International humanitarian lawFootnote 1 is predicated on a delicate equilibrium between the competing demands of military necessity and humanity.Footnote 2 In the development of any norm of international humanitarian law the challenge is, as stated in the 1868 St Petersburg Declaration, to establish ‘the technical limits at which the necessities of war ought to yield to the requirements of humanity’.Footnote 3 This subtle balancing act finds expression in Article 51(3) of Protocol I Additional to the Geneva Conventions of 1949,Footnote 4 which provides that civilians are protected from attack ‘unless and for such time as they take a direct part in hostilities’. By permitting the targeting of civilians who participate directly in hostilities, Article 51(3) is a fulcrum balancing the humanitarian impulse to protect civilians with the dictates of military necessity, which permit attacks on civilians who are harmful to the military.
Civilians – that is, those who are not combatants under Article 4A of the Third Geneva Convention of 1949Footnote 5 or Article 43 of Protocol IFootnote 6 – have come to play roles of increased military significance in armed conflict. Developments in weapons technology,Footnote 7 the asymmetric nature of many conflictsFootnote 8 and increased outsourcing of war-related work to private (civilian) contractorsFootnote 9 have seen growing classes of civilians become potentially harmful to enemy forces. While military exigencies may mean that such people are viewed as valuable targets, military necessity is subject to the laws of war,Footnote 10 including the general prohibition on attacking civilians. Determining the circumstances in which civilians lose immunity from attack for participating directly in hostilities becomes essential. Not only do civilians who participate directly in hostilities become legitimate targets,Footnote 11 they may also be prosecuted under national laws on the basis that they are not combatants who are entitled to so participate.Footnote 12
Although Article 51(3) is novel in its codification and phrasing of the ‘direct participation’ exception, the basic idea underpinning it – that humanity demands the protection of citizens, provided they are harmless – is not. The general concept that non-combatants who engage in hostile acts may be exposed to attack (and punishment) dates back several centuries. Against this historical backdrop, this paper will trace the factors and ‘mischiefs’ which influenced the formulation of Article 51(3) and which continue to affect its application.
Recognizing that Article 51(3) closes the conceptual gap between civilians entitled to protection from attack and combatants permitted to participate directly in hostilities, this inquiry commences by considering the development of these interlinked categories of persons. Starting with Grotius in the seventeenth century and proceeding to Rousseau in the eighteenth, it traces the limited right to participate in hostilities and the immunity of non-combatants. This paper illustrates that the development of the two categories has been heavily informed by the paradigms of war in which they have arisen, and also by notions of guilt and innocence, military necessity, chivalry and humanity. Having established these foundations, the second part of the paper considers the challenges posed by civilians participating in hostilities in the nineteenth and twentieth centuries and the international legal responses to such participation. While it is beyond the capacity of this paper to critique these activities or the responses to them in any detail, the hidden factors behind the laws, and how they fit within the dominant paradigm of conflict, are examined. The third part of this paper examines the legislative path to the introduction of Article 51(3) and the increased legal protection it provides for civilians. Such an examination is relevant not only for the current interpretation of the article,Footnote 13 but also for an understanding of the factors underpinning it. Finally, drawing on these factors, this paper analyses some trends in the development of Article 51(3) and the compatibility of these trends with changes in the nature of contemporary armed conflict, particularly the shifting demands of military necessity.
The paper does not consider the specific challenges raised by modern warfare in depth; this has been done elsewhere.Footnote 14 Rather, in recognition of the changing methods of warfare adopted over the last three decades, it considers the relevance of some of the major assumptions and biases which underpin Protocol I. In so doing, this paper provides a useful backdrop against which to view current debates about the circumstances in which civilians should forfeit their immunity as non-combatants for taking a direct part in hostilities under Article 51(3).
Background: Grotius to Rousseau
This paper's inquiry into the development of the ‘direct participation’ exception to civilian immunity begins with Grotius, since his work, along with that of Francisco de Vitoria, is acknowledged as the analytical basis of the contemporary law of land warfare.Footnote 15 Grotius, writing in the midst of the Thirty Years War in 1625, recorded the state of the law of nations – the nascent international law – as he perceived it. By the seventeenth century, modern nation-states, although in an incipient form, had emerged as the only legitimate authorities in Europe that could make war on their neighbours and suppress rebellion within their own realms.Footnote 16 Their status as such was cemented with the adoption, in 1648, of the Peace of Westphalia, which abolished private armies and conferred a legal monopoly on states for the maintenance of armies and for fighting wars.Footnote 17 Developing on the ‘just war’ theories, ideas of military honour and chivalry required that wars be fought ‘publicly and openly’.Footnote 18 Those who fought in wars without the authority of the state were considered marauders, brigands and freebooters outside the law of nations,Footnote 19 and perfidy was repugnant to the fighting classes.Footnote 20 In accordance with the ideas of the medieval law of war (the jus militare), those who engaged in other than ‘open and public wars’ met short shrift at the hands of the fighting classes.Footnote 21 It is against this background that the writings of Grotius are considered.
Grotius, the practice of nations and restraint in war
Grotius' starting point was his conception of the effect of a declaration of war on a sovereign's subjects. In his view, a public war (that is, a war waged between two or more sovereign authorities) was ‘declared at the same time … upon all a sovereign's subjects’.Footnote 22 Accordingly, the ‘right to kill’ which arises in warFootnote 23 extended ‘not only to those who actually bear arms, or are subjects of him that stirs up the war, but in addition to all persons who are in the enemy's territory’.Footnote 24 Indeed, Grotius explicitly stated that the slaughter of infants, women, old men, hostages and ‘suppliants’ seeking to surrender was permissible in a public war.Footnote 25 He found ample evidence of the slaughter of non-combatants in the writings of ancient scholars and the ‘common practice of nations’.Footnote 26
Grotius explicitly distinguished, however, between actions which are ‘permissible’ according to the law of nations (such as those outlined above) and those which were ‘right’, ‘praiseworthy’ or ‘honourable’.Footnote 27 He wrote,
when I first set out to explain this part of the law of nations I bore witness that many things are said to be ‘lawful’ or ‘permissible’ for the reason that they are done with impunity, in part also because coactive tribunals lend to them their authority; things which, nevertheless, either deviate from the rule of right (whether this has its basis in law strictly so called, or in the admonitions of other virtues), or at any rate may be omitted on higher grounds and with greater praise among good men.Footnote 28
His treatment of legally permissible actions may thus be seen as a forced concession to past verdict and practice.Footnote 29
On the question of what is ‘right’ or ‘honourable’ in war (or the lex ferenda), Grotius stated as a basic principle, ‘One must take care, so far as is possible, to prevent the death of innocent persons, even by accident.’Footnote 30 While he did not expressly define ‘innocent persons’, he appears to have been referring to those who are unarmedFootnote 31 and have not committed any serious crimes.Footnote 32 Citing Livy and Josephus, Grotius observed,
By the law of war armed men and those who offer resistance are killed. … [I]t is right that in war those who have taken up arms should pay the penalty, but that the guiltless should not be injured.Footnote 33
In Grotius' view, children should always be spared, as should women, unless they ‘have committed a crime which ought to be punished in a special manner, or unless they take the place of men’.Footnote 34 Similarly, men ‘whose manner of life is opposed to war’ – specifically those who perform religious duties or men of letters – should be spared.Footnote 35
In urging restraint in relation to these categories of civilians, unless they took up arms, Grotius trod a path well worn by earlier commentators such as Gentili, Suarez and Vitoria.Footnote 36 It was, however, Grotius' reluctant view that the lex lata Footnote 37 permitted the slaughter of these categories of civilians as they were, according to Grotius' conception, ‘enemies’ in a public war.
Rousseau's maxim
From Grotius' concept of the effect of a declaration of war, the significance of Rousseau's commentary becomes apparent. In contrast with Grotius, Rousseau took the view that war is a relation between governments, involving the citizens of a state only ‘accidentally’. Writing in 1762, Rousseau said,
War, then, is not a relationship between man and man, but between State and State, in which private persons are only enemies accidentally, not as men, nor even as citizens, but simply as soldiers; not as members of their fatherland, but as its defenders …Footnote 38
Rousseau's maxim explicitly recognized that non-combatant citizens are not, in any real sense, the enemies of an opposing army and should not be made its object.Footnote 39 Prior to Rousseau's contribution, the separate identity of the individual and his or her state was not recognized by the law of nations; the identification of one with the other was total.Footnote 40 Rousseau's maxim is, accordingly, seen by many as forming the modern jurisprudential basis for the principle of non-combatant immunity,Footnote 41 or as Best wryly put it, ‘the non-combatant's supreme talisman’.Footnote 42
Rousseau's statement appears to have reflected contemporary practice at the time he was writing.Footnote 43 Cassese has observed that during the period from 1648 to 1789, war became very much a game between professionals without a great deal of involvement of the civilian population.Footnote 44 In contrast with the bloodiness of the Thirty Years War, during which Grotius wrote his treatise, wars in Rousseau's era were fought by professional armies, the expense of which kept conflicts small.Footnote 45 Moreover, military professionalism ensured that a soldier's focus was on mastering armed opponents, not on the civilian population.Footnote 46 Publicists such as Vattel were already cautiously moving towards a judicial statement of non-combatant immunity to match the practical immunity increasingly being achieved in conflict.Footnote 47 Rousseau's statement, however, was appealing for its ‘surpassing simplicity’.Footnote 48 It set up an unbridgeable conceptual divide between combatants and non-combatants.Footnote 49
Over the years, many people have criticized aspects of Rousseau's maxim.Footnote 50 It is outside the scope of this paper to examine the merits of these criticisms in detail. Suffice it to note that, although the maxim was, and is, far from universally accepted,Footnote 51 its influence is undeniable.Footnote 52 The conceptual gulf it established, coupled with the idea (alluded to by Rousseau and codified in 1868 in the St Petersburg Declaration)Footnote 53 that the only legitimate object of war is to weaken the military forces of the enemy, brought Grotius' conception of the lex ferenda to life. In this way, the concept of innocence, on which Grotius and his contemporaries had focused, expanded and metamorphosed into notions of civilian status and the protection of civilians from attack.Footnote 54 Against the above backdrop, the second part of this paper considers some of the different ways in which non-combatants have participated in hostilities throughout history, and how the laws of war responded to them.
Law-making and the ‘wars of nations’
Beginning with the revolutionary wars of the late eighteenth century and early nineteenth century, war passed through a transition from the dynastic war of kings to the war of nations-at-arms, in which entire populations were mobilized to support the war effort.Footnote 55 An increasing range of activities involving peasants, such as providing food to partisans and passing on information about the occupying army, came to be regarded as political participation in conflict.Footnote 56 Armed resistance increased alongside political participation, and took many forms, including spontaneous armed resistance, organized acts of resistance in the form of guerrillas and francs-tireurs,Footnote 57 and the levée en masse.Footnote 58 Any currency Rousseau's maxim once held was undermined.
During the late eighteenth and nineteenth centuries, the practical response to non-uniformed fighters was usually ferocious.Footnote 59 Due to the ‘treacherous’ threat they posed, armed civilians, regardless of gender, were attacked with ‘a draconian severity’ by opposing armed forces.Footnote 60 Moreover, according to Nabulsi, there was no real legal or practical distinction between non-violent political behaviour and violent resistance.Footnote 61 Peasants who, for instance, passed on information about the occupying army, hid escaped prisoners of war or fed illicit fighters, were deemed as criminal as those who physically killed soldiers from the occupying force.Footnote 62 They were also exposed to risks, including being shot, for certain conduct, such as hiding their own crops.Footnote 63 Due to the relative paucity of historical records on political resistance in the nineteenth century,Footnote 64 it is difficult to determine whether political resisters were targeted directly or executed as a matter of law enforcement. From the few records that do exist, however, it appears that, primarily, they were executed as a matter of law enforcement.Footnote 65
How the positive laws of war would deal with the various forms of civilian conduct was yet to be determined. The growing involvement of civilians in political life, including in armed conflict, compelled governments in the nineteenth century to discuss the question of ‘normalizing’ their involvement in war.Footnote 66 As discussed below, European countries which relied more heavily on the civilian population in armed conflict advocated their recognition as legitimate combatants. The debates around this issue were heavily influenced by the work of Dr Francis Lieber, a German émigré to America.
Lieber and his Code
It was not only in Europe that the participation of non-combatants in war demanded attention. Across the Atlantic, the methods used by the South in the American Civil War compelled the Union government to find ways of addressing the legal status of guerrilla warfare.Footnote 67 According to Hartigan, in the early years of the conflict the Union army tended to equate all irregular troops with ‘guerrillas’, who in turn were classified as criminals.Footnote 68 As in Europe during the revolutionary wars, this generalization applied not only to those who bore arms for the South, but also to non-combatant civilians who either actively or passively supported irregular troops.Footnote 69 The rebel authorities, on the other hand, claimed the right to engage in guerrilla warfare and be treated as combatants.Footnote 70 Writers on the laws of war had not dealt with the status of these troops in any comprehensive manner.Footnote 71 In apparent recognition of the conundrum, Henry Wager Halleck, the general-in-chief of the Union armies, wrote to Lieber in 1862 requesting his assistance in defining guerrilla warfare.Footnote 72 Lieber obliged, initially producing an essay on the topicFootnote 73 and later completing a more comprehensive field manual. Due to Lieber's substantial influence on the subsequent codification of the laws of war, it is worthwhile examining his contribution in some detail.
Lieber's essay on guerrilla warfare and the laws of war dealt comprehensively with the treatment of ‘armed parties loosely attached to the main body of the army, or altogether unconnected with it’.Footnote 74 Lieber observed that while several categories of armed bands (the freebooter, marauder, brigand, partisan, free corps, spy, war-rebel or conspirator, highway robber and levée en masse or ‘arming of the peasants’) were dealt with by the laws of war, ‘guerrilla parties’Footnote 75 were not.Footnote 76 Guerrillas, according to Lieber, are ‘peculiarly dangerous, because they easily evade pursuit, and by laying down their arms become insidious enemies; because they cannot otherwise subsist than by rapine, and almost always degenerate into simple robbers or brigands’.Footnote 77 In reflection of this ‘peculiar’ threat, Lieber argued that, when guerrilla parties aid the main army of the belligerent in ‘fair fight and open warfare’, they should be treated as regular partisans.Footnote 78 If, however, they resort to ‘occasional fighting and the occasional assuming of peaceful habits, and to brigandage’, they should not be protected by the laws of war.Footnote 79
Lieber's treatment of ‘guerrilla parties’ may be contrasted with his treatment of the levée en masse. After noting that most constitutions enshrined the right of the people to possess and use arms, Lieber concluded that it was generally agreed that the rising of the people openly to repel invasion entitled them to the privileges of the laws of war.Footnote 80 The absence of a uniform was immaterial, provided such absence was not used for the purpose of concealment or disguise.Footnote 81 Lieber thus gave emphasis to the idea of ‘openness’, a dominant indicator of a lawful war in medieval and post-Westphalian warfare.Footnote 82
Lieber subsequently completed his manual on the laws of war, which became ‘General Orders, no. 100: Instructions for the Armies of the United States in the Field’.Footnote 83 Lieber, whose political affiliations were with the anti-slavery North, felt that in order to preserve the Union and free the slaves, it was essential to bring discipline to the Union army and to define precisely the status of the enemy troops and greater population.Footnote 84 In the light of this attitude, it is, perhaps, unsurprising that his Code constituted ‘an admixture of military sternness with basic humanitarianism’.Footnote 85 This delicate balance is evident in Article 15 of the Code, which codified the permissible destruction of life during war. It provided,
Military necessityFootnote 86 admits of all direct destruction of life or limb of armed enemies, and of other persons whose destruction is incidentally unavoidable in the armed contests of the war …
The Code defined the term ‘enemy’ to include citizens.Footnote 87 Article 15, therefore, emphasized the status of armed enemy citizens as legitimate targets. Other dangerous enemies or people of importance to the government could lawfully be captured.Footnote 88
Article 22 of the Code provided for the immunity of civilians. It stated, ‘The principle has been more and more acknowledged that the unarmed citizen is to be spared in person, property, and honour as much as the exigencies of war will admit.’Footnote 89 The question of precisely how much ‘the exigencies of war will admit’ was not further developed in the Lieber Code. It was clear from Article 15 that all armed enemy citizens may be directly attacked.Footnote 90 However, the Code was less direct on the protection from attack provided to hostile, but unarmed, civilians.Footnote 91 Some, such as spies and certain types of war traitors,Footnote 92 were to be dealt with under severe rules of law enforcement, which included capture and execution.Footnote 93 Moreover, those who ‘held intercourse’ with the enemy were also treated as war traitors and faced ‘severe’ punishment.Footnote 94 According to General Order No. 30, those who harboured and fed the enemy were to ‘suffer death’ or such other punishment as ordered by a court martial.Footnote 95 This Order suggests that the ‘military exigencies’ referred to in Article 22 of the Lieber Code would, under the banner of law enforcement, have permitted the killing of certain unarmed hostile citizens for their participation in hostilities.
Brussels to The Hague
Lieber's authority as to the laws of war was held by his contemporaries in high regard,Footnote 96 and his Code was widely adopted in Europe.Footnote 97 The Lieber Code formed the basis of the draft text for the Brussels Conference in 1874, which was convened at the behest of Emperor Alexander of Russia.Footnote 98 The Conference gave rise to a Protocol and a Declaration on the laws of war.Footnote 99 The Declaration was not ratified due to the unwillingness of the ‘great powers’, who, according to Jochnick and Normand, considered it too ‘humanitarian’.Footnote 100 Many of its rules were, nonetheless, reproduced in military manuals.Footnote 101
The Brussels Protocol adopted, in slightly looser terms, the principle of restraint laid down six years earlier in the St Petersburg Declaration,Footnote 102 stating that the only legitimate object of war is to weaken the enemy without inflicting unnecessary suffering.Footnote 103 The Brussels Declaration went on to specify, in Articles 9 to 11, the classes of people who should be recognized as ‘belligerents’ under the laws of war. Under Article 9, the laws of war applied to armies, and militia and volunteer corps who were commanded by a responsible person, had a fixed distinctive emblem recognisable at a distance, carried arms openly, and conducted their operations in accordance with the laws of war. Like the Lieber Code, this definition of ‘belligerent’ focused on the requirement of ‘openness’ of warfare and reflected an aversion for perfidious methods. According to Risley, writing at the close of the nineteenth century, Article 9 accurately expressed the generally accepted laws of war.Footnote 104 Article 10 conferred belligerent status on the members of a levée en masse, while Article 11 granted prisoner of war status to non-combatant members of the armed forces.
On the controversial issue of those outside the armed forces or the levée en masse who resisted invasion or occupation, the Brussels Declaration said nothing. This omission was not for want of trying. The Russian draft text, for instance, proposed that individuals not qualifying as combatants, but who ‘at one time take part independently in the operations of war, and at another return to their pacific occupations …, do not enjoy the rights of belligerents, and are amenable, in case of capture, to military justice’.Footnote 105 The draft article was, however, withdrawn as a consequence of the opposition of the smaller powers, many of whom viewed the Brussels Conference as an attempt by the military powers to prevent resistance being offered by the civilian population against an invader.Footnote 106 The Brussels Declaration did not, therefore, explicitly forbid guerrilla warfare and other forms of civilian participation in hostilities falling outside Article 9.Footnote 107 Rather, it merely enumerated conditions (in Articles 9 and 10) under which combatants were to be regarded as lawful.Footnote 108 Thus, in the absence of rules protecting civilians, individuals who participated in hostilities in any way continued to do so at their own risk.Footnote 109
The Brussels Declaration, while not ratified, provided an important basis for the work of the jurists of the Institute of International Law, who produced the ‘Oxford Manual’ in 1880.Footnote 110 The Oxford Manual, which purported to codify ‘the accepted ideas of our age so far as this has appeared allowable and practicable’,Footnote 111 provided in Article 1,
The state of war does not admit of acts of violence, save between the armed forces of belligerent States. Persons not forming part of a belligerent armed force should abstain from such acts.Footnote 112
The armed forces of a state were defined in Article 2 and included bodies other than the regular army which, among other things, wore a uniform or ‘fixed distinctive emblem’ and carried arms openly.Footnote 113 The manual went on to forbid the ‘maltreatment’ of ‘inoffensive populations’, on the basis that ‘The contest (is) carried on by “armed forces” only.’Footnote 114 Like the Brussels Declaration, however, the Manual did not give further consideration to the question of people who fell in the gap between the ‘armed force’ and ‘inoffensive population’, such as civilians who engaged in hostile acts, whether bearing arms or not. The exception to this rule was the treatment of individuals as spies, who could not demand treatment as prisoners of war.Footnote 115
Although the Manual was ignored by most countries and derided by its contemporaries,Footnote 116 it formed, along with the Brussels Declaration, the basis of the Hague Conventions on the conduct of land warfare which were adopted in 1899 and 1907.Footnote 117 These Conventions were, according to their preambles, ‘inspired by the desire to diminish the evils of war, as far as military requirements permit’.Footnote 118 Both Conventions adopted a set of Regulations Respecting the Laws and Customs of War on Land which were in most material respects identical. As Best has observed, the Hague Regulations have provided the basis for the laws of land warfare ever since.Footnote 119
In contrast with the Oxford Manual and, to a lesser extent, the Brussels Protocol,Footnote 120 the Hague Conventions did not refer specifically to the immunity of civilians from direct attack.Footnote 121 The definition of ‘belligerent’ used in the Brussels Declaration was, however, reproduced in the Hague Regulations without change.Footnote 122 Participants in the levée en masse who carried arms openly and respected the laws of war were likewise considered belligerents entitled to the rights and subject to the duties of the laws of war.Footnote 123 The continued focus on openness was also apparent in the definition of spies. A spy was a person who, acting clandestinely or on false pretences, attempted to obtain information in the zone of operations of a belligerent, with the intention of communicating it to the hostile party.Footnote 124 Soldiers not wearing a disguise who managed to penetrate the zone of operations of the hostile army to obtain information were not considered spies. Nor were civilians who ‘openly’ carried out their mission to deliver despatches.Footnote 125
The Regulations did not expressly address the fate of other civilians who, falling short of the definition of belligerent, took up arms against an invading or occupying power, or resisted in other ways. This issue therefore remained the province of customary international law.Footnote 126 The 1866 edition of Wheaton's classic text, Elements of International Law, stated the law as follows:
[N]o use of force against an enemy is lawful, unless it is necessary to accomplish the purposes of war. The custom of civilized nations, founded upon this principle, has therefore exempted … all … public or private individuals engaged in the ordinary civil pursuits of life, from the direct effect of military operations, unless actually taken in arms, or guilty of some misconduct in violation of the usages of war, by which they forfeit their immunity.Footnote 127
The editor's accompanying note elaborated that non-combatants who ‘make forcible resistance, or violate the mild rules of modern warfare, give military information to their friends, or obstruct the forces in possession … are liable to be treated as combatants’.Footnote 128 Thus, hostile but unarmed civilians could lawfully be targeted for their participation in conflict.
Risley, writing in the closing years of the nineteenth century, appears, at first blush, to have interpreted the law more restrictively in favour of civilians. In language reminiscent of Common Article 3 of the four Geneva Conventions of 1949,Footnote 129 he argued that subjects of a belligerent ‘are not liable to be killed or taken as prisoners of war so long as they do not actively engage in hostilities’.Footnote 130 Further examination of Risley's argument, however, suggests that participation in hostilities need not in fact have been ‘active’ for a non-combatant to lose immunity. Risley contended that the immunity of non-combatants was granted on the implicit understanding that the distinction between the classes of combatant and non-combatant be maintained in good faith.Footnote 131 Accordingly, immunity was ‘forfeited by a non-combatant who commits any hostile act’.Footnote 132 He stated,
Combatants must be open enemies, known and knowable, and non-combatants must be harmless. As soon as an individual ceases to be harmless, he ceases to be a non-combatant, and must be reckoned a combatant; and unless he bears the distinguishing marks of an open combatant, he puts himself outside the laws of war, and is, if captured, liable to be shot as a bandit instead of detained as a prisoner of war.Footnote 133
In accordance with Risley's analysis, civilians who ‘actively engaged in hostilities’, ‘ceased to be harmless’, or ‘[committed] any hostile acts’ would have become subject to attack.
The breadth of, and lack of precision in, these phrases would have provided cold comfort to civilians in conflict zones at the turn of the twentieth century. Regrettably, significant gaps in the positive law, such as the lack of a definition of ‘civilian’ and the failure of the world's powers to agree on the status of individuals who took up arms or otherwise participated in hostilities without meeting the requirements of belligerent status,Footnote 134 persisted for some time. These deficiencies opened the door for arguments in the early part of the twentieth century that all non-combatants whose destruction would be of military value should lose their immunity from attack.
The challenges of ‘total war’
The twentieth century saw a blurring of the distinction between combatants and civilians in armed conflict.Footnote 135 With no positive law protecting (or, indeed, defining) civilians, their immunity from attack was precariousFootnote 136 and vulnerable to arguments that military necessity permitted them to be targeted. Wright observed in his authoritative A Study of War in 1942,
As the proportion of the population contributing directly or indirectly to the making of the policy and the military of the enemy have increased, economic and propaganda measures have gained in relative importance. Attacks upon civilians … have increased under the plea that traditional rules must be applied in the light of ‘military necessity’ as developed under changing technical conditions.Footnote 137
As discussed below, by the Second World War arguments of military necessity were used to justify widespread bombing of civilian and industrial targets. The notion that humanity required the protection of ‘harmless’ civilians, which had come to be the norm at the turn of the twentieth century,Footnote 138 proved largely ineffective in sparing civilian populations from attack.
Quasi-combatants
With increasingly sophisticated methods of warfare, the civilian population became more involved in the war-making machine, including in supplying arms and ammunitions.Footnote 139 One of the great debates during the two world wars revolved around the status of armament and munitions workers. Such persons did not fall within the definitions of ‘combatant’ or ‘non-combatant members of the armed forces’ under the Hague Regulations.Footnote 140 In the absence of a positive definition of ‘civilian’, it was also logically problematic to regard them as such, given that their workplaces constituted legitimate military targets under customary international law.Footnote 141 The need to clarify the status of such workers, whose contribution to the war effort was arguably on a par with that of soldiers,Footnote 142 was compounded by the advent of aerial warfare. Aerial warfare enabled belligerents to attack military objectives – such as munitions factories – on a larger and less discriminate scale than previously.Footnote 143 The status of those inside the factories became crucial.
Accordingly, as early as 1916 arguments were made that armament workers should be treated as a category of quasi-combatants who lost their immunity as non-combatants and should be treated as combatants while they were engaged in activities harmful to the enemy. In an article in the Revue de Droit International Rolland wrote that armament workers
… occupy a position intermediate between the combatants proper and the non-combatants who are still employed on their peacetime trades and professions. The reasons for sparing them are losing force. Fundamentally they are almost in exactly the same position as the men of the auxiliary services of the armies, and the latter are certainly legitimate objects of attack.Footnote 144
This argument was taken up with great vigour by some scholars over the next thirty years.Footnote 145 In 1938 Spaight famously urged international law to ‘move with the times’ by accepting that ‘the old clear-cut division of enemy individuals into combatants and non-combatants is no longer tenable without some qualification’.Footnote 146 This argument was, moreover, extended by some to civilian workers who directly supported the war effort, such as those who transported munitions.Footnote 147
The Draft Convention for the Protection of Civilian Populations Against New Engines of War, approved in principle by the International Law Association in 1938,Footnote 148 reflected Spaight's argument that munitions factory workers should be, while at work, legitimate objects of war. It protected the ‘civilian population’ from ‘[forming] the object of an act of war’. ‘Civilian population’ was defined as ‘all those not enlisted in any branch of the combatant services nor for the time being employed or occupied in any belligerent establishment’. ‘Belligerent establishments’ were defined to include ‘military, naval or air establishment, or barracks, arsenal, munition stores or factories, aerodromes or aeroplane workshops or ships of war, naval dockyards, forts, or fortifications for defensive or offensive purposes, or entrenchments’.Footnote 149 Thus emerged a precursor to Article 51(3) in the form of the quasi-combatant; that is, a person ‘for the time being employed or occupied in a belligerent establishment’, such as a munitions store or factory. On its face, it appears that once such people ceased employment within belligerent establishments or, arguably, returned to their homes, they would once again receive the protection offered to the civilian population. The Draft Convention, however, was neither signed nor adopted due to the onset of the Second World War.Footnote 150 Accordingly, the question of whether, and to what extent, munitions workers constituted a legitimate target remained a topic of dispute for the duration of the Second World WarFootnote 151 and, indeed, well beyond.Footnote 152
Contribution to the war effort
Armament and munitions workers were not the only civilians in danger of attack, particularly from the air, during the two world wars. By the dawn of the Second World War, war had come to be viewed as a totalitarian affair to which all a nation's citizens contributed through industry and morale.Footnote 153 While states initially sought to avoid the direct targeting of civilians, as the conflict progressed the perceived demands of military necessity eroded this standard.Footnote 154 Consequently, the area and extent of aerial bombardment continually expanded during the Second World War.Footnote 155 As Nurick stated,
At first, the bombing was confined to military objectives in the actual theater of operations. Then bombing was extended to military objectives, such as factories, communications, and the like in the rear of the enemy's lines, with some regard for the civilian population. Finally, it was extended in many instances to the bombing of cities in order to affect the morale of the civilians.Footnote 156
Best aptly observed that plausible economic reasons for injuring civilians had multiplied, and ‘their own apparently willing participation in the decisions to make or to continue war seductively suggested that they deserved to be damaged’.Footnote 157 The widespread practice of saturation bombing of civilian targets made it difficult to assert that the direct targeting of civilians remained contrary to international law.Footnote 158 Scholarly consensus existed, however, on one point: the illegality of targeting the civilian population for the mere purpose of terrorizing the population.Footnote 159
The practice of bombing civilians to negate their general contribution to the war effort or to terrorize them into submitting had a significant effect on the development of Protocol I; terror bombing was clearly prohibited by Article 51(2). Further, as will be discussed in the following section of this paper, the requirement in Article 51(3) that participation in hostilities be direct appears to have been formulated largely to ensure that general contribution to the war effort not be sufficient to expose civilians to attack.Footnote 160
The Geneva Conventions of 1949
Enemy attacks by belligerents, whether directly aimed or incidental, were, according to Best, one of the principal causes of civilian suffering in 1939–45.Footnote 161 Thus, when the International Committee of the Red CrossFootnote 162 took up its longstanding project to improve the protection of civilians following the Second World War, the issue was in dire need of attention.Footnote 163 As a result of political circumstances, however, the issue of enemy attacks on civilians was not taken up.Footnote 164 Rather, the 1949 Diplomatic Conference was tasked only with updating the ‘Geneva law’Footnote 165 and not the Hague Regulations governing the conduct of military operations.Footnote 166 The ICRC observed in its Commentary to the Fourth Geneva ConventionFootnote 167 that any provisions in the Convention's draft text designed to protect the civilian population from the dangers of military operations were systematically removed.Footnote 168 Accordingly, the Fourth Geneva Convention only protects those who ‘find themselves … in the hands of a party to the conflict of which they are not nationals’Footnote 169 from arbitrary actions by the enemy. It does not protect civilians against the ‘whole series of dangers which threaten them in warfare’.Footnote 170
There is, nonetheless, one provision of the four Geneva Conventions of 1949 which deserves mention in the present context: Common Article 3. This Article sets out minimum guarantees applicable in non-international armed conflict and protects ‘persons taking no active part in the hostilities’ against ‘violence to life and person, in particular murder of all kinds’.Footnote 171 Although one could argue that Common Article 3 sets out the principle of distinction later codified in the Additional Protocols, this was probably not the intention behind the provision, given the Conference's focus on the Geneva law.Footnote 172 Common Article 3 is, however, significant for its use of the language ‘taking no active part in the hostilities’, a precursor to the phrase later adopted in Article 51(3) of Protocol I and Article 13(3) of Protocol II.Footnote 173
The road to Article 51(3)
Gathering momentum: the work of the ICRC
The failure to update the Hague Regulations following the Second World War left a significant void in the codified laws of war. In the light of this, in 1956 the ICRC moved beyond its traditional focus on ‘Geneva law’, promulgating the Draft Rules for the Limitation of the Dangers incurred by the Civilian Population in Time of War.Footnote 174 The 1956 Draft Rules reaffirmed some principles of customary law and offered concrete solutions to resolve problems resulting from changes and developments in weaponry.Footnote 175 They were directed primarily towards protecting ‘civilian populations efficiently from the dangers of atomic, chemical and bacteriological warfare’.Footnote 176 Ultimately the 1956 Draft Rules made no headway with the governments intended to implement them.Footnote 177 Nonetheless, they remained an important document in the push for an authoritative revision of the laws of war given the ‘ever more distressing varieties of war and war techniques’.Footnote 178
The 1956 Draft Rules sought to require parties to ‘confine their operations to the destruction of … military resources, and leave the civilian population outside the sphere of armed attacks’.Footnote 179 ‘Civilian population’ was negatively (and somewhat awkwardly) defined as all persons who were not
(a) Members of the armed forces, or of their auxiliary or complementary organizations; and
(b) Persons who do not belong to the forces referred to above, but nevertheless take part in the fighting.Footnote 180
Under this definition, individuals who took part in the fighting were not considered civilians, even if their usual activities were primarily peaceful. The range of people protected by civilian immunity under the 1956 Draft Rules was, accordingly, considerably narrower than those later protected under Article 51(3). As discussed later in this paper, one of the reasons for broadening the protection afforded to civilians under Article 51(3) was the increased use of citizens engaged in guerrilla warfare in the years preceding 1977.
The ICRC's next, and more subdued, attempt to revise the law of armed conflict was manifested in a draft resolution, which was ultimately adopted by the Twentieth International Red Cross Conference in 1965 in Vienna.Footnote 181 The resolution on the ‘protection of civilian populations against the dangers of indiscriminate warfare’ relevantly provided that
distinction must be made at all times between persons taking part in hostilities and members of the civilian population to the effect that the latter be spared as much as possible.Footnote 182
Like the 1956 Draft Rules, the resolution, on its face, provided no protection for civilians who intermittently took part in hostilities. According to Pictet's 1966 analysis of the resolution, the principle it espoused was one of the ‘general principles of customary law which now regulate the question’.Footnote 183 The resolution was, Pictet observed, ‘the only pronouncement of the kind made by an assembly in which governments are represented since the Second World War’.Footnote 184 However, no further moves were made to convert the resolution into a binding agreement.Footnote 185
It was not until the 1968 Teheran United Nations International Conference on Human RightsFootnote 186 that the need to clarify the rules protecting the civilian population gained the necessary traction within the international community.Footnote 187 One of the dominant themes at the conference was the law relating to guerrilla warfare.Footnote 188 Thanks largely to the behind-the-scenes work of the International Commission of Jurists, the conference resulted in the adoption of Resolution 2444, ‘Respect for Human Rights in Armed Conflicts’.Footnote 189 Resolution 2444 affirmed the principle of distinction exactly as drafted in the 1965 Red Cross resolution. Moreover, the resolution from Teheran provided the impetus needed for the International Conference of the Red Cross to request, in 1969, the ICRC to ‘[propose], as soon as possible, concrete rules which could supplement the existing humanitarian law’.Footnote 190
The negotiation of Article 51(3)
Beginning in 1971, the ICRC held a series of Conferences of Government Experts on the Reaffirmation and Development of International Humanitarian Principles in Armed Conflicts.Footnote 191 The conferences set an essentially conservative path for the future codification of the laws of war by agreeing not to revise or ‘overhaul’ the Geneva Conventions of 1949, but rather to reaffirm them.Footnote 192 Moreover, the government experts rejected the proposal of the ICRC to deal with guerrilla warfare sui generis via a specific protocol, preferring for it to be addressed in the context of all the other forms of warfare.Footnote 193 Consequently, in 1973 the ICRC distributed drafts of two protocols which built on the Geneva Conventions and addressed the question of guerrilla warfare alongside the other forms of warfare.Footnote 194
The Draft Protocols formed the basis for Protocols I and II. A number of the provisions now found in Protocol I relating to the protection of the civilian population underwent only minor amendment during the negotiation process. For example, Draft Protocol I (on international armed conflict) set out the expansive, and much needed,Footnote 195 definition of ‘civilian’ now found in Article 50.Footnote 196 ‘Civilian’ was, and remains, defined as any person who is not a combatant within the meaning of Article 43 of Protocol I or Article 4A of the Third Geneva Convention.Footnote 197 Further, Draft Protocol I expressly prohibited making the civilian population or individual civilians the object of attack, a rule which is now found in Article 51(2).Footnote 198 This rule has, of course, been made subject to the exception now set out in Article 51(3) of Protocol I involving civilians who participate directly in hostilities.
Draft Article 46(2), the predecessor to Article 51(3), provided that ‘[c]ivilians shall enjoy the protection afforded by this Article unless and for such time as they take a direct part in hostilities.’Footnote 199 In contrast with the principle as set out in earlier resolutions, draft Article 46(2) made it clear that civilians lost their immunity only for the period during which they took part in hostilities.Footnote 200 Once they returned to peaceful activities they would once again be protected by their civilian status. In so providing, draft Article 46(2) shifted the law's balance between military necessity and humanity further towards the latter than previous formulations of the norm. This progression was consistent with the purpose of draft Article 46 as expressed in the provision's heading, ‘Protection of the civilian population’.Footnote 201
True to the observation by Suter that the law tends to be concerned with the last war,Footnote 202 discussions during negotiations seem to have focused on civilians in two historical contexts. On the one hand, a number of states referred to the need to increase the protection afforded to the civilian population in the light of the experience of the Second World War.Footnote 203 Other delegations focused attention on the protection of civilians and guerrilla fighters in the context of wars of national liberation. The Chinese representative, for example, argued that
People's militia and guerrilla fighters in wars of national liberation should be protected, since they were basically civilians who had been forced to take up arms in self-defence against imperialist repression in order to win independence and safeguard their right to survival. When not participating directly in military operations, members of people's militia or guerrilla movements should have civilian status and benefit from the protection granted to civilians.Footnote 204
Similar sentiments advocating the broad protection of civilian populations fighting wars of national liberation were expressed by other states.Footnote 205 It appears that both sets of concerns were allayed by draft Article 46(2), as it underwent only minor change during the negotiation processFootnote 206 and was adopted by consensus in the Third Committee.Footnote 207
The general acceptance of Article 46(2) may have been partially due to the malleability of the phrase, ‘direct participation in hostilities’.Footnote 208 The ICRC Commentary on Draft Protocol I stated,
What should be understood by direct part in hostilities? The expression covers acts of war intended by their nature or purpose to strike at the personnel and matériel of enemy armed forces.Footnote 209
The meaning was not, however, so broad as to include a civilian's participation in the ‘war effort’. The commentary to the draft expressly distinguished the ‘direct part which civilians might take in hostilities’ from the ‘part in the war effort which they are called upon to carry out at highly different levels’.Footnote 210 The latter should not expose a civilian to attack, because ‘in modern warfare, all the nation's activities contribute in some way or other, to the pursuit of hostilities, and even the people's morale plays its part in this context’.Footnote 211 One might surmise from this discussion that the adjective ‘direct’ was included for the primary purpose of ensuring that general contribution to the war effort was excluded as a ground for the loss of civilian immunity.
The malleability of the phrase ‘direct participation in hostilities’ is apparent from discussions surrounding an alternative proposal advanced during negotiations about draft Article 46(2). Brazil, Canada, the German Democratic Republic (GDR) (East Germany) and Nicaragua suggested rewording draft Article 46(2) to read, ‘Civilians shall enjoy the protection afforded by this Section except when they commit hostile acts or take a direct part in military operations.’Footnote 212 This proposal appears to have been considered by some delegations as being synonymous with draft Article 46(2). The Danish representative, for example, interpreted the proposed amendment as having the same meaning as draft Article 46(2).Footnote 213 Similarly, the Australian representative considered the proposal superfluous, ‘since a civilian committing a hostile act, even an isolated one, would be taking a direct part in hostilities’.Footnote 214 According to Fleck, the GDR representative, the proposed amendment was intended to make draft Article 46 ‘clear and applicable for the serving soldier’.Footnote 215 The proposal was, however, rejected by other states,Footnote 216 and ultimately the ICRC draft was preferred.
The view that Article 51(3) encompasses the carrying out of ‘hostile acts’ nevertheless made its way into the final ICRC commentary on the provision.Footnote 217 The commentary unhelpfully conflated the notion of ‘direct participation in hostilities’ with engaging in ‘hostile acts’.Footnote 218 Thus, while Protocol I on its face narrowed the exception to the norm of non-combatant immunity, the discussions surrounding Article 51(3) demonstrate that there remained considerable latitude in the interpretation of ‘direct participation in hostilities’.
On the whole, the adoption of Protocol I substantially strengthened the protection afforded to civilians in comparison with previous manifestations of the norm of civilian immunity. First, Protocol I adopted a broad definition of ‘civilian’ which includes all those who do not qualify as a ‘combatant’.Footnote 219 Second, it provided that only civilians who take a direct part in hostilities lose their protection from attack; indirect participation in hostilities is not sufficient to cause the forfeiture of civilian immunity.Footnote 220 Arguments that would see civilian populations lose their immunity due to their ‘war sustaining’ activitiesFootnote 221 are therefore beyond the legal pale of Article 51(3).Footnote 222 Finally, Protocol I expressly recognized that immunity is lost only for such time as civilians participate directly in hostilities.
In the light of these important differences between Article 51(3) and previous formulations of the norm, one might well have doubts about the view expressed by the UK representative during negotiations that Article 51(3) simply reaffirmed existing rules of international law designed to protect civilians.Footnote 223 Rather, Article 51(3) has involved a fairly substantial shift in the balance between military necessity and humanity, with a weighting in favour of protecting civilians. The precise scope of Article 51(3), however, remains unclear. With these factors in mind, the final part of this paper analyses the relationship between the evolving requirements of military necessity on the one hand, and the increasingly protective framework of international humanitarian lawFootnote 224 (and, in particular, Article 51(3)), on the other.
Challenges facing Article 51(3)
At the centre of international humanitarian law is – and always has been – a tug of war between the competing principles of military necessity and humanity. Military necessity admits of such force as is necessary to subdue the enemy.Footnote 225 In accordance with the dictates of military necessity, the immunity afforded to civilians has, historically, been predicated on their remaining unarmed and harmless. For Grotius and his contemporaries, notions of honour and righteousness required that only the ‘innocent’ – that is, those who were unarmed and not guilty of serious crime – be protected from attack.Footnote 226 Throughout the nineteenth century the requirement of harmlessness prevailed. The influential Lieber Code, for instance, required the sparing of the ‘unarmed citizen … as much as the exigencies of war [would] admit’.Footnote 227 If military necessity required the targeting of harmful civilians, the Lieber Code, with its allowance for the exigencies of war, was sufficiently flexible to accommodate their being put to death, at least as a matter of law enforcement. Subsequent nineteenth-century instruments did nothing to displace this approach or disavow the paramountcy of ‘the exigencies of war’; thus, while civilians were generally protected, their protection remained subject to military demands.Footnote 228 By the end of the nineteenth century, therefore, it was established that in order to enjoy immunity from attack, ‘non-combatants must be harmless. As soon as an individual ceases to be harmless, he ceases to be a non-combatant.’Footnote 229 This approach, which effectively deprived those who were harmful ‘[took] part in the fighting’Footnote 230 from the protection afforded to civilians (thus roughly reflecting Grotius' conception of the lex ferenda), continued until Protocol I.
Protocol I went against the tide of history by expressly conferring civilian status on all those who were not combatants properly so called, regardless of whether or not they were harmless.Footnote 231 The inclusive definition of ‘civilian’ in Article 50 of Protocol I meant that classes of people not fitting the traditional civilian mould were nonetheless entitled to immunity against attack. This definition may be contrasted with the ICRC's 1956 Draft Rules, which, on their face, would have deprived those who took part in the fighting of civilian status altogether.Footnote 232 As a result of Protocol I's undifferentiating conception of civilians, international humanitarian law found itself, in the words of Best, ‘teetering on the edge of a credibility gap’, with the law bestowing on all classes of non-combatants the same protection.Footnote 233
Since the adoption of Protocol I in 1977, ways for civilians to harm the enemy have developed and diversified, such that the conduct of a civilian may be as integral to military operations as that of the soldier.Footnote 234 We are now far removed from the post-Westphalian paradigm of war as a conflict which takes place on a battlefield between the armed forces of states or state-like entities.Footnote 235 As seen in Operation Iraqi Freedom, for instance, military functions are being outsourced to private contractors to an unprecedented degree.Footnote 236 Recent years have also seen the rapid development of weapons technology that enables attacks (for example, against computer networks) to be waged with great precision from remote locations by either civilian or military personnel.Footnote 237 Advanced weapons technology is, however, far from universally available, and the technological capabilities of the parties to several contemporary international conflicts have been significantly disparate.Footnote 238 This military asymmetry appears to have created incentives for the weaker party to resort to more covert or perfidious methods of war which frequently involve civilian participants, such as those seen in the latest conflicts involving international forces in Afghanistan and Iraq.Footnote 239 The support of the civilian population becomes, to paraphrase Mao Zedong, as essential to the combatant as water to the fish.Footnote 240 Increased reliance on civilian populations by the parties to armed conflicts, together with an often pronounced divergence in their ideological and structural make-up and motivations,Footnote 241 blurs conventional understanding about who is and is not a civilian.
The increasing military value of civilians may create doubt as to whether the ‘direct participation’ exception in Article 51(3) strikes an appropriate balance between the competing considerations of military necessity and humanity. For example, some argue that civilians in ‘societies with malevolent propensities’ should, as a matter of military necessity, be exposed to attack.Footnote 242 That such arguments continue to arise despite rules prohibiting terror bombingFootnote 243 and attacks against civilians for indirect participation in hostilitiesFootnote 244 is testimony to the tension (around which international humanitarian law revolves) between fluctuating perceptions of military necessity and considerations of humanity.
Notions of military necessity suggest that civilians whose actions are harmful to the enemy should lose their immunity from attack. In contrast with previous manifestations of the exception to civilian immunity, however, Article 51(3) does not permit the targeting of all civilians whose attack is necessary from a military perspective. Rather, only those who are participating directly in hostilities may be subject to attack. People whose conduct is indirectly harmful to the enemy are, under Article 51(3), protected from attack. In this respect, Article 51(3) gives rise to complex questions about how to distinguish between direct and indirect participation in hostilities. To this end, it is hoped that the guidelines being published in this issue of the Review, a culmination of the international expert process which took place under the auspices of the ICRC and the TMC Asser Institute, will provide much needed interpretative guidance. With the increased role of civilians in armed conflict since 1977, the importance of the ‘direct participation’ exception has grown. It now requires clarification and reinterpretation if it is to remain workable in the context of modern hostilities. Whether or not this aspect of the law can be interpreted and applied to accommodate the competing dictates of humanitarian protection and military necessity remains to be seen.
