, and they do not have the right to PoW status; but they do retain a claim to certain fundamental guarantees regarding their detention and any judicial proceedings against them.

The distinction between lawful and unlawful combatants is important. Article 5 of the 1949 Geneva Convention III provides that, in cases of doubt, prisoners shall be treated as PoWs 'until such time as their status has been determined by a competent tribunal'. While this Article does not specify the nature of the 'competent tribunal', Article 45 of the 1977 Geneva Protocol I, in elaborating these provisions, allows for considerable leeway in the procedure by which a tribunal could reach such a decision. The possibilities that the proceedings could take place after a trial for an offence, and also in camera in the interest of state security, are not excluded. In US official manuals the general principle that Article 5 tribunals must be held is not contested. The US Army manual states unequivocally: 'When doubt exists as to whether captured enemy personnel warrant PW [prisoner of war] status, Art. 5 Tribunals must be convened'.⁴⁵

The fact that certain prisoners may be viewed as unlawful combatants, and may (after a tribunal has so determined) be denied PoW status, does not mean that they have no legal rights at all. A strong argument can be made that, whether or not they are formally entitled to such rights, they should have certain of the basic safeguards accorded to PoWs. Furthermore, Article 75 of the 1977 Geneva Protocol I elaborates a range of fundamental guarantees that are intended to provide minimum rules of protection for all those who do not benefit from more favourable treatment under other rules. Any state with a claim to act legally in international relations, even if not itself a party to the 1977 Geneva Protocol I (neither the US nor Afghanistan is a party), must take the rules in Article 75 seriously as the minimum standards, especially as these provisions are viewed as customary law.⁴⁶

The United Kingdom's long engagement against terrorism in Northern Ireland provides one precedent for applying international rules to prisoners whose status is contested. While denying that there was an armed conflict whether international or otherwise, and strongly resisting any granting of PoW status to detainees and convicted prisoners, the UK did come to accept that international standards had to apply to their treatment. The UK Commission of Inquiry whose report in 1972 led to this conclusion is an interesting example of asserting the wider relevance, even in an internal conflict, of certain international legal standards, including some from the main body of the four 1949 Geneva Conventions.⁴⁷

In 1967-68, during the US involvement in Vietnam, the US issued directives to classify Viet Cong main force and local force personnel, and certain Viet Cong irregulars, as PoWs. This was despite the existence of doubts and ambiguities as to whether these forces met all the criteria in Article 4 of the 1949 Geneva Convention III. However, there was a significant exception in respect of terrorism. Viet Cong irregulars were only to be classified as PoWs if captured while engaging in combat or a belligerent act under arms, 'other than an act of terrorism, sabotage, or spying'. There was provision for establishing Article 5 tribunals to determine, in doubtful cases, whether individual detainees were entitled to PoW status. Those not entitled to such status were to be transferred to the South Vietnamese authorities.⁴⁸

US policy regarding prisoners taken in Afghanistan appeared initially to leave only limited room for application of the rules of protection contained in the laws of war. By referring to these prisoners generally as 'battlefield detainees' and 'unlawful combatants', the US signalled its unwillingness to classify al-Qaeda and Taliban prisoners as PoWs. On 11 January 2002, when asked whether the ICRC would have any access to the prisoners who had just been taken to the US naval base at Guantanamo Bay in Cuba, Rumsfeld stated:

I think that we're in the process of sorting through precisely the right way to handle them, and they will be handled in the right way. They will be handled not as prisoners of war, because they're not, but as unlawful combatants. The, as I understand it, technically unlawful combatants do not have any rights under the Geneva Convention. We have indicated that we do plan to, for the most part, treat them in a manner that is reasonably consistent with the Geneva Conventions, to the extent they are appropriate, and that is exactly what we have been doing.⁴⁹

In the event, ICRC officials started interviewing detainees at Guantanamo on 18 January, and were able to establish a permanent presence there. On 22 January Rumsfeld, contrary to his earlier statement, recognised that 'under the Geneva Convention, an unlawful combatant is entitled to humane treatment'.⁵⁰ On 7 February, the White House, in the first major policy statement on the issue, announced that Taliban prisoners were covered under Geneva Convention III, but al-Qaeda members were not. While neither group was accorded full PoW status, the White House gave detailed assurances about their treatment.

Two considerations contributed to the US determination not to classify as PoWs prisoners taken in Afghanistan: the first related to conditions of detention of prisoners, and the second to the conduct of judicial proceedings. On conditions of detention, the 1949 Geneva Convention III famously states that PoWs are only obliged to give name, rank, date of birth and personal or serial number. The US was anxious to obtain considerably more information from them, although whether a different classification actually improves the prospects of securing accurate information is debatable. As regards judicial proceedings, from early on in the war, the US reportedly intended to prosecute a number of al-Qaeda and Taliban leaders, including Osama bin Laden if captured. The US has been reluctant to pursue the procedure laid down in the Geneva Convention, which specifies that any sentence of a PoW must be 'by the same courts according to the same procedure as in the case of members of the armed forces of the Detaining Power'.⁵¹ Such procedures, US officials feared, could provide opportunities for al-Qaeda suspects and their lawyers to prolong legal processes and attract publicity. There was also concern that in cases involving defendants with no documents and no willingness to collaborate with any of the procedures, and where evidence might be largely based on intelligence sources, it could be difficult to provide evidence that met high standards of admissibility, and equally high standards of proof of direct personal involvement in terrorist activities. Further, al-Qaeda might learn valuable information, for example, about its vulnerability to intelligence gathering, from evidence in open court. In addition, following the normal US military procedures for appeals was seen as posing problems.⁵²

In certain other respects, too, there could be difficulties in treating some of the prisoners as normal PoWs. For example, a practice that is normally pursued after a war - releasing and repatriating prisoners - is complicated in this case by three considerations. First, while the war in Afghanistan may be concluded at a definite date, it may be decades before the 'war on terror' can be declared to be over for the US. Second, unlike PoWs in a 'normal' inter-state war, the prisoners concerned might continue to be extremely dangerous after release, given their training and motivation to commit acts of terrorism. Third, their countries of origin might refuse to accept them back, except perhaps as prisoners.

President Bush's Military Order of 13 November 2001 provides for the option of trying certain accused terrorists by military commissions operating under special rules. It specifies that individual terrorists, including members of al-Qaeda, can be detained and tried 'for violations of the laws of war and other applicable laws', and that the military commissions would not be bound by 'the principles of law and the rules of evidence generally recognized in the trial of criminal cases in the United States district courts'. It also contains some extremely brief provisions for humane conditions of detention, and provides for the Secretary of Defense to issue detailed regulations on such matters as the conduct of proceedings of the military commissions.⁵³ The provision for trial by military commission is not unprecedented: for example, President Roosevelt's Proclamation of 2 July 1942, bluntly entitled 'Denying Certain Enemies Access to the Courts of the United States'.⁵⁴ On 30 November 2001, the President's Counsel offered several assurances, including that such commissions are one option, but not the only option.⁵⁵ Nevertheless, President Bush's Military Order remained the subject of considerable legal and political debate in the US and elsewhere as to its constitutionality, practicability and advisability. One test of the detailed regulations, which had not been issued at the time of writing, will be whether the procedures of the military commissions conform with the ten recognised principles of regular judicial procedure outlined in Article 75(4) of 1977 Geneva Protocol I.

Overall, the US handling of questions relating to the treatment and status of prisoners, especially those under Northern Alliance control, has caused widespread concern and criticism. The principles briefly indicated in the above-mentioned USAF document of 21 September 2001 were not consistently followed, and practical arrangements, especially around the time of the rebellion at Mazar-e Sharif, were inadequate. Although many key US positions were defensible, especially that certain prisoners might not qualify for PoW status, aspects of US policy and procedures were poorly presented, and in some cases did not appear to be fully thought-out. The prisoner issue - always sensitive anyway - was especially significant in this war: if the coalition was perceived to have treated prisoners inhumanely, or to have regarded their status and treatment as being in an international legal limbo, there would be risks of a general weakening of the prisoner regime, including for any coalition personnel taken prisoner in the ongoing war on terrorism. The handling of this issue was a potential threat to coalition unity. The controversies over the prisoner question had a special resonance because of the concern of other countries that the US had been moving towards unilateralism generally, on a wide range of matters: in this perspective, fairly or unfairly, the US reluctance to accept the full application of the 1949 Geneva Convention III on PoWs to those particular prisoners was seen as one more example of a selective approach to international law. The White House statement of 7 February, while not answering all concerns, provided reassurance that US polices would follow provisions of the Geneva Conventions.

Conclusion

Bush administration policies on these laws-of-war issues have evolved in a generally sensible direction. However, neither the United States nor its critics have shown a clear understanding of how the laws of war should be applied to military counter-terrorist operations. This is in no small part because the application of those laws is complicated, as a return to the three questions set out at the beginning of this essay shows:

• First, according to a strict interpretation of their terms, the main treaties relating to the conduct of international armed conflict are formally and fully applicable to anti-terrorist military operations only when those operations have an inter-state character. Where anti-terrorist operations have the character of civil war, the parties must apply, as a minimum, the rules applicable to civil wars.

• Second, in anti-terrorist military operations, certain phases and situations may well be different from what was envisaged in the main treaties on the laws of war. They may differ from the provisions for both international and non-international armed conflict. Recognising that there are difficulties in applying international rules in the special circumstances of anti-terrorist war, the attempt can and should nevertheless be made to apply the law to the maximum extent possible. This conclusion is reinforced by decisions of commissions of inquiry, a resolution of the UN Security Council, some practice of states and considerations of prudence.

• Third, the great majority of prisoners taken in war meet the criteria for PoW status laid down in international treaties, and must be so treated if they continue to be held. However, in an anti-terrorist war, as in other wars, there are likely to be certain individuals who do not meet the criteria. Such individuals, for example, members of a terrorist organisation, may present special problems as prisoners, and may pose a continuing threat even after the end of a war. The standard presumption outlined in treaty law and in US military manuals is that such people should be accorded the treatment, but not the status, of a PoW until a tribunal convened by the captor determines the status to which the individual is entitled. In cases where it is determined that they are not PoWs, there are certain fundamental rules applicable to their treatment, including those outlined in Article 75 of 1977 Geneva Protocol I. Any prisoner, whether or not classified as a PoW, can be tried for offences, including those against international law, that were committed prior to capture.

There are ample grounds for questioning whether military operations involving action against terrorists constitute either a new, or a wholly distinct, category of war. The coalition operations in Afghanistan, and the larger war against terrorism of which they are a part, are not completely unlike earlier wars. Many forms of military action and issues raised are similar to those in previous military operations, and concern issues addressed by the laws of war.

Events in Afghanistan have confirmed that there are particular difficulties in applying the laws of war to anti-terrorist operations. A war that has as a fundamental purpose the pursuit and bringing to justice of people deemed to be criminals involves many awkward issues for which the existing laws of war are not a perfect fit. The use of proxies in an anti-terrorist war risks creating a situation in which major powers are at the mercy of their local agents, whose commitment to the laws of war may be slight.

Despite such problems, treating, or appearing to treat, the law in a cavalier manner risks creating new problems. If a major power is perceived as ignoring certain basic norms, this may have a negative effect in a coalition, or on enemies. It may also affect the conduct of other states in other conflicts. In that wider sense, the principle of reciprocity in the observance of law retains its value.

There has been no serious suggestion that the existing legal framework can or should be abandoned, and no proposals for alternative detailed rules. The existing laws of war, however imperfect, are irreplaceable. Since issues relating to the laws of war arise with great frequency in anti-terrorist military operations, and will no doubt continue to do so in the continuing 'war against terrorism', there is a need for greater clarity about observance of the basic laws of war, and about the principles to be followed if and when parties consider that specific circumstances justify specific derogations from that body of law.


Adam Roberts is Montague Burton Professor of International Relations at Oxford University and Fellow of Balliol College. He is co-editor, with Richard Guelff, of Documents on the Laws of War (Oxford and New York: Oxford University Press, 2000.) An earlier version of this paper, dated 16 January 2002 and entitled 'The Relevance of the Laws of War in Anti-Terrorist Wars', appeared on the SSRC website in January-March 2002.

Notes

¹ For texts of treaties and other international documents on terrorism, and useful discussion thereof, see esp. Rosalyn Higgins and Maurice Flory (eds), Terrorism and International Law (London: Routledge, 1997). For more recent treaties and UN resolutions see the information on terrorism on the UN website, www.un.org.

² 'Crimes against humanity', defined in the Charter and Judgment of the International Military Tribunal at Nuremburg in 1945-46, are more fully defined in Article 7 of the 1998 Rome Statute of the International Criminal Court (not yet in force).

³ 1949 Geneva Convention III Relative to the Treatment of Prisoners of War, Articles 4(B)(2) and 122. See also the references to 'neutral and other States not Parties to the conflict' in 1977 Geneva Protocol I Additional to the Geneva Conventions of 12 August 1949, and Relating to the Protection of Victims of International Armed Conflicts, Articles 9, 19, 31 etc.

⁴ UN Security Council Resolution 1267 of 15 October 1999. See also Security Council Resolutions 1076 of 22 October 1996 and 1193 of 28 August 1998, both of which, in addressing the ongoing conflict in Afghanistan, referred to the problem of terrorism there, and called upon states to take specific actions, most notably to end the supply of arms and ammunition to all parties to the conflict.

⁵ UN Security Council Resolutions 1368 of 12 September 2001 and 1373 of 28 September 2001.

⁶ President George W. Bush, Address to a Joint Session of Congress and the American People, Washington DC, 20 September 2001.

⁷ Rumsfeld, at the end of his opening statement at a news briefing at the Pentagon on 4 December 2001.

⁸ US Army, The Law of Land Warfare, FM 27-10, Department of the Army Field Manual, Washington DC, July 1956, revised 15 July 1976, paragraph 41.

⁹ US Army, The Law of Land Warfare, FM 27-10, paragraph 3. A subsequent official US exposition of the principle states: 'Only that degree and kind of force, not otherwise prohibited by the law of armed conflict, required for the partial or complete submission of the enemy with a minimum expenditure of time, life and physical resources may be applied.' US Navy, The Commander's Handbook of the Law of Naval Operations, NWP 1-14M, Department of the Navy, 1995, paragraph 5.2.

¹⁰ See particularly Rumsfeld's comments on proportionality in relation to the suppression of the revolt at Mazar-e-Sharif, below.

¹¹ In ratifying 1977 Geneva Protocol I in 1998, the United Kingdom made a statement that the term 'armed conflict' denotes 'a situation which is not constituted by the commission of ordinary crimes including acts of terrorism whether concerted or in isolation'.

¹² 1977 Geneva Protocol II Additional to the Geneva Conventions of 12 August 1949, and Relating to the Protection of Victims of Non-International Armed Conflicts, Article 1.

¹³ For fuller discussion, and evidence that concern about the hazards of coping with terrorism in a laws-of-war framework is not new, see my chapter in Lawrence Freedman et al. (eds), Terrorism and International Order, Chatham House Special Paper (London: Routledge & Kegan Paul, 1986), esp. pp. 14-15.

¹⁴ President Reagan's letter of transmittal of the 1977 Geneva Additional Protocol II to the US Senate. Treaty Doc. No. 2, 100th Congress, 1st Session, at III (1987), reprinted in American Journal of International Law, vol. 81, no. 4, October 1987, pp. 911-12.

¹⁵ Professor Yehuda Blum, Permanent Representative of Israel, at UN Security Council, 5 June 1982. Security Council Official Records, 2374th meeting, p. 7. The Security Council unanimously demanded an end to all military activities and a withdrawal of Israeli forces from Lebanon in Resolutions 508 and 509 of 5 and 6 June 1982 respectively.

¹⁶ Commission of Inquiry into the Events at the Refugee Camps in Beirut: Final Report, Jerusalem, 1983, pp. 53-4. The report is reprinted in International Legal Materials , vol. 22, no. 3, May 1983, at p. 473. The Commission was chaired by Yitzhak Kahan, President of the Supreme Court of Israel.

¹⁷ In a case concerning detainees in Ansar Prison, on which the Israeli Supreme Court issued a judgment on 11 May 1983, the Israeli authorities asserted that the prisoners were 'hostile foreigners detained because they belong to the forces of terrorist organisations, or because of their connections or closeness to terrorist organisations'. Israel, while refusing them PoW status, claimed to observe 'humanitarian guidelines' of the 1949 Geneva Convention IV on civilians. For details of the case see Israel Yearbook on Human Rights 1983, vol. 13, pp. 360-64.

¹⁸ Chairman of the Joint Chiefs of Staff Instruction, Standing Rules of Engagement for US Forces, Ref. CJCSI 3121.01A, 15 January 2000, p. A-9. A similar but not identical statement had appeared in the Standing ROE of 1 October 1994 which this document replaces. A number of other US military-doctrinal statements are equally definite that US forces will always apply the law of armed conflict.

¹⁹ In its confidential messages to the US and UK governments on 28 September 2001, the ICRC stated that 'the use of nuclear weapons is incompatible with international humanitarian law'. This is undoubtedly wrong as a statement of law, and the wording was omitted in a revised text sent to governments on 6 October.

²⁰ See e.g. 'Afghanistan: ICRC calls on all parties to conflict to respect international humanitarian law', Communication to the press 01/47, ICRC, Geneva, 24 October 2001.

²¹ See esp. Hans-Peter Gasser, 'Internationalized Non-International Armed Conflicts: Case Studies of Afghanistan, Kampuchea and Lebanon', American University Law Review, vol. 33, no. 1, Fall 1983, pp. 145-61.

²² Afghanistan is nonetheless bound by the complete prohibition on possession and use of biological weapons in the 1972 Biological Weapons Convention, which it ratified on 26 March 1975. It is not a party to the 1993 Chemical Weapons Convention, which it signed on 14 January 1993 but never ratified.

²³ Eight reporters died in the period October-December 2001, several of them due to banditry rather than military operations. Information supplied by Nik Gowing of BBC World Television, 26 January 2002.

²⁴ 1949 Geneva Convention IV Relative to the Protection of Civilian Persons in Time of War, Part II (i.e. Articles 13-26). The issue of humanitarian relief is only touched on briefly in this essay as, while of critical importance in Afghanistan, only to a limited extent does it raise problems specific to anti-terrorist military operations.

²⁵ President Bush, televised address announcing the start of military strikes in Afghanistan, 7 October 2001. Text published in International Herald Tribune, 8 October 2001, p. 3.

²⁶ At all stages of the operations in Afghanistan, US military press briefings contained extensive and sometimes detailed references to the delivery of humanitarian aid. See e.g. the briefing by General Tommy Franks at Tampa, Florida, on 18 January 2002, http://www.defenselink.mil/news/Jan2002/briefings.html.

²⁷ Reports from UNHCR border monitors, summarized in a press briefing by a UNHCR spokesman in Geneva, 25 January 2002; and Afghanistan OCHA Situation Report No. 37, 29 January 2002, both on the UN Reliefweb site, http://www.reliefweb.int/w/rwb.nsf.

²⁸ Richard Myers, interview with This Week on ABC TV, 21 October 2001, http://www.defenselink.mil/news/Oct2001/briefings.html.

²⁹ A figure of 3,767 was given by Marc W. Herold, 'A Dossier on Civilian Victims of US Aerial Bombing of Afghanistan: A Comprehensive Accounting', 19 December 2001, University of New Hampshire website http://pubpages.unh.edu/~mwherold/ . There are certain updates on this site. This report was produced while the US bombing campaign (of which it is extremely critical) was ongoing. Its methodology is imperfect, because of the following factors: (1) The total figure is spuriously exact, and the calculation leading to it is not transparent. The author has informed me that the figure was not intended to suggest total accuracy. (2) Unavoidably, in view of time constraints, the study relied heavily on media reports, some of them dubious. (3) In some instances al-Qaeda deaths, and possibly Taliban deaths, may have been reported as civilian deaths. (4) It is probable that some civilian casualties of bombing went unreported and were thus omitted from the report. For a strong critique, see the paper by Jeffrey C. Isaac of Indiana University, 'Civilian Casualties in Afghanistan: The Limits of Herold's "Comprehensive Accounting"', 10 January 2002, available at http://www.indiana.edu/~iupolsci/docs/doc.htm .

³⁰ Rumsfeld, Interview on C-SPAN, 8 January 2002, http://www.defenselink.mil/news/Jan2002/briefings.html.

³¹ Many of the internally displaced in, and refugees from, Afghanistan testified eloquently to the disastrous effects of the bombing on civilians and their property. See e.g. Taghi Amirani's documentary film, The Dispossessed, made in November-December 2001, about the Makaki Camp in Nimruz Province near the Afghan-Iranian border. The camp was initially under Taliban, and then Northern Alliance, control.

³² Rumsfeld, Remarks outside ABC TV Studio, 28 October 2001, http://www.defenselink.mil/news/Oct2001/briefings.html.

³³ William M. Arkin, 'Fear of Civilian Deaths May Have Undermined Effort', Los Angeles Times, 16 January 2002, p. A12.

³⁴