5.1 Introduction
5.1.1 Fundamentals
The law of armed conflict, which is also known as international humanitarian law or laws of war, is a set of rules specifically designed (lex specialis) for the conduct of hostilities.1 In the current text, the term ‘law of armed conflict’ or its abbreviation loac will be used. As explained in Section 4.1, the application of loac is premised on the factual situation of belligerent hostilities and the necessities that entails.2 Although loac was developed as a means to reduce suffering, it was not intended to prohibit war.3 Instead, the rules developed are the result of “a compromise based on a balance between military necessity, on the one hand, and the requirements of humanity, on the other”.4 As Schmitt explains, “no state likely to find itself on the battlefield would accept norms that place its military success, or its survival, at serious risk”.5
There is some debate whether loac provides permissions or authorisations, prohibiting all that is not authorised, or whether it imposes prohibitions, leaving anything not prohibited as authorised.6 While it may historically have been the case that States considered themselves to be unconstrained when waging war, with the exception of certain generally accepted prohibitions,7 modern loac is no longer merely prohibitive, neither with regards to its nature nor the contents of the rules. The view that will be applied here is that loac provides the legal basis for the use of force during armed conflict, and that its rules are formulated both as permissions (such as military necessity) and restrictions (such as proportionality) that must be complied with in order for the acts of war to be lawful.8 loac rules commonly include terms of a discretionary nature, such as proportionate, necessary and feasible. This enables military forces, and particularly Commanders, to assess how to strike the best balance between the driving force of military necessity and the limitations required by humanitarian considerations, within the parameters set out by the applicable rules.9 As a result, loac is constant in what applies, but how it will apply depends on the relevant facts and how they are perceived at the time.
Most of the rules are centred on the concept of ‘attack’, which is defined to include the use of force in offence or defence.10 There are, for example, rules on who or what may be considered a lawful target; limitations on the means and methods of attack; and regulation of the permissible consequences of attack. There are also more general rules, as in the case of the treatment of persons under control of a party to the conflict and occupation. Because the focus of the book is on the use of force against persons during an armed conflict and how this is regulated by roe and applicable law, the loac rules most relevant here are those on the conduct of hostilities. More specifically, the focus will be on the rules determining who is a lawful target, what it entails to be a lawful target, and how attacks are to be carried out, including identification of lawful targets and precautions in attack. As explained in Section 3.5, the most complicated nato use of force roe, namely the nato hostile act and hostile intent roe, authorise attack on individuals or groups on the basis of their conduct, and the loac concept of ‘direct participation in hostilities’ (dph) will therefore be examined in detail. To a large extent, the rules governing the conduct of hostilities considered here may be deemed customary international law equally applicable to iacs and niacs, and in those cases, a distinction will not be made between the two types of conflict.11
loac is commonly summed up in a varying number of principles or rules, such as distinction, military necessity, and proportionality, and unnecessary suffering or humanity.12 It is, however, important to note that these are the driving forces behind the developed rules rather than additional rules, and may therefore not be used as a justification for violating loac rules.13 This is especially the case with respect to military necessity.14 Although it may be debated whether they strictly speaking are principles, in the sense that they have an independent legal impact,15 this discussion will not be dealt with here, and the term principle will be used.
The principle of distinction16 demands that the parties to the conflict “at all times distinguish between the civilian population and combatants and between civilian objects and military objectives and accordingly shall direct their operations only against military objectives”.17 This principle is easy to describe though challenging to apply in practice,18 a challenge that lies at the heart of this book.
Military necessity is the hardest of the three principles to define. Although it has been the subject of discussions too extensive to cover here,19 the UK loac Manual captures the essence of the principle: “Military necessity permits a state engaged in an armed conflict to use only that degree and kind of force, not otherwise prohibited by the law of armed conflict, that is required in order to achieve the legitimate purpose of the conflict, namely the complete or partial submission of the enemy at the earliest possible moment with the minimum expenditure of life and resources”.20 Military necessity must, as mentioned, be balanced with humanity, namely the prohibition on inflicting “suffering, injury, or destruction not actually necessary for the accomplishment of legitimate military purposes”.21
Finally, the proportionality rule requires those who plan or decide upon an attack to “refrain from deciding to launch any attack which may be expected to cause incidental loss of civilian life, injury to civilians, damage to civilian objects, or a combination thereof, which would be excessive in relation to the concrete and direct military advantage anticipated”.22 Attacks expected to cause excessive incidental loss of civilian life, injury to civilians, damage to civilian objects are classified as “indiscriminate” and are hence prohibited.23
With the exception of certain prohibitions such as perfidy, most loac conduct of hostilities rules focus on the effects rather than on the means or methods of attack. Thus, they remain relevant even for current conflicts. However, certain aspects of loac have been challenged. One example that is particularly relevant to the current topic is the intended dichotomy between combatants and civilians. These are considered mutually exclusive categories, with separate rights and duties. However, as will be discussed in Section 5.2, current conflicts often involve a non-state party, which, for the sake of targeting may be viewed as combatants, but in other respects qualify as civilians.24 The result is an unclear legal framework. Because of this unclarity, proposals have been made for a third category on members of organised armed groups.
5.1.2 Lawful Acts of War
loac is premised on State control over the use of force, and the principle that military forces that act on behalf of the State in accordance with loac will not be acting unlawfully. As Baxter explained, “the act of the soldier who conforms to the law of war and does not engage in private acts of warfare is an act of state depriving the enemy state of jurisdiction”.25 This is known as the ‘combatant’s privilege’ or ‘combatant immunity’, and is premised on the idea of ‘lawful acts of war’. The application of loac makes behaviour that would otherwise be unlawful permissible, provided the person acting is bestowed the privilege to take part in the hostilities, as is the case for combatants.26 The concept of ‘lawful acts of war’ is central to this book. It illustrates how loac permits the use of force rather than merely regulating it; it means that the use of force against persons will not be arbitrary, affecting the scope of application of personal self-defence by providing another and more robust legal basis for the use of force.27
There is no express authority for this concept in loac, but it is implicit both in the rules providing who may lawfully participate in hostilities28 and the requirement that a detaining power only prosecutes prisoners of war for acts they would also prosecute their own forces for, thereby excluding lawful acts of war.29 It has also been argued that it is part of the customary international law applicable to armed conflicts,30 and is reflected in the Rome Statute to the icc Article 31(3) when read together with Article 21(b).31 The obligation on the opposing State not to prosecute prisoners of war and other opposing forces for acts committed in accordance with loac will naturally only apply in an iac, where the opposing side is a State. States, however, are also likely to grant immunity to their own forces for acts during a niac that do not violate the law of niacs.32
There are two ways in which the concepts of ‘combatant privilege’ and ‘lawful acts of war’ may be approached in national legislation, whether this is accomplished for instance by changing existing law or interpreting existing law in a new way. First, the combatant privilege may be viewed as an authorisation, forming part of the authority provided to military forces to use force on behalf of the State. Where such an authority is provided, soldiers will not need to rely on personal self-defence to justify the use of force, since their use of force is lawful. This may be complemented by a provision in the penal code providing that the use of force in accordance with international law will not violate national criminal law, as the examples of national approaches given below illustrate.
Alternatively, lawful acts of war may be an exculpatory ground in criminal law that justifies the use of force, either on its own or as part of a ‘public authority’ defence. By requiring a defence in criminal law, it is implied that the use of force in accordance with loac is not itself an exception to the national criminal law. This distinction is particularly important for the relationship between the two main legal authorities for military forces to use force during an ongoing armed conflict, namely loac and personal self-defence. If the use of force in accordance with loac is itself lawful, self-defence will not be relevant in the same circumstances because there is no alleged crime to defend against. By contrast, if the ‘lawful acts of war’ concept is a criminal defence, this implies that there is a need for a criminal defence, and other criminal law defences such as self-defence may also be relevant.33
As explained by Radin and Schmitt, “[u]nlike human rights law and practice, which view any death as a possible breach of the right to life that must be investigated, ihl bestows immunity on combatants for deaths that are in accordance with its strictures”.34 ‘Lawful acts of war’ as a concept is therefore most likely to be considered and applied in the context of a criminal prosecution. First, it may function as a basis for not charging the suspect with committing a crime because the elements of the crime are negated by way of being lawful acts of war. For instance, the killing of a lawful target will in most cases not be unjustified, which is commonly a criterion for unlawful killing. In this sense it will negate the need for criminal defences such as self-defence since there is no unlawful force to justify.
Second, in the case of a trial, it will function as a defence in that it will be argued that the elements of the crime are not met, although as will be explained, the concept of ‘lawful acts of war’ is strictly speaking an authority to use force rather than a criminal defence like self-defence and necessity. As will be shown below, this practical application of ‘lawful acts of war’ when a trial is initiated is reflected in the U.S. approach. It is mainly set out in the U.S. military’s Manual for Courts-Martial, and is therefore formulated in the form of a defence. In such situations, especially where it is unclear whether the use of force had fully complied with loac, self-defence may be relevant as a secondary defence. However, as will be explained below in Section 8.2, the scope of application of the criminal law concept of self-defence during an armed conflict is relatively limited, and if it is reasonable to kill someone in self-defence, it is likely to have been reasonable to believe that the use of force was lawful under loac.
Most States have provisions that to some degree reflect the concept of ‘lawful acts of war’ in their domestic legislation.35 A comprehensive comparative study of national approaches is beyond the scope of this book, and will only function to illustrate that different countries have taken different approaches.36 Instead, a selection of national approaches will be included to illustrate the variety in solutions.
In Norway, there is no general law on the use of force by the military.37 The traditional approach appears to have been that a lawful act of war is something the State is authorised to carry out without express foundation in law.38 This was confirmed by the Norwegian Supreme Court in 1946.39 The issue has not been considered by Norwegian courts since the trials arising from the Second World War. ‘Lawful acts of war’ is, in other words, a customary law notion. Continued use of the Norwegian armed forces in operations involving armed conflict indicates that this customary rule still applies. There is a tendency in academic literature to deal with ‘lawful acts of war’ together with criminal law justifications, as another reason that the use of force may be justified and hence lawful.40 However, a suggestion to codify ‘lawful acts of war’ as a criminal law defence in the recent update of the penal code was unsuccessful.41 The relationship between ‘lawful acts of war’ and Norwegian law, in particular the penal code, was considered sufficiently established through a general article stating that “[t]he criminal legislation applies subject to the limitations that follow from agreements with foreign states or otherwise by international law”.42 The provisions of the Penal Code must be interpreted restrictively in light of Norwegian international law obligations, with the result that lawful acts of war will not violate its provisions.43
Provisions similar to that found in the Norwegian Penal Code, providing that acts carried out in accordance with international law will not amount to a violation of national criminal law, are also found in the laws of other European countries. According to the Danish Penal Code §12, the application of the penal code is limited by internationally recognised exceptions.44 Article 25 of the German Basic Law states that “The general rules of international law shall be an integral part of federal law. They shall take precedence over the laws and directly create rights and duties for the inhabitants of the federal territory”.45 In the Netherlands, the Dutch Military Criminal Code provides that “an otherwise unlawful act committed during an armed conflict that is in conformity with ihl and within the limits of a soldier’s competence cannot be punished”.46 In 2005, France also introduced a similar provision, clarifying the authority for military forces to use force in accordance with international law as part of military operations conducted beyond French territory and territorial waters.47
The same approach is taken in the U.S. Model Penal Code48 with regards to the use of lethal force. In Section 3.03(2)(b), it is proposed that killing that “occurs in the lawful conduct of war” should be treated as an exception from the criminalisation of taking another’s life.49 Other conduct “is justifiable when it is required or authorized by (…) the law governing the armed services or the lawful conduct of war”.50 However, as mentioned above, in the U.S. military’s Manual for Courts-Martial (mcm), all lawful acts of war are treated as part of defences. Under rule 916(c), ‘Justification’, “A death, injury, or other act caused or done in the proper performance of a legal duty is justified and not unlawful”.51 The comments provided (entitled ‘Discussion’) emphasises that “killing an enemy combatant in battle is justified”.52 Although this appears to suggest that the approach in the U.S. military legal system differs from those already discussed, it must be remembered that the mcm only applies when a case is prosecuted and tried. The argument that an act was lawful by way of being a lawful act of war will therefore function as a defence. When the use of force is considered reasonable, the act is considered lawful and an exception to criminal law prohibitions; there will be no prosecution. The application of ‘lawful acts of war’ in the U.S. military law is in other words similar to the approach taken in many European systems. This is confirmed by the argument that the application of ‘lawful acts of war’ excludes self-defence, so that self-defence would only apply to “other kinds of killings, where the decedent is not a lawful combatant but, rather, a civilian”.53
The position in British law is not laid down anywhere and is therefore more difficult to discern. Documents from the Iraq Inquiry reveal that the UK Ministry of Defence considers the matter to be regulated by the traditional common law definition of murder. The definition usually referred to is that set out by Sir Edward Coke in 1797: “murder is when a man of sound memory, and of the age of discretion, unlawfully killeth within any country of the realm any reasonable creature in rerum natura under the King’s peace, with malice aforethought, either expressed by the party or implied by law, so as the party wounded, or hurt, etc. die of the wound or hurt, etc. within a year and a day after the same”.54 The reference to “any reasonable creature in rerum natura under the King’s peace” is interpreted to exclude the killing of an enemy during an armed conflict, thereby providing the necessary exception for lawful acts of war.55 The use of force will of course have to comply with loac, with the result that the killing of persons who are no longer participating in the hostilities, such as prisoners, is not permitted.56 This common law rule is considered to include acceptance for the loac rules on ‘combatant immunity’ even for the opposing forces.57
The national approaches taken to give effect to combatant immunity is likely influenced by the fact that criminal liability for actions of military forces participating in an armed conflict will only be considered if there are indications that the forces have acted unlawfully.58 In fact, allegations of suspected breaches are only required to be investigated if they are credible.59 The actions of military forces are continuously reviewed by their Commanders and the command group in order to evaluate mission success. If someone is suspected of committing or having committed an unlawful act, Commanders have a duty to prevent them and, where applicable, initiate the appropriate disciplinary or penal actions.60 Actions that are considered reasonable and lawful will usually not be subject to further scrutiny,61 and the legal authority for carrying out lawful acts of war is unlikely to be questioned.
Members of armed groups that are not part of the State’s military forces or do not meet the requirements set out in Article 4A of the Fourth Geneva Convention will not be entitled to combatant immunity. They are not considered to have a right to participate in hostilities,62 and their acts are therefore not considered lawful acts of war. Although this rule is only expressly set out in AP I, it was recognised long before the 1970s.63 In a commentary on the Norwegian post-Second World War trials, Andenæs commented that members of the Norwegian resistance movement were not combatants, and if captured, “would not be entitled to treatment as prisoners of war, but had to be prepared for being held responsible and maybe shot”.64 He further emphasises that this did not make the organisation unlawful under international law, they were merely not afforded combatant privileges.65
In addition to risking potential prosecution, the fact that non-state actors in a niac are not lawful combatant can also affect the potential application of personal self-defence. As explained in Section 8.2.3, because armed groups in a niac are not lawful combatants, their attack on nato forces is not afforded prosecutorial immunity under international law; it therefore is likely to be considered unlawful under domestic law.66 This means that the first requirement of self-defence is met, namely that of an unlawful or unjustified attack; however, the scope of application of self-defence remains limited and hence unsuitable for many combat situations.
5.1.3 Attack: the Use of Force in Offence or Defence
Many of the loac rules are centred on the notion of attack. Examples include who or what may be the object of attack, who or what enjoys special protection from attack, and precautions in attack. The attack rules are also the most relevant to the discussions in this book. During military operations, force is used in connection with dynamic or deliberate targeting, or in response to threats, hostile acts or attack by opposing forces (combat engagement). The loac rules examined in this chapter will be applicable to all of these situations.
According to Article 49(1) of AP I, ‘attacks’ means “acts of violence against the adversary, whether in offence or in defence”.67 ‘Acts of violence’ includes all acts with violent consequences, not just those of a kinetic nature.68 The emphasis on ‘offence or defence’ entails that attack should be interpreted to entail all use of force, including where the opposing side attacks first and the use for force by nato forces is in response to that first attack.69 Although this statement may be contrary to the etymological meaning of the word attack and hence a source for confusion,70 it is an important specification of the application of loac. It is particularly important in the discussion of the relationship between loac and self-defence, and therefore also between roe and self-defence. When loac applies to the use of defensive force as well as offensive force, there is limited need for applying the exceptional legal regime of self-defence.71
It should be noted that loac refers to both attack and hostilities. ‘Hostilities’ is a broader concept than ‘attack’, and includes military acts not involving violence, such as psychological operations (psyops).72 Furthermore, according to Article 49(3), the definition of attack in Article 49(1) and the subsequent rules in AP I on the general protection against effects of hostilities formally only apply to warfare which affects civilians or objects on land.73 However, the rules are generally applied as customary law in air and maritime operations as well, especially the rules on proportionality and precautions.74
5.2 Who Is a Lawful Target
5.2.1 Introduction
The use of force against persons by military forces participating in an armed conflict must be directed at persons who are lawful targets according to loac.75 The only exception to this rule is the use of force in self-defence, as defined and regulated by the respective troop contributing nations. The two main categories of persons in loac are combatants and civilians,76 and the principle of distinction between these categories is “the undisputed cornerstone” of loac.77 While combatants may be lawfully attacked, civilians must not be the object of an attack and must be protected from the effects of the hostilities. The dichotomy is, however, a simplification of the categories of persons in loac, especially in the context of targeting.78 Other categories of persons who are not civilians may also not be the object of attack, namely medical and religious personnel,79 civil defence personnel of the armed forces,80 and persons hors de combat, such as prisoners of war and wounded soldiers.81
Civilians are defined negatively as all those who are not or are no longer combatants.82 They are entitled to “general protection against dangers arising from military operations”83 and shall not “be the object of attack”.84 The protection will be lost by the civilians if and “for such time as they take a direct part in hostilities”.85 By contrast, combatants are targetable at all times, unless they have become hors de combat. Combatants have traditionally been defined as members of the armed forces of a party to the conflict, both regular and irregular such as militias and volunteer corps.86 In addition to these two traditional categories, members of (non-State) organised armed groups are increasingly considered a separate category; they do not meet the requirements of lawful combatants, but do not resemble civilians either.87 They are not entitled to participate in hostilities, but to the extent that they do, they are not entitled to civilian protection from being the object of an attack or being exposed to excessive harm, either. The challenge is that loac generally deals with the two traditional categories, and the rules applicable to members of organised armed groups are less clear. The different categories will be considered in further detail below.
The distinction between combatants and civilians directly participating in hostilities is important for understanding the operational concepts of conduct-based and status-based targeting and the corresponding nato roe.88 Combatants are lawful targets due to their status as combatants, regardless of current activity, thereby enabling status-based targeting. Civilians who lose their protection due to their direct participation in hostilities are lawful targets based on their conduct, but continue to have status as civilians.89 Non-state forces can be seen either as members of an organised armed group or as civilians taking a direct part in hostilities on a continued or regular basis. While the former would enable status-based targeting, the latter would remain a form of conduct-based targeting and therefore require information about the continued participation in order for an attack to be lawful.90
Status-based targeting has the benefit of removing the requirement of catching the person ‘red handed’.91 However, it is not clear what the criteria for ‘membership in an organised armed group’ should be or how it should be proved, considering that formal membership is unlikely. By contrast, conduct-based targeting requires information about the current activities of the person or persons to be attacked, which may also be challenging to attain. These issues will be further discussed in Section 5.2.3.4, on the possibility for continuous loss of protection on the basis of dph, and Chapters 11 and 12 when examining the relationship between nato attack roe and loac. Finally, although the question of conduct-based versus status-based targeting is most relevant to the determination of when an attack may be carried out and how persons should be dealt with if captured, it can also affect the discussion of the relationship between loac and self-defence. Because conduct-based targeting is more akin to self-defence, there is a risk of confusing the applicable legal frameworks. Some have even argued that “loac deals with status based targeting and not conduct based targeting”,92 presumably implying that conduct-based use of force amounts to self-defence.93 This concern is examined more fully in the context of operational concepts of self-defence.94
The question of lawful targets for a directed attack must be distinguished from the legality of effects of an attack on persons who were not the intended target. This distinction between direct attack and so-called collateral damage will be further discussed in Section 5.3 below, while the issue of mistake or doubt will be examined in Section 5.4.
5.2.2 Combatants
The 1949 Geneva Conventions do not set out clear criteria for being a combatant.95 Instead, definitions are provided in Article 4(A) of GC iii of who is entitled to prisoner of war status.96 Because both combatant and non-combatant members of the armed forces are entitled to this status, this provision does not provide a definition of who is a lawful target.97 A clearer definition is provided in AP I Article 43(2): Combatants are those members of the armed forces of a party to the conflict who are not medical and religious personnel. The armed forces of a Party to the conflict is defined as “all organized armed forces, groups and units which are under a command responsible to that Party for the conduct of its subordinates, even if that Party is represented by a government or an authority not recognised by an adverse Party”.98 The forces must further be subject to an internal disciplinary system enforcing compliance with loac.99 Whereas the main focus here is on nato forces and therefore on combatant members of State armed forces, other groups of persons may also be considered members of the armed forces with the corresponding rights when they meet the criteria set out in Article 43(1). These include members of other militias and members of other volunteer corps, including organised resistance movements, belonging to a Party to the conflict and operating in or outside their own territory, and participants in a levée en masse.100
Combatants must also, as a general rule, distinguish themselves from the civilian population by wearing a uniform or other distinctive signs.101 If these requirements are not met, the person will remain a combatant and a lawful target, but will not be entitled to pow status if captured while not complying with the principle of distinction.102
Combatants are lawful targets at any time, regardless of their current activity.103 They cease to be lawful targets if they become hors de combat, that is, a combatant who “is in the power of an adverse Party”, “clearly expresses an intention to surrender”, or “has been rendered unconscious or is otherwise incapacitated by wounds or sickness, and therefore is incapable of defending himself”.104 In addition, in order to remain hors de combat, the combatant must abstain from any hostile act and not attempt to escape.105
Unlike, for instance, civilians taking a direct part in hostilities, combatants have the right to participate in hostilities,106 with the result that their acts are considered lawful acts of war, provided they comply with loac.107 They are also entitled to the favourable treatment set out in GC iii concerning prisoners of war.108 This right is not an individual right but rather afforded because the individual is a member of the armed forces.109
The term ‘combatant’ was traditionally reserved for military forces representing a State in an iac. Therefore, the concepts of ‘combatant immunity’ and ‘lawful acts of war’ are only regulated in treaties dealing with such conflicts.110 As mentioned above in Section 5.1.2, it is common to refer to the State forces in a niac as combatants as well, and States will grant their own forces the right of combatant immunity for acts carried out on behalf of the State in all conflicts.111 Unless the application of the combatant privilege in niacs is secured in domestic legislation, this will of course be contingent on the State remaining in power at the end of the conflict. The other right arising from the status of lawful combatant, namely to be treated as a prisoner of war if captured by the adversary, will not be applicable during a niac because the adversary is not a State and is therefore not obliged to afford this status and treatment.
Even though the term ‘combatant’ historically and legally has been used to describe armed forces fighting on behalf of the State, there is a tendency to refer to all those who participate in hostilities as ‘combatants’.112 As a result, it has become common to refer to those who are entitled to participate in hostilities are as ‘privileged combatants’ or ‘lawful combatants’ in order to distinguish them from persons who participate in hostilities without having the right to do so.113 The latter group is sometimes referred to as ‘unprivileged combatants’ or ‘unlawful combatants’,114 although the term ‘unlawful’ is somewhat misleading because there is no provision in loac prohibiting their participation.115 Their participation is, however, likely to be prohibited under applicable domestic legislation because it commonly includes the use of force not authorised by the State and thus in violation of national criminal law.116 Furthermore, as explained in the UK loac manual, the use of the terms is also misleading in the sense that a person who does not have the right to participate in hostilities does not meet the AP I definition of combatant.117 For AP I States, ‘unlawful combatant’ would therefore be an oxymoron.118
The application of the term ‘combatant’ to categories of persons other than State armed forces is likely the result of imprecise regulation. The description and status of non-State forces in a niac is particularly controversial and unclear.119 Although the term ‘combatant’ should be reserved for iacs, its application to all forms of armed conflicts has resulted in a corresponding expansion of the term ‘unlawful combatants’ to niacs as well.120 States have not clearly described or defined this category in international treaties. For instance, while Common Article 3 refers to “members of armed forces”,121 AP II distinguishes between armed forces of a High Contracting Party and “dissident armed forces or other organized armed groups”.122 In the Manual on the Law in Non-International Armed Conflict, the term ‘fighters’ was proposed as an alternative to combatants in order to avoid confusion.123 However, as pointed out in the icrc Customary International Law Study, “this term would be translated as ‘combatant’ in a number of languages and is therefore not wholly satisfactory either”.124 Although there are debates about the appropriate term for this category, State practice indicates that States generally consider such persons to be lawful targets and to not have the right to participate in hostilities.125
The existence of a third category in addition to combatants and civilians, to account for members of non-State organised armed groups, is further examined in Section 5.2.3.4.
5.2.3 Civilians Taking Direct Participation in Hostilities (dph)
5.2.3.1 Background
Civilians are generally protected both from direct attacks and from attacks that are expected to cause excessive civilian losses compared to the military advantage anticipated to be achieved from an attack. These protections are lost if and when a civilian directly participates in hostilities. When civilians no longer directly participate, they regain civilian protection, although they may be prosecuted under domestic law for any unlawful acts carried out while taking a direct part in hostilities. As pointed out above, civilians do not have the right to participate in hostilities, as this right is reserved for combatants. This does not, however, mean that international law prohibits such participation. Instead, their acts are likely prohibited by domestic legislation.126
There is considerable debate and disagreement over the requirements for ‘direct participation in hostilities’ (dph) and how the concept should be applied. Another complex aspect is the time requirement (‘for such time’), that is, how long the protection is lost. May a person be a farmer by day and fighter by night, or will he lose his protection for an extended period of time? And when does the participation end?127 While the rule is only expressly set out in the Additional Protocols of 1977,128 it is considered part of customary law applicable in all armed conflicts.129 The following examination of what is entailed in directly participating in hostilities will therefore not distinguish between iacs and niacs.130
In 2009, after six years of discussions between the icrc and international experts, the icrc issued an ‘Interpretive Guidance on the Notion of Direct Participation in Hostilities’.131 The group failed to reach consensus, and some of the international experts published their counterarguments to the icrc approach in a separate publication.132 Although few States have published similar declarations on the icrc’s proposal, lawyers in the defence sector commonly voice their concerns and criticisms when the topic is discussed in international fora. The main concern is that the proposed interpretation is too difficult to apply in practice.133 Even though aspects of the Interpretive Guidance are controversial, it serves as a useful starting point for discussions, and will be used in that manner in this book.134
The concept of dph has played an increasingly important role in military operations, both in relation to who may be the object of an attack and who must be taken into account when conducting a proportionality analysis or assessing whether precautions in attack are required. When the opposing side fails to comply with the principle of distinction, making it harder to distinguish them from protected civilians, conduct-based targeting becomes essential for the operation.135 For instance, during Operation Enduring Freedom in Afghanistan, U.S. and its coalition partners did not declare any forces hostile, which would have enabled status-based targeting. Instead, they introduced the concept of ‘likely and identifiable threat’ (lit) as a form of conduct-based targeting.136 As will be further explained below in Chapter 11, dph is probably the most relevant legal basis for conduct-based targeting on the basis of the nato hostile act and hostile intent roe.137
Furthermore, the demands to reduce civilian casualties beyond the requirements of loac have made it necessary to identify the status of persons in the vicinity of a known lawful target to a greater extent. If all persons expected to be harmed by an attack can all be classified as either civilians taking a direct part in hostilities or members of an organised armed group, the attack will not be expected to cause civilian casualties and may therefore be carried out. This has resulted in an increased focus on what amounts to dph and on how to apply the criteria. Combined with the failure of the opposing forces to wear uniforms or other distinctive signs, this civcas policy has pushed Commanders and their legal advisers “to place a priority on identifying fighters, and on developing rules for the use of force that allow targeting of these fighters, while also protecting civilians”.138 Having identified and located a person who is a lawful target, the status of other persons in the vicinity must be assessed on the basis of information available at the time, such as their interaction with the target, the location of the target, the presence of weapons, and the nature of these weapons. It will, for example, be easier to make the determination if the intended target is in a building or vehicle known for being used by the opposing forces than if it is an unknown location or location known for civilian use. Furthermore, while civilians may carry arms for self-defence, it is less likely they carry military type weapons such as rpgs or sniper rifles.
The scope and application of the dph concept has been subject to extensive research and debate in the aftermath of the release of the icrc Interpretive Guidance. It is not the purpose of the following presentation to provide a complete overview of those discussions. Instead, the focus is on providing enough insight into the topic to enable a discussion of the relationship between nato roe and the use of force during armed conflict, in particular in response to persons demonstrating a hostile intent (not constituting an imminent attack) or who commit or directly contribute to a hostile act (not constituting an actual attack). However, as a result of the controversies surrounding the application of dph, both with regard to individual acts and continuous loss of protection, the examination of dph will be relatively detailed.
5.2.3.2 Direct Participation in Hostilities
In order to take a direct part in hostilities, or active part as it is sometimes described,139 a person must be found to be involved in hostilities, and that involvement must amount to direct participation.140 According to Melzer, “‘hostilities’ refers to the collective resort by parties to an armed conflict to means and methods of warfare”, while “‘participation’ refers to individual involvement of a person in these hostilities”.141 In addition, the participation must be direct as opposed to indirect. Melzer explains the distinction in the following way: “Direct participation refers to hostile acts carried out as part of the conduct of hostilities between parties to an armed conflict and leads to the loss of protection against direct attack. Indirect participation, by contrast, may contribute to the general war effort, but does not directly harm the enemy and, therefore, does not entail the loss of protection against direct attack”.142
According to the icrc Interpretive Guidance, three cumulative requirements must be met in order for a civilian’s behaviour to qualify as dph: there must be a certain threshold of harm expected from the act; there must be a direct causal relation between the act and that harm; and the act must be part of the ongoing hostilities.143
In order to meet the first criterion, “a specific act must be likely to adversely affect the military operations or military capacity of a party to an armed conflict or, alternatively, to inflict death, injury, or destruction on persons or objects protected against direct attack”,144 such as civilians and civilian property. The harm need not have materialised but there must be an objective likelihood that it occurs.145 When the harmful act is directed at a military operation or capacity, there is no requirement of ‘quantitative gravity’, but it will not be sufficient that the civilian refuses to act in support of a part to the conflict.146 In fact, “any consequence adversely affecting the military operations or military capacity” would suffice.147 In addition to killing and physically injuring or damaging persons or property, the Interpretive Guidance includes as examples “armed or unarmed activities restricting or disturbing deployments, logistics and communications”; “denying the adversary the military use of certain objects, equipment and territory”; “electronic interference with military computer networks”; and “transmitting tactical targeting information for an attack”.148 If the act is not adversely affecting the military capacity or operations of a party to the conflict, the act must, in order to meet the threshold requirement, cause “at least death, injury or destruction” to protected persons or objects.149 This excludes inconvenience and even arrest and deportation.150
By the second criterion, the threshold requirement, “there must be a direct causal link between a specific act and the harm likely to result either from that act, or from a coordinated military operation of which that act constitutes an integral part”.151 Because the treaty law refers to ‘direct’ participation, ‘indirect’ participation must be excluded. Support to the general war effort or war-sustaining activities such as building roads, ports, airports and bridges, or economic activities such as oil production will be indirect rather than direct.152 According to the Interpretive Guidance, direct participation must be interpreted to require “one causal step”.153 As a result, activities such as scientific research, transport of weapons and equipment not “carried out as an integral part of a specific military operation designed to directly cause the required threshold of harm”,154 recruitment and training of personnel, or “assembly or storing of an improvised explosive device (ied)”,155 will not qualify. It is contended that “[o]nly where persons are specifically recruited and trained for the execution of a predetermined hostile act can such activities be regarded as an integral part of that act and, therefore, as dph”.156 An act that does not in itself meet the direct causation test but forms part of a collective operation may still amount to direct participation if it “constitutes an integral part of a concrete and coordinated tactical operation that directly causes such harm”.157 Examples of such acts include transmitting tactical intelligence to attacking forces and giving instruction and assistance to troops “for the execution of a specific military operation”.158
Finally, the third criterion proposed by the icrc requires an act to be “specifically designed to directly cause the required threshold of harm in support of a party to the conflict and to the detriment of another”.159 This requirement of a belligerent nexus excludes acts that are likely to directly cause harm of a certain threshold but that are unrelated to the conflict and should be dealt with by law enforcement.160 It also means that the use of force in self-defence cannot amount to dph because its purpose clearly is defensive rather than to support a party to the conflict.161 The belligerent nexus requirement does not mean that the opposing forces must identify a subjective intent. Instead, the requirement refers to “the objective purpose of the act”, which is “expressed in the design of the act or operation and does not depend on the mindset [sic] of every participating individual”.162 It may nevertheless be difficult to determine whether this requirement is met. Based on the conduct of the civilian, the circumstances prevailing at the time and place, it must be reasonable to perceive the act as having the necessary belligerent nexus.163
The icrc Interpretive Guidance has been the subject of extensive debate and criticism. According to Yoram Dinstein, for instance, these criteria “are abstractions”. In his view, “[t]he pivotal question is whether the activities engaged in ‘by their nature or purpose, are likely to cause actual harm to the personnel or matériel’ of the opposing side”.164 He argues that the following parameters should be applied instead of the icrc criteria. First, “every instance of use of weapons in combat” will quality as dph, even if the person is acting far from the contact zone, as will be the case with those launching a missile. Second, dph is not limited to the use of force; it “goes beyond the straightforward commission of acts of violence and includes such ancillary acts as gathering military information”.165 Finally, activities that are not part of military operations should not be considered direct participation of hostilities, even if they are “ultimately harmful to the enemy”.166 Whereas the icrc criteria require further interpretation in order to be applied and therefore are difficult to apply in practice, especially for military forces lacking easily accessible legal advisers,167 the parameters set out by Dinstein provide clearer guidance on how to approach the issue of dph.
Perhaps the most significant difference between this approach and that of the icrc is that it does not require a ‘direct causal link’ between the act and the harm. The direct causation criterion is one of the more controversial aspects of the icrc proposal for interpreting dph.168 According to Schmitt, there is broad agreement that those who personally conduct attacks are directly participating in hostilities. However, international experts strongly disagree with the icrc approach to civilians who are not directly involved in the attack or operation causing harm, but nevertheless provide support.169 The proposed requirement of “one causal step” is so strict that Schmitt suggests that it may be “a poorly drafted explanation of the agreed upon need for a clear link between the act and the ensuing harm”.170 Instead, the test should be whether the act “constitute[s] ‘an integral part’ of the operation causing the harm”; integral is not to be equated with necessary.171 Rather than providing a definition for the difficult cases, these ought to be resolved “on a case-by-case basis”.172 A similar approach has been taken by the icty: “[I]t is unnecessary to define exactly the line dividing those taking an active part in hostilities and those who are not so involved. It is sufficient to examine the relevant facts of each victim and to ascertain whether, in each individual’s circumstances, that person was actively involved in hostilities at the relevant time”.173
In cases where the law needs to be flexible and is accordingly formulated in a general manner, attempting to define the rule further by setting out additional criteria has the potential of undermining the necessary and inherent flexibility of that rule. The application of dph will depend on the circumstances at the time, including the pattern of life in the area, the nature of the conflict, and the tactics, techniques and procedures of the relevant party to the conflict. For example, a person carrying a rifle may be considered a likely threat in some places, while in others it is normal for all men to carry arms for self-protection. This is also made clear in the U.S. Law of War Manual: “Whether an act by a civilian constitutes taking a direct part in hostilities is likely to depend highly on the context, such as the weapon systems or methods of warfare employed by the civilian’s side in the conflict. For example, in some contexts, training and logistical support may be viewed as taking a direct part in hostilities, while in other contexts it might not”.174 Rather than creating a more extensive and abstract definition than the single line in AP I Article 51(3) and AP II Article 13(3), the better approach may be to train military forces in the relevant parameters for making the determination and provide mission-specific examples of what is clearly within the definition and what will clearly fall outside it.175
The icrc Interpretive Guidance provides numerous examples, some already mentioned above. Others include: “guarding captured military personnel of the adversary to prevent them being forcibly liberated (as opposed to exercising authority over them)”;176 “clearing mines placed by the adversary”;177 “wiretapping the adversary’s high command”;178 “sniper attacks against civilians and the bombardment or shelling of civilian villages or urban residential areas”;179 lookouts during an ambush;180 and “the delivery by a civilian truck driver of ammunition to an active firing position at the front line”.181 However, some of the examples provided in the icrc Interpretive Guidance are particularly controversial,182 thereby revealing the inherent disagreement between international experts and the icrc. As Boddens Hosang points out, “since those examples show the intensions of the icrc behind the recommendations and how to interpret it, they are more than mere illustrations and therefore deserve careful scrutiny”.183 The most controversial examples concerns human shields and ieds.
The example of human shields is commonly used to explain the distinction between voluntary and involuntary participation. There is general agreement that persons who are forced to participate in hostilities, for instance by being used as human shields, will not lose their civilian protection.184 However, in many cases it will be difficult to determine whether a person is acting out of free will or not. In the particular case of human shields, where there is considerable practice of persons being forced to stand before military targets in an attempt to dissuade the opposing forces from attack, the icrc Interpretive Guidance has attempted a novel solution. It proposes that the determination of whether voluntary human shielding amounts to direct participation depends on the effect it has on the opposing forces, which will in turn depend on the opposing forces’ mode of attack. If the human shields create a physical obstacle for land forces or give physical cover to those fighting, their act will qualify as dph. By contrast, if the attack is carried out from the air, the attack could physically still be carried out, but if the act of shielding does not cause the civilians to lose their protection, it would be expected to cause excessive civilian losses and hence be unlawful. It is, in other words, a legal rather than physical obstruction, and should in the view of the icrc not result in a loss of protection because it is too indirect.185
This proposal has been met with much criticism. Boddens Hosang, for instance, calls it “flawed on several levels”.186 As he correctly points out, when a civilian engages in a physical act that adversely affects the military operations of the opposing party, the fact that the adverse effect is caused by legal or physical effects “has no relevance”.187 As long as the act and the intent of the civilian engaged in the act adversely affecting the opposing forces is the same, the qualification of that act should not depend on how the opposing forces carry out their attack. According to Schmitt, the attempt to limit the circumstances in which any human shields lose protection is driven by a desire to prevent them from becoming the object of an attack. However, as he explains, the main reason others argue that voluntary human shields should lose their protection is not so they can be directly attacked, but because they should not count as civilians in the proportionality assessment.188 There is limited military advantage gained from attacking civilians who choose to act as human shields, but their involvement should not be rewarded by making the military objective they are shielding unlawful to attack on the basis that it would cause excessive harm.189
The other controversial example concerns those involved in the manufacturing and storage of improvised explosive devices, also known as ieds. According to the icrc, “the assembly and storing of an improvised explosive device (ied) in a workshop, or the purchase or smuggling of its components, may be connected with the resulting harm through an uninterrupted causal chain of events, but, unlike the planting and detonation of that device, do not cause that harm directly”.190 Although it is generally recognised that persons working in an munitions factory generally do not lose their protection from direct attack, applying this approach to ied production is inappropriate. As Schmitt explains, “ieds are often assembled and stored in close proximity to the battlefield by members of armed groups. Although the precise location and time at which they will be used may not be known in advance, they will likely be employed soon after their assembly”.191 In such cases their activities would amount to dph.192 The contrary conclusion by the icrc is a result of the “one causal step” criterion and illustrates the weakness of their proposed test.193
The disagreements on some of the icrc examples may be the result of a more systematic disagreement: the icrc appears to focus only on the tactical level when setting out actions that may constitute dph.194 For instance, many of the examples provided relate to the carrying of weapons.195 This limitation on the scope of direct participation appears unwarranted. According to Watkin, “it has long been recognized that insurgent campaigns, like conventional warfare, are fought with strategic as well as tactical goals in mind. As a result, the exercise of command, planning, intelligence, and even logistics functions can involve dph above the tactical level”.196 This focus on tactical actions is also reflected in the assessment of the duration of dph; the direct participation ends for instance when a shooter lays down his weapon.197
5.2.3.3 Beginning and End of Participation
Apart from the ‘direct causal link’ criterion, the ‘for such time’ requirement is potentially the most controversial aspect of the definition of dph set out in the Additional Protocols and Interpretive Guidance.198 The criterion consists of two separate but interlinked considerations: when will the protection be lost, and for how long? This section will look at the beginning and end of specific acts or operations amounting to dph, while the question whether the right to civilian protection may be lost for a longer period of time will be considered in Section 5.2.3.4 below. The icrc position will be presented first.
While it is clear that acts such as attacking will amount to dph, certain preparatory measures to the act or operation involving dph may also entail the loss of civilian protection if it is an integral part of that act or operation.199 The icrc proposes that in order to qualify as dph, the preparatory measures must be “aiming to carry out a specific hostile act” rather than “aiming to establish the general capacity to carry out unspecified hostile acts”.200 It is neither necessary nor sufficient that the act is close to the execution of a specific hostile act in time or space, nor is it sufficient that the act is indispensable for its execution.201 Thus while loading bombs onto a plane for a specific attack will amount to dph even if the attack is carried out the next day and far away, the loading of bombs onto a truck to be transported to a storage unit for future use will not.202 According to the icrc, a distinction must be made between preparatory acts for the execution of a specific hostile act and general preparation. For example, although general training of personnel is considered insufficient to warrant a loss of civilian protection, instruction on how to carry out a specific act will amount to dph.203
Civilian protection will be lost during the deployment to and return from the execution of a specific act or operation if and when this is considered an integral part of the act or operation that amounts to dph. According to the icrc, “[a] deployment amounting to dph begins only once the deploying individual undertakes a physical displacement with a view to carrying out a specific operation. The return from the execution of a specific hostile act ends once the individual in question has physically separated from the operation, for example by laying down, storing or hiding the weapons or other equipment used and resuming activities distinct from that operation”.204
The icrc approach to the timing issue was not supported by many members of the group of experts participating in the development of the Interpretive Guidance. They considered it too restrictive.205 It excludes, for instance, a finding that persons involved in assembly, smuggling or hiding weapons, including acquiring materials and building an ied, could be directly participating in hostilities, a view which clearly departs from State practice over the past years. In view of the experts, “the period of participation should extend as far before and after a hostile action as a causal connection existed”.206 This approach will, to a greater extent, include preparatory acts that are an integral part of harmful act. Boothby goes so far as stating that the icrc approach is simply wrong: “The icrc is wrong to limit the notion to ‘military operations preparatory to an attack’, and is also wrong to limit it to preparation for a particular attack. The distinction is finer than that. It is between the generation of a general capacity to undertake military activity and preparation for combat or hostilities”.207
The experts’ criticism also concerned the question of when protection is regained. Boothby is especially critical of the proposal that physical separation from the operation, such as laying down weapons, is sufficient. He argues that such acts should instead be viewed as “preparatory to the next act of dph”.208 The icrc approach is clearly too restrictive, encouraging misuse of civilian protection by simply laying down weapons every time the opposing forces appear to be able to respond. A person who has attacked the opposing forces should not be able to claim civilian protection as soon as the weapons are hidden and prepared for the next attack.209 However, not all acts physically separating the person from the operation should be assumed to be preparation for continued direct participation; it may be that the act actually constitutes the end of the direct participation. The better approach therefore seems to be to reassess whether the acts amount to preparation for further participation, or, as proposed below, whether the participation is sufficiently regular to warrant continued loss of protection.
The question of when civilians directly participating in hostilities regain protection is an integral aspect of the ‘revolving door’ debate. The issue is often referred to by the example ‘farmer by day, warrior by night’210 and has long been controversial.211 According to the icrc Interpretive Guidance, the revolving door is “an integral part, not a malfunction, of ihl. It prevents attacks on civilians who do not, at the time, represent a military threat”.212 Schmitt points out two errors in this argument. First, persons taking a direct part in hostilities do not lose their protection because they are a threat but instead because they choose to take part in the hostilities.213 The fact that they do not pose a threat between acts of dph is therefore not decisive.214 Second, it does not make sense from a military perspective. When insurgent warfare is based on surprise attacks, it becomes difficult to prepare for them. Instead, the opposing forces are dependent on targeting insurgents when they are preparing their next attack.
The restrictive approach taken by the icrc, however, entails that the insurgents are protected again once the attack is carried out, unless they can be proved to have a continuous combat function in an organised armed group.215 It will, in other words, function as a shield that those directly participating in hostilities can hide behind,216 as the reference by the icrc to “the revolving door of protection” suggests.217 The result is to throw “the military necessity-humanitarian considerations balance wildly askew”.218 Boddens Hosang goes so far as to argue that “the restrictive approach taken by the icrc seems unworkable in practice and eventually runs the risk of eroding, at least at the tactical level, the incentive and understanding for compliance with the principle of distinction as a fundamental component of ihl”.219 The icrc position that the protection comes and goes with each individual act is arguably also at odds with the AP Commentary, which refers to protection being lost “for as long as his participation lasts”.220 According to Boothby, this phrase “implies a period of time and sits rather uncomfortably with the Interpretive Guidance”.221
The ‘revolving door’ question is particularly challenging in the context of operational level or pre-planned operations where the purpose is to target known persons, regardless of their current activity. On the tactical level, however, it is likely to pose fewer challenges. Bagwell and Kovite go as far as to argue that “for tactical level targeting, where typically the person is recognized as a direct participant by his commission of acts that fall well within the “measures preparatory” standard of the Interpretive Guidance”, the proposal to start the period of participation earlier “is largely insignificant”.222 Similarly, unless the persons committing an act of dph are identified, once visual contact is lost, they can no longer be attacked because it will be impossible to distinguish them from civilians.223 However, where the person is identified and known, as is more likely the longer the armed conflict lasts, the revolving door problem will be equally relevant at the tactical level.
A potential alternative to the restrictive icrc approach would be to consider a civilian who repeatedly takes direct part in hostilities “a valid military objective until he or she unambiguously opts out of hostilities through extended non-participation or an affirmative act of withdrawal”.224 The result would be that the stricter ‘for such time’ criterion, applicable to individual acts of direct participation, would only apply to ‘true’ civilians, rather than persons functioning as combatants.225 It may, of course, be difficult at times to discern whether a person has opted out. loac requires that in cases of doubt, a person shall be presumed to be a protected civilian.226 However, it could be argued that if “the civilian has decided to refrain from further participation in hostilities, but an attacker is unaware – and has no reason to be aware – of that fact”, it will be a case of mistake of fact, which loac accepts may occur as a result of the fog of war.227 Reducing the scope for repeatedly stepping in and out of civilian protection, and hence the scope for misusing the civilian protection to gain an operational advantage, also opens the possibility of treating all non-State actors as continuous dph, rather than having to develop generally applicable legal criteria for membership in an organised armed group. The analysis will now turn to considering this issue in further detail.
5.2.3.4 Continuous Loss of Protection: a Third Category for Organised Armed Groups?
When a person directly participates in hostilities on a regular and continuous basis, the question becomes whether the protection will be regained in between those activities. As explained above, the ‘revolving door’ question has been a source of criticism of the ‘for such time’ criterion. In the icrc Interpretive Guidance, it is argued that “it would contradict the logic of the principle of distinction to place irregular armed forces under the more protective legal regime afforded to the civilian population merely because they fail to distinguish themselves from that population, to carry their arms openly, or to conduct their operations in accordance with the laws and customs of war”.228 Requiring States to only attack the opposing forces when they are actually fighting would furthermore ‘fudge’ the difference between those forces and civilians, thereby weakening the protection of civilians.229 The question is how best to approach the issue of continued loss of protection. Although it should be possible to target those who continue to take direct part in hostilities on a regular basis, it is important to find a way to distinguish those who participate continuously and those who do so sporadically. Furthermore, when the loss of protection is based on the expectations of future acts, how can persons considered to have lost their protection for a longer period communicate that they have no intention of doing so and regain their civilian protection?
There are two ways of approaching this problem. Either continuous loss of protection can be limited to those connected to an organised armed group, or the focus can be on the current and expected future acts, with membership in an organised armed groups as a useful, but not decisive, indicator.
In the icrc Interpretive Guidance, it is suggested that persons who have a “continuous combat function” in an organised armed group lose their protection throughout the period of their membership in the group.230 For targeting purposes, they are not civilians, but rather a form of armed forces to which the ‘for such time’ criterion does not apply.231 The concept was initially developed in the context of niacs, but during the project was equally applied to iacs.232 For the purposes of the Guidance, this category includes both dissident armed forces and other organised armed groups.233
To the extent dissident armed forces continue to wear uniforms or otherwise distinguish themselves from the civilian population, they present less of a challenge in relation to targeting. The main challenge to the principle of distinction and for targeting are organised armed groups that “recruit their members primarily from the civilian population but develop a sufficient degree of military organization to conduct hostilities”.234 They often fail to distinguish themselves from the civilian population, and the membership is difficult to determine. Because membership is rarely formalised or visible, the icrc Interpretive Guidance suggests that “the decisive criterion for individual membership in an organized armed group is whether a person assumes a continuous function for the group involving his or her dph”.235 Continuous combat function may be “openly expressed through the carrying of uniforms, distinctive signs, or certain weapons”, or “be identified on the basis of conclusive behaviour, for example where a person has repeatedly directly participated in hostilities in support of an organized armed group in circumstances indicating that such conduct constitutes a continuous function rather than a spontaneous, sporadic, or temporary role assumed for the duration of a particular operation”.236 Individuals whose function does not involve dph are not considered members of the group and hence not targetable on the basis of their relationship with the group.237
In addition to being organised and armed, the Interpretive Guidance requires such groups to “conduct hostilities on behalf of a Party to the conflict”.238 While this is relevant for determining an armed group’s status and associated rights if captured during an iac, it is not clear how it has any implications in the context of targeting, whether in iacs or niacs.239 It does, however, serve as a useful reminder of the requirement that it is not sufficient to be organised and armed: the acts carried out must also meet the criteria of dph, thereby excluding armed gangs of criminals making use of the situation. Furthermore, it excludes groups involved in a parallel armed conflict taking place in the same area.
The icrc proposal has the benefit of making the distinction between those who have lost their protection from attack and those who remain protected as civilians clearer. In particular, it acknowledges the possibility of status-based targeting in niac.240 However, in order to reduce the risk of persons entitled to protection being mistaken for someone who has lost their protection, the icrc has set the threshold of participation very high, making it difficult to apply in practice.241 ‘Continuous combat function’ requires that the criteria for dph be met, including the controversial ‘one causal link’ criterion, as well as that the person is a member of an organised armed group.242 This is problematic for several reasons. First, as a starting point, the icrc test excludes a significant number of supporting personnel who are an integral part of the group’s combat effectiveness.243 The icrc contends that those who are involved in the preparation, execution or command of acts amounting to dph are members of the organised group for the purpose of targeting, while recruiters, trainers, financiers and propagandists are not, even though they may actually be members of the armed group.244 This is likely to be a result of the proposed and controversial high threshold for participation to be considered direct.
While it may make sense to have a high threshold for concluding that an act amounts to dph when there is no connection to an organised armed group, the criteria should be applied differently in cases where such a connection exists. Supporting functions are both integral and essential for an organised armed group, whether State armed forces or non-State groups, to function. The icrc excludes from its definition of members of organised armed groups persons who would have been a lawful target if they were members of the State armed forces. It is difficult to see why they should be able to carry out the same functions for an organised armed group and remain a protected civilian.245 As Watkin explains, “there is a commonality to fighting wars that requires an organization to consist of commanders, planners, intelligence personnel, and fighters to carry out the military action. There is also a requirement to provide logistical support”.246 It does not make sense to claim that in the context of non-State organised armed groups, only those carrying out attacks will lose their protection as civilians. The function of a person within a group should only be relevant to the extent it distinguishes that person from those without a combat-related function, such as medical personnel or members of a political wing of the group. Admittedly it may be harder to ascertain whether persons with support functions such as trainers have lost their protection from attack, but this should not exclude the possibility of coming to that conclusion when their involvement in the hostilities is clear.
Second, although membership in an organised armed group is a useful indication that a person is taking direct part in hostilities on a regular basis, it cannot be an exhaustive requirement for continued loss of protection. It may sometimes not be possible to define the connection between a person of interest and the organised armed group, or it may be that the person is acting out of support for the group without their knowledge. However, there may still be sufficient information to warrant the conclusion that someone is taking direct part in hostilities on a regular basis.
In the last few years, there has been an increasing trend of individuals or small groups carrying out attacks on civilians. In order to avoid detection, they have little or no contact with the group they support. If such attacks are carried out in the context of an armed conflict and over a period of time, they could be considered to qualify as taking a direct part in hostilities such that their civilian protection is lost for a period extending beyond the immediate time before and after attacks. When a person is deemed to take direct part in hostilities over a longer period or on a regular basis, the lack of membership in an organised armed group should not be sufficient to shield that person from direct attack.247
When faced with a group of persons considered to be members of an organised armed group, it will be impossible in many cases to discern their function in that group, as the icrc expects.248 Some therefore argue that if the membership criterion is to remain central, it will only make practical sense if all members of an organised armed group are considered lawful targets, either by way of being armed forces or civilians continuously participating in hostilities.249 The result is that for targeting purposes they will be treated as combatants, although like civilians, they have no right to directly participate in hostilities, and will therefore not benefit from the combatant privileges of prisoner of war status and prosecutorial immunity for lawful acts of war.250
This status-based approach to members of organised armed groups clearly appears less complicated from an operational perspective. However, in many cases organised armed groups do not distinguish themselves from civilians in such a way that they may be identified based on their appearance. As a result, their acts will remain central for determining whether they are lawful targets. Furthermore, certain members of an organised armed group should not be considered lawful targets, especially the medical personnel, provided they refrain from carrying out acts detrimental to the opposing forces. Because medical personnel are unlikely to bear distinguishing marks, their actions will again be important for the determination of their status. Membership in an organised armed group will therefore not be sufficient on its own as a basis for determining whether a person is a lawful target, and should consequently not be the central and conclusive criterion. This does mean that a distinction must be made between those with a fighting function and, inter alia, the cook.251 However, if the distinction is impossible to make, as is commonly the case, the mistake is likely to be considered reasonable.252
The alternative approach to the icrc ‘continuous combat function’ proposal is to apply the criteria for dph and to interpret the temporal scope to mean loss of protection for as long as the person continues to directly participate in hostilities on a regular or recurrent basis.253 As Boothby argues, “the correct approach is to distinguish between, first, isolated and sporadic acts by civilians and, second, repeated or persistent acts of dph. Only the former would involve resumption of protected status after the act of dph, while the latter would involve continuous loss of protected status while such persistent or repeated involvement in hostilities continues”.254 This approach enables the focus to remain on the test set out in loac, namely on dph, rather than having to develop criteria for membership of an organised armed group as well.255
The focus on dph rather than on membership does not, however, solve the question of how long the protection is lost. Information about membership or connection to an armed group will be useful in determining whether the acts amount to dph and whether the participation is regular. The point is that while the lack of such information may make it harder to anticipate future actions, it does not make it impossible to conclude that a person is participating in hostilities on a regular basis. However, the requirements for information about the type of acts that have been carried out, the degree to which they are carried out on a continuous basis, and the quality of the source of the information, should be higher if the membership is unclear. A higher standard reduces the scope of targeting individuals on the basis of regular participation without a clear connection to an organised armed group, thereby ensuring protection of non-participating civilians, while at the same time enabling those known to be participating on a regular basis to be targeted.
The concept of continuous dph without a requirement of membership in an organised armed group is expressly rejected by the icrc Interpretive Guidance. It claims that “[i]n practice, confusing the distinct regimes by which governs the loss of protection for civilians and for members of State armed forces or organized armed groups would provoke insurmountable evidentiary problems.” The Guidance goes on to claim that “[i]n operational reality, it would be impossible to determine with a sufficient degree of reliability whether civilians not currently preparing or executing a hostile act have previously done so on a persistently recurrent basis and whether they have the continued intent to do so again. Basing continuous loss of protection on such speculative criteria would inevitably result in erroneous or arbitrary attacks against civilians”.256
This seems harsh. loac has clear rules on the presumption on civilian status and requires those who plan and carry out attacks to do everything feasible to verify that the target is not a protected civilian or otherwise protected person.257 If it is not possible to sufficiently distinguish between protected civilians and lawful targets, the attack may not be carried out. Furthermore, as Boothby points out, “decisions, including as to whether a targeted individual is a recurrent participator in hostilities, are based on information, not evidence”.258 The icrc is, however, right in emphasising that the criteria for making the determination of continued loss cannot be “speculative”. Proving recurring past participation to a sufficient degree to determine the likelihood of future participation may be difficult, especially where the connection to an organised armed group is unclear. Still, the icrc appears to overestimate the complexity of this determination. According to Watkins, “it is not difficult logically, operationally, or factually to determine future activity from past conduct. This is an intelligence issue involving the same considerations as determining who is performing a ‘continuous combat function’”.259 When explaining how to determine whether someone is having such a function, the icrc also suggests that conclusive previous behaviour may be sufficient.260
Even if future activities are difficult to determine, this does not make it impossible. Furthermore, military operations are not driven by a focus on those who meet the criteria of being a lawful target, but rather on persons who are a threat or potential threat, and who can with reasonable certainty be considered a lawful target. As Prescott explains, “targets are not identified on the basis of their conformance with precise legal guidelines and perfect situational awareness and then recommended for engagement”.261 Instead, “roe are applied to targets as a result of the target validation process and during the actual engagement, however deliberate or expedite these events might be”.262 The decision to use force against someone is based on the intelligence and experience, and whether it is reasonable to consider someone to be a lawful target. In cases where a person’s involvement in activities amounting to dph is very uncertain, it is therefore, first, unlikely that they will draw the attention of nato forces in the first place, and second, it is unlikely that an attack is considered reasonable.
The fact that the law is difficult to apply should not exclude the potential for applying force in those situations where the criteria may reasonably be considered met. As Boothby stresses, “no decisions in war are ever based on certainties. (…) Decision makers thus inevitably have to base decision in part on inference”.263 They are, however, obliged to make such a decision in good faith.264 Furthermore, in order to avoid ‘erroneous or arbitrary attacks against civilians’, States may decide to restrict or retain the authority to determine that a person who is not considered to be a ‘member’ of an organised armed group has lost the protection as civilian beyond the individual acts of dph. By contrast, because targeting on the basis of membership in an organised armed group typically will be easier to apply in practice, the authority for making that determination may be released to a lower level in the chain of command. The relationship between the operational regulation of the use of force and loac is further examined in Chapters 11 and 12.
State practice clearly indicates that opposing forces are considered lawful targets at any time, not just while taking a direct part in hostilities.265 The question is therefore how best to ensure that military forces only direct their attacks at persons who are considered to be lawful targets. Whether the approach to continuous loss of protection is a membership approach or continuous dph or a combination thereof, the most important way to ensure accurate application in an ongoing operation is to provide the military forces with examples of what is considered to meet the tests and not. After all, as emphasised in the U.K. Manual of the Law of Armed Conflict, “[w]hether civilians are taking a direct part in hostilities is a question of fact”.266 And the interpretation of facts will always be context dependent. While it will be important to emphasise the legal criteria, in particular that direct participation must be distinguished from indirect participation, it must be remembered that concrete and relevant examples are more useful to military forces than complex criteria that require interpretation.
A final issue to be considered in the context of continuous loss of civilian protection is how to know whether it has ended. According to the icrc, the disengagement from a continuous combat function “need not be openly declared; it can also be expressed through conclusive behaviour, such as a lasting physical distancing from the group and reintegration into civilian life or the permanent resumption of an exclusively non-combat function”.267 The same will hold true for disengagement from continuous direct participation; the disengagement need not be expressed, but must somehow be affirmative, based on an objective and reasonable assessment of the facts.268
As pointed out above in Section 5.2.3.3, complete certainty is rarely possible during an armed conflict and loac therefore permits mistakes to be made. However, the decision that someone has lost their civilian protection must be based on a good faith assessment of the information reasonably available at the time. If the focus is on the acts of the person, rather than merely his or her ‘membership’ in an organised armed group, the information collected about the person’s previous activity must be regularly updated in order to ensure that the person is still participating.
One way to operationalise this would be to require new information within a set time limit in order for the person to remain on the list of lawful targets.269 Furthermore, since resources are generally limited both in the form of time and means of warfare, military forces will not direct them at persons to whom they do not expect to gain a military advantage by attacking. However, it will not be feasible to keep all such information updated on a daily or even weekly basis. A challenge may therefore arise where a person is attacked on the basis of previous regular participation, but has just decided to disengage from such activities. On the other hand, in such cases the wrongful determination of a person as a lawful target is likely to be an honest and reasonable mistake; it is not feasible to know that the person has recently disengaged. As is emphasised in the Commentary to the Geneva Conventions, “[t]hose who take part in the struggle while not belonging to the armed forces are acting deliberately outside the laws of warfare”.270 It is therefore not unreasonable to argue that “the direct participant should bear the risk of mistake, not his or her opponents, as ihl does not envision the participation of the former in the first place”.271
5.3 What Does It Entail to Be a Lawful Target
5.3.1 Lawful to Attack and Detain
The most important implication of being a lawful target is that you may be attacked at anytime and anywhere, regardless of current activity. The only exceptions arise if the person becomes hors de combat, for example because of injuries, or indicates an intention to surrender. Persons who are lawful targets may be “pursued without warning or any other preliminaries.”272 There is no requirement of force only being used as a last resort.273 Rather, the only limitation on the use of force lies in the explicit prohibitions on the use of certain means and methods such as biological and chemical weapons, denial of quarter and perfidy. It is also prohibited to employ “weapons, projectiles and material and methods of warfare of a nature to cause superfluous injury or unnecessary suffering”.274
The icrc Interpretive Guidance suggests that “it would defy basic notions of humanity to kill an adversary or to refrain from giving him or her an opportunity to surrender where there manifestly is no necessity for the use of lethal force”.275 It is therefore the icrc’s view that “the kind and degree of force which is permissible against persons not entitled to protection against direct attack must not exceed what is actually necessary to accomplish a legitimate military purpose in the prevailing circumstances”.276 Although it is uncontroversial to state that belligerents do not have unlimited rights to choose means and methods of warfare,277 using this as an argument for an obligation to capture rather than kill seems unwarranted. In fact, Dinstein points out that it “would be utterly incompatible with the general practice of States”.278 To be fair, the Interpretive Guidance does state that the obligation to capture rather than kill is not necessarily a legal one; however, either there is an obligation or there is not, and such an obligation would be impossible both to apply and to assess compliance with.279 Furthermore, the basis for the proposed restriction on the use of force is the principle of military necessity. Yet, this principle is already taken into consideration when creating the rules.280 Unless explicitly cited as a factor to be considered, it should not be applied as “a separate restriction that constitutes an additional hurdle over which an attacker must pass before mounting an attack”.281 The opposite conclusion would have the undesirable side effect of permitting loac to be set aside if deemed sufficiently necessary.282
States may still decide to encourage military forces to capture rather than kill opposing forces where this is possible. However, this is most likely done on the basis of military considerations such as perception management among the civilian population and the desire to gather intelligence from captured persons.283 In operations that require close interaction with the civilian population, such as counter insurgency (coin) operations, States may also decide to impose general restrictions on the use of force, for instance through a requirement of ‘minimum use of force’.284 Such policy decisions should not be interpreted as State practice in the context of customary law development without the necessary indications of the required corresponding opinio juris.285 In the case of nato, it may even be a result of States having different views on whether or not the situation should be classified as a conflict. By imposing a requirement of ‘minimum use of force’ or limiting force to the degree, intensity, and duration necessary to achieve the objective, the concerns of nato States that do not consider the operation to involve participation in an armed conflict will be respected.
Even if the capture versus kill debate appears to have favoured the conclusion that there is no new legal duty to capture, it has had the beneficial effect of clarifying the requirements for becoming hors de combat and, in particular, when a person should be deemed captured and thus no longer a lawful target. It is unlawful to attack a person who is hors de combat, but such persons may still be detained. Especially in the context of non-State actors, it has been argued that a person should be considered captured when clearly under the control of the opposing forces and unable to continue fighting.286 This seems like a sound approach, and although it may be difficult to apply at the tactical level, he or she should be captured if it becomes apparent that the person has no means of resisting. Furthermore, as Schmitt points out, “In the case of the weaponless surrounded fighter, the fact that he is unwilling affirmatively to surrender does not preclude his having been captured. (…) But if a person who appears capable and ready to resist has not been captured; he may be killed even if resistance would be completely futile”.287
5.3.2 No Protection against the Effects of Hostilities
Being a lawful object of attack is not the only consequence of being a lawful target. The other consequence is that there is no protection against the effects of hostilities beyond the prohibition on unnecessary suffering and superfluous injury.288 Most importantly, persons who are lawful targets are not required to be taken into account in a proportionality assessment, and there is no duty to minimise losses as generally required as precautions in attack.289 This also means that in situations where the only persons or objects expected to be affected by an attack are lawful targets, there is no requirement to assess whether precautions in attack are required.
As previously explained,290 an increased focus on reducing civilian casualties has resulted in an increased need for determining the status of those expected to be harmed by an attack. Because persons who are lawful targets are not required to be taken into account in the proportionality analysis, incidental loss of persons who are lawful targets are not violating the zero civilian casualty policy. This has resulted in an increased focus on and need for defining the limits of dph, both with regards to the type of actions that may amount to direct participation and the degree of certainty required to make the determination that someone has lost civilian protection. After all, these are not necessarily persons who would have been attacked had they not been in the vicinity of the primary target, that is, a person or object subject to attack. Furthermore, the determination will often need to be made on the basis of the limited facts available at the time, and the decision on whether an attack should be continued, or cancelled or suspended because it is expected to cause excessive incidental loss to civilians, will in many cases have to be taken quickly.291
Despite its important practical implications, this consequence of becoming a lawful target often seems forgotten. In the Interpretive Guidance, the icrc, for instance, mainly focuses on protection against direct attack and the loss thereof. As a result, the organisation misunderstands why many States are sceptical of their restrictive approach to human shields. The concern is not the ability of military forces to direct their attacks at human shields, but rather the effect that the use of human shields have on the operation.292 The issue of human shields is, however, particularly complex because it is in many cases impossible to determine with any degree of certainty whether the persons are there voluntarily or involuntarily.293
5.4 Identification of Lawful Targets: Precautions in Attack and the Issue of Doubt
5.4.1 General Rule on Precautions in Attack
loac requires that those who conduct military operations to take “constant care” (…) “to spare the civilian population, civilians and civilian objects”.294 This is commonly referred to as the requirement to take all feasible precautions. There are two aspects of this duty. First, as a general obligation to spare civilians in the conduct of all aspects of military operations, laid out in AP I Article 57(1). This will apply both to attack and to operations not involving attack, such as the movement or deployment of forces.295 Second, Article 57(2)-(5) sets out the more specific duties, operationalising the general obligation in the context of attack.296 These include requirements concerning verification of target, choice of means and methods of warfare, proportionality, and giving effective warnings. The most relevant aspects of precautions for the current book are verification of targets and proportionality, and these will be considered in further detail below. The duty to take all feasible precautions also extends to taking feasible precautions to reduce the risk of harm to protected persons and objects from the effects of enemy attacks.297 However, this will not be examined in further detail here.
The term “feasible” is generally understood as meaning “that which is practicable or practically possible, taking into account all circumstances ruling at the time, including humanitarian and military considerations”.298 It is in other words clear that “[t]he obligation to do everything feasible is high but not absolute”,299 as was emphasised in the otp Report. Examples of circumstances that may be taken into account are risk to own forces or other security risks; the expected humanitarian benefits from the precaution; and the availability of alternative approaches or weapon systems.300 While some suggest that the financial cost of taking the precaution also may be taken into account,301 this is controversial, although resource considerations, such as the availability of expensive weapon or ammunition, are generally accepted to be relevant for the feasibility consideration.302
But we are obliged to judge the situation as it appeared to the defendant at the time. If the facts were such as would justify the action by the exercise of judgment, after giving consideration to all the factors and existing possibilities, even though the conclusion reached may have been faulty, it cannot be said to be criminal.306 (…) We are concerned with the question whether the defendant at the time of its occurrence acted within the limits of honest judgement on the basis of the conditions prevailing at the time.307
5.4.2 Verification of the Target before and During an Attack
As was explained in Section 5.2 above, individuals or groups become lawful targets under loac on the basis of either their status or their conduct. When planning or conducting an attack, everything feasible must be done to verify that the objectives to be attacked are military objectives and not civilians or subject to special protection.308 If it becomes apparent that the person is not a lawful target after all, the attack must be cancelled or suspended.309 As the ila Study Group on the Conduct of Hostilities in the 21st Century points out, “[t]he obligation to verify is crucial for two reasons: first, it functions to minimize the risk that a target will be mistakenly assessed as qualifying for deliberate attack; second, it functions to maximize the probability of that attacks will only be directed at targets that genuinely contribute to bringing the enemy into submission”.310
Coupled with this duty to take precautions is the presumption of civilian status, dictating that if there is doubt whether a person is a civilian, that person shall be considered to be a civilian.311 The same applies to doubt as to whether a civilian is taking a direct part in hostilities; if there is doubt, the person should be presumed to be protected.312 Although the doubt rule and the precaution requirement are separate rules, they are in practice closely interlinked, and it may be hard to determine where the issue of doubt ends and becomes a question of what is reasonably expected based on the information available at the time.
In the practical application of the principle of civilian immunity and the rule of doubt, (a) commanders and others responsible for planning, deciding upon, or executing attacks necessarily have to reach decisions on the basis of their assessment of the information from all sources which is available to them at the relevant time, (b) it is only in cases of substantial doubt, after this assessment about the status of the individual in question, that the latter should be given the benefit of the doubt and treated as a civilian, and (c) the rule of doubt does not override the commander’s duty to protect the safety of troops under his command or to preserve the military situation.317
Rather than requiring a civilian presumption to be applied in any cases of doubt, the UK loac manual sets out a higher threshold of substantial doubt. The Norwegian manual similarly stresses that due to the fog of war, some degree of doubt must be tolerated. However, rather than focusing on the degree of doubt, it stipulates that if doubt remains after all feasible precautions have been taken to verify the target, attacks may only be carried out if it is considered highly probable or most likely that the person is a lawful target.318 The U.S. has also objected to the rule that any doubt should result in presumption of civilian protection.319 They emphasise that the determination cannot be based merely on “hypothetical or speculative considerations regarding their possible current status as a military objective”,320 but must be made “in good faith based on the information available to them in light of the circumstances ruling at the time”.321
It is clear that States do not support the interpretation “even if there is only a slight doubt”,322 more information must be sought, as suggested in the icrc Commentary. However, rather than defining the degree of doubt, the focus should be on what is expected of the Commander. As Henderson explains, “the level of verification required to reduce doubt, and the degree of acceptable doubt, will vary depending upon the likely adverse consequences of a wrong decision”.323 In practice, the test is therefore whether the attacker acted in honest belief and whether a reasonable attacker would have made the same decision.324
The first question when assessing the legality of targeting is therefore what the attacker honestly believed: Did he honestly believe he had sufficient information about the target and that the proposed target was a lawful one?325 If not, he would be bound by the presumption of civilian protection, and must either cancel the attack or take further measures to gather information about the target and then reassess whether it is a military objective. Second, the targeting decision must be assessed on the basis of the objective standard of reasonableness: “whether a reasonable attacker, having employed all reasonably available means of verification under the circumstances, would initiate the attack”.326 There are two aspects to this question. Have the attackers employed all reasonable available means of verification under the circumstances? The attacker is required to take into account all reasonably available information; if information is ignored, this would undermine the reasonableness of the decision.327 Furthermore, would a reasonable attacker initiate the attack on that basis? It need not be the only possible conclusion to be drawn, but it must be a decision a reasonable attacker could have made.328 As Merriam concludes, the decision to target will be lawful “[w]hen the attacker has taken all feasible measures to resolve his doubt and verify the character of the target, and he ‘reasonably concludes’ that it is military”.329 When assessing whether a decision to initiate an attack was reasonable, the question will therefore focus less on the degree of doubt and more on whether the attacker had done enough to gain information about the target, that is, whether sufficient precautions had been taken.330
In order to remove or at least reduce doubt concerning a person’s status as a lawful target, or said differently, in order to make a good faith assessment of this issue, it is required to do ‘everything feasible to verify’ the status. As mentioned above,331 the requirement is to do as much as possible in the circumstances, taking into consideration amongst others the risk to own forces, military advantage expected to be achieved, and available resources.332 In other words, it is very context dependent. For instance, the situation will differ for those planning an attack from a distance and those with the opposing forces in sight. As the UK loac Manual explains, “[t]he former has more time to make up his mind; the latter is more easily able to verify the target”.333 Time sensitivity will in many cases be the main limitation on how much may be done to verify the target.334 For instance, if a patrol is unable to remain for a longer period of time in the area without drawing fire from nearby opposing forces, this will limit the amount of information they have the opportunity to gather about a person of interest. Similarly, if a person known to be a lawful target is potentially located but may soon disappear again, the military forces tasked with attacking that person will only be able to do so much to verify his identity before it is too late.
Whether it is reasonable to carry out the attack based on the information available will depend, inter alia, on the importance of carrying out the attack: if the person is a high value target or considered to be a considerable threat to the forces, it may be reasonable to carry out the attack even if uncertainty remains.335 It is also suggested that the expected risk of incidental harm to civilian persons or objects resulting from attacking the target should influence what measures of verification may reasonably be expected to be carried out before determining that the target is a military objective.336 If the target is less important or the situation less urgent, it would be feasible to take further measures to verify that the person is a lawful target. Still, there is a limit to the amount of resources military forces may be expected to employ in order to verify the target. If it is not feasible to do more or it is not reasonable to expect more to be done to ensure the verification, the question becomes what is reasonable to conclude based on the information available.
The requirement to take all feasible precautions differs depending on the stage of the targeting process. In the initial stages, when the attack is being planned or initiated, the requirement of taking all feasible measures in order to verify the status of the target applies in full. The attack should not be planned to be carried out unless the target is deemed lawful. This requires knowledge of what information is reasonably available and procedures that ensure that such information is collected and employed. For instance, Commanders, or those who plan and decide upon an attack on their behalf, must be expected to know what Information Surveillance and Reconnaissance (isr) assets the forces have available and what they can provide.337 Although the provision makes reference to ‘those who plan or decide upon an attack’,338 this should not be interpreted as excluding others from having an obligation to stop the attack if they become aware that the target is not lawful.339
During the force execution phase, however, a different threshold is applied – the attack may be carried out unless it becomes apparent that the target is not lawful.340 This makes operational sense. First, the forces carrying out the attack may not have access to all the information that the planners used to determine the legality of the attack.341 Furthermore, once the attack has begun, it will in most cases be difficult to continue to check that the target is lawful, and some degree of doubt is likely to continue to exist or even arise. This will apply equally to pre-planned or ordered attacks and targets selected by the person carrying out the attack. It should be emphasised that although ‘becomes apparent’ may appear to be a passive requirement, the attacker “has a duty to monitor the attack as long as it is feasible to do so”.342
The person who executes the order is responsible for the execution, but not anything else. He must be able to presume that the decision made by competent authorities within the organisation is correct. Even if it later became clear that the killing was based on incorrect information or the order was unreasonable, the person executing the order must be free from responsibility provided he did not know of the errors.345
However, if for some reason it becomes clear to those carrying out the attack that the initial assessment of the target was erroneous, the attack must be halted.
The provision makes reference to either ‘cancel or suspend’.346 If it is expected that the status of the person as a lawful target will change, such as when the person is no longer hors the combat or resumes a direct part in hostilities, the attack may be temporarily suspended. In general, however, it is more appropriate to consider suspending rather than cancelling an attack if the target is deemed lawful but attacking it at a certain time is expected to cause excessive collateral damage. Although the duty to cancel or suspend an attack applies to all levels of the military chain of command, it should be noted that not all persons involved in the execution of an attack will have the authority to cancel or suspend an attack, or have the practical opportunity to do so.347 This should be taken into consideration when assessing whether the duty to cancel or suspend attacks has been complied with.348 This does not, however, mean that the absence of military authority to halt an attack in itself is sufficient reason not to prevent or attempt to prevent an unlawful attack.
5.4.3 ‘Positive Identification’
The process of identifying the target is commonly referred to as ‘positive identification’, or pid. ‘Positive identification’ is an operational concept intended to ensure that the proposed target is a legitimate target.349 The concept was originally used by the U.S. Air Force in the aftermath of the 1991 Gulf War in Operations Provide Comfort (later Northern Watch) and Southern Watch to set the standard for identification when enforcing no-fly zones. It required an almost no-mistake standard.350 When it was introduced to ground operations in Afghanistan and subsequently Iraq, it caused some confusion.351 What did ‘positive’ entail? Did it require more than the loac standard for verification of lawful targets?352 The definition eventually agreed upon was that “pid is a reasonable certainty that the proposed target is a legitimate military target”, which was met with similar questions and criticisms.353
Despite the initial confusion caused by the application introduction of an operational concept intended for the air to air context to other domains, it has now evolved into a useful tool for ensuring that the requirements of military necessity and distinction are met.354 As such, ‘positive identification’ is nothing new. In fact, it has been argued that “pid has always been an implicit requirement before using force”.355 The added value in requiring pid is that it emphasises the need for both legal and operational assessment of the target and its so-called ‘geospatial location’;356 the person must be assessed to be a lawful and approved target (also known as target validation),357 and the location or presence of the lawful target must be verified. The distinction may appear artificial, but as will be explained below, it is an important distinction in operations where the legal status may be determined by other people than those who carry out the attack.358 Furthermore, because pid integrates both the legal and operational requirements for target validation, it is a contextual operationalisation of the loac rules on distinction, rather than a restatement of the law. In addition to ensuring compliance with applicable law, the procedures and requirements for vetting and validation of targets may be subject to mission specific requirements that are stricter than the loac rules.359 They will usually also be more detailed than the law requires, for instance in setting out requirements for type and quality of information, and number and forms of means of identification.
The first requirement in a pid analysis focuses on whether the person is a lawful target and hence whether it is militarily necessary to engage the target. The determination should also entail a clarification of whether the person is approved for targeting, that is, whether the person falls within a pre-approved target set or is subject to limitations, such as those set out in the Restricted Target Lists (rtl) or No Strike Lists (nsl), and the attack on the person is authorised by roe and other targeting procedures.360 Because these may vary from operation to operation, or even within an operation, the pid standard varies, in the same way that the roe differ.361 Second, the person or persons deemed to be lawful (and approved) targets must be accurately located at the time of attack. Conversely, it is not sufficient to declare that a person of interest is located; that person must be assessed to be a lawful target.362 By way of example, locating a person suspected of involvement in a previous attack may not be sufficient to warrant attack on that person. In some cases, it is also necessary to conclude that the person is a lawful target at the time of being located, for instance, due to dph on a regular basis or membership in an organised armed group.
The emphasis on the two aspects of pid will vary depending on the operation. If the operation is based on pre-planned target sets, the assessment of whether the person may be attacked is already undertaken, and the focus is on locating that person. By way of example, if person X is approved for lethal targeting through a deliberate targeting process, the emphasis will be on locating that person in order to make sure he or she actually is person X. If the operation is dynamic, the military forces may come under attack or face other hostile acts in response to which they may use force in accordance with loac. In such cases, the focus will be on locating the origin of the threats.363 As Pennekamp explains, ‘positive identification’ will require that the military forces get “a close enough look at the situation to be reasonably certain that the individual the soldier is about to engage is the actual source of hostility”.364 The focus during dynamic operations may also be on identifying persons of interest as lawful targets. For instance, military forces may come across persons perceived to be behaving in a suspicious manner, and will need to determine whether they are lawful targets.
In joint operations, where the forces identifying a person as a lawful target are not the ones carrying out the attack, both aspects of pid may play a central role. First, the determination must be made that a person is a lawful target. Second, the persons carrying out the attack must verify that the person they are about to attack is the same person. For instance, this situation may arise if ground forces have identified a person as a lawful target, but the attack is carried out by air forces. The person directing the Close Air Support may be able to provide the coordinates for the person to be attacked, or rely on accurate descriptions of the person’s whereabouts and movements. Once established, pid must be maintained until the attack is carried out.365 If a person who has been assessed to be a lawful target, whether by planners or by those about to execute an attack, has been located but then lost again, those carrying out the attack must relocate the target in order to re-establish pid and ensure that the attack is in fact directed at a lawful target.
Positively identifying a person is not, however, sufficient for carrying out an attack lawfully. In particular, it is necessary to assess and minimise the risk of harm to protected persons and objects, and to cancel or suspend the attack if this harm is expected to be excessive. This is further considered in the next section.
5.5 Carrying Out Attacks
The preceding sections have primarily focused on the categories of persons that are lawful targets and their identification. This final section will present a brief overview of the rules regarding how attacks should be carried out once it is established that the target is considered to be lawful. These requirements will usually not be reflected in the roe themselves, beyond the reference to necessity and proportionality in the gentext.366 Instead, they will be set out in documents such as sops and Tactical Directives providing context specific guidance on the application of roe. For instance, procedures on Collateral Damage Estimation will set out requirements for the proportionality assessment for use of force in the joint targeting process.367
Even when directing military operations at military objectives, loac makes it clear that “the right of the Parties to the conflict to choose methods or means of warfare is not unlimited”.368 For instance, it is prohibited to employ means or methods of warfare that are of a nature to cause unnecessary suffering or superfluous injury or which are intended, or may be expected, to cause widespread, long-term and severe damage to the natural environment.369 Furthermore, those who plan and decide upon attacks are required to take all feasible precautions to minimise the adverse effects on the civilian population and protected objects, and to refrain from carrying out attacks that are expected to cause excessive collateral damage.370 Attacks that are expected to cause excessive collateral damage are considered indiscriminate and hence unlawful.371 If it becomes apparent during the force execution phase that the attack, despite the initial assessment of legality, may be expected to cause excessive collateral damage, it must be cancelled or suspended.372 In order to reduce the risk of collateral damage, it is required to give effective warning of the attack, provided the circumstances so permit;373 and, if it is possible to choose between several military objectives for obtaining a similar military advantage, the Parties must choose the one that is expected to cause least collateral damage.374
deciding to launch an attack which may be expected to cause incidental loss of civilian life, injury to civilians, damage to civilian objects, or a combination thereof, which would be excessive in relation to the concrete and direct military advantage anticipated.376
Rather than using to the phrase “injury to civilians, damage to civilian objects, or a combination thereof” when discussing loac proportionality, it is common to only make reference to collateral damage, as is done in this book. Strictly speaking, collateral damage refers to harm to civilian objects, while harm to civilians is labelled ‘incidental injury’. However, this distinction is not usually made.377 Collateral damage may amongst others be the result of: civilians working with military forces or living in military bases; civilians working or living in the vicinity of military objectives; or civilians temporarily being present on or near a military objective at the time it is being attacked.378
As mentioned in the introduction to this section, military forces are required to take all feasible precautions in the choice of means and methods of attack in order to minimise the risk of collateral damage, and must refrain from deciding to launch attacks expected to cause excessive collateral damage. Furthermore, ‘feasible’ is generally understood to mean that which is practicable or practically possible, taking into account all circumstances ruling at the time, including humanitarian and military considerations.379 According to Schmitt and Widmar, this entails that “they need only select less harmful means or methods that do not involve sacrificing military advantage and that are feasible”.380 For example, while using Precision Guided Munitions (pgms) in most cases will reduce the risk of collateral damage, military forces usually will not have unlimited access to such munitions, and it will therefore often not be feasible to use them for all attacks involving such a risk.381 The decision may be made to only use such munitions for attacks on particularly important targets or attacks that would be expected to cause excessive collateral damage if executed with less precise means.382 Similarly, while the use of ground forces may enable greater precision and more opportunity to verify both the presence of lawful targets and absence of protected persons and objects than using air assets, sending ground forces sufficiently close to the target may involve too great a risk to them.383
It is not sufficient to have taken all feasible precautions to minimise the risk of collateral damage; if the expected collateral damage is deemed excessive to the concrete and direct military advantage anticipated to be achieved, the attack cannot be carried out. According to Dinstein, “‘excessive’ means that disproportionality is not in doubt”.384 Excessive must not be confused with extensive; if the target is sufficiently important, extensive collateral damage need not be excessive.385 Similarly, if the target is of minimal importance, even limited collateral damage could be excessive. Even if the attack was not initially expected to cause excessive collateral damage, it must be cancelled or suspended until the circumstances permit the attack to be carried out if it later becomes apparent that the attack nonetheless must be anticipated to cause excessive damage.386 The distinction between the planning and force execution phases of an attack is examined in detail in the context of verification of targets above, and applies equally here.387
The proportionality rule refers to ‘expected’ collateral damage and ‘anticipated’ military advantage. Compliance with the rule is therefore not assessed on the actual outcome of the attack, but rather on the information reasonably available at the time.388 This includes the planning, approval and force execution phases of the attack.389 Although expectations or anticipation necessarily entail a subjective evaluation of the facts and the military advantage to be gained from the attack, it must nonetheless be a reasonable conclusion.390 If the attack was expected not to cause excessive collateral damage, but the battle damage assessment (bda) or other information reveals that the attack nonetheless did cause excessive collateral damage, this should be taken into consideration when assessing the expected collateral damage for future similar attacks.391
The military advantage anticipated must be concrete and direct, not merely hypothetical.392 It does not need to be substantial, however.393 It is also generally accepted that the military advantage of an attack may be based on an overall assessment of the operation or campaign as a whole rather than just the attack itself.394 For instance, the military advantage expected from the nato bombing campaign on the Serbian TV and Radio Station, radio and TV relay sites, and electricity grids, was the impact the attacks would have on the Serbian C3 (Communications, Command and Control) assets, not merely the effect on the respective targets.395 As a result, it may be difficult for those who execute an attack to determine the proportionality of that attack; they may not have the overview of the entire attack needed to assess its overall advantage.396 The inclusion of more than merely tactical or local gains in the assessment of military advantage must not, however, be interpreted to permit the armed conflict as a whole to be taken into account.397
On the other side of the calculus, only loss of life, injury or damage is considered collateral damage; mere inconveniences are not taken into account.398 Loss of economic income or temporary disruption to civilian life are not considered collateral damage.399 There has been some debate regarding the inclusion of indirect or secondary harm to civilians and civilian objects.400 In order to have the potential for being expected, the collateral damage must be foreseeable. If the harm is foreseeable, it is difficult to see how this could not be taken into account, even if it is indirect.401 Rather than focusing on whether the effect is direct or indirect, the focus should be on whether it is reasonably foreseeable and should therefore be expected.402 The issue of indirect or reverberating effects have been particularly emphasised in the context of explosive weapons being used in densely populated areas.403
Although loac the rules on proportionality only makes express reference to the avoidance of harm to civilians and civilian objects, attacks on military hospitals and medical personnel in the recent years have initiated a debate as to whether harm to them should be viewed as included in the proportionality assessment.404 Similar discussions have been raised concerning the wounded, sick and injured military personnel,405 and pow camps.406 There are two approaches. Either the aforementioned persons and objects are civilians or civilian objects, or the rule is not limited to these categories. While the definition of military objects entails that all objects that are not targetable are civilian, it is harder to make the argument that military medical personnel or combatants hors de combat are civilians.407 In order to include these categories of persons in the proportionality analysis, the argument must therefore be made that the reference to civilians should be read as referring to all protected persons, both civilians and those with special protection.408
In Corn’s view, while the inclusion of specially protected persons and objects in the precautionary rules in general makes sense, expanding the proportionality rule beyond civilians and civilian objects would entail “an unrealistic constraint on tactical action”.409 It would in essence entail a reassessment of proportionality after every shot is fired. In the case of an attack on two soldiers, for instance, once the first is injured and therefore hors de combat, the attack on the remaining soldier could be deemed to cause excessive harm. Perhaps more important, however, is the fact that the violation of the prohibition on excessive collateral damage would be a war crime,410 and the principle of legality in criminal law demands that the scope of a crime must be clearly defined and have a clear legal basis.411 The crime cannot be expanded by way of an extension by analogy.412 The non-application of the principle of proportionality does not, however, mean that such persons or objects should not be taken into account. Customary law, as reflected in the Marten’s clause,413 demands that constant care must be taken to spare them, and to mitigate harm to the extent feasible.414
Recent nato operations such as Afghanistan and Libya have, as a result of the emphasis on reducing civilian casualties to an absolute minimum, resulted in the loac proportionality rule playing a less central role. As explained by Schmitt, “in Afghanistan the operational concern was the mere fact of collateral damage, not whether that damage expected to be caused was excessive relative to military advantage. Rules of engagement so embraced this casualty aversion that the legal principle of proportionality never loomed large”.415 For isaf, the policy on avoiding civilian casualties was set out in the previously mentioned civcas Directives.416 Although the civcas Directives were a reflection of policy rather than an expression of opinio juris, it may still be viewed as an example of precautions the Commander considered feasible to carry out, especially in an operation where the protection of civilians and civilian support were central to the mission. As Johansen explains, “[p]recautionary measures in order to reduce collateral damage therefore simultaneously enhance mission accomplishment. Consequently, the threshold for feasibility becomes lower. When precautionary measures align with mission accomplishment, these measures are also more easily assessed as feasible”.417
The lex specialis concept is further examined in Section 4.4.
See also Section 1.3.1 where armed conflict is defined, including the distinction between iacs and niacs.
See also Dino Kritsiotis, ‘War and Armed Conflict: The Parameters of Eenquiry’, Chapter 1 in Liivoja, Rain and Tim McCormack (eds.), Routledge Handbook of the Law of Armed Conflict (Routledge, Milton Park, Abingdon, Oxon, 2016) pp. 9–10. For a detailed overview of the historic evolution and sources of the law concerning targeting, see for instance William H. Boothby, The Law of Targeting (Oxford University Press, Oxford, 2012) pp. 13–29. On the history of loac in general, see e.g. Frits Kalshoven, ‘The History of International Humanitarian Law Treaty-Making’, Chapter 2, in Rain Liivoja and Tim McCormack (eds.), Routledge Handbook of the Law of Armed Conflict (Routledge, Milton Park, Abingdon, Oxon, 2016), and Geoffrey S. Corn et al., The Law of Armed Conflict – an Operational Approach (Wolters Kluwer Law & Business, New York, 2012) Chapter 2.
Yves Sandoz, Christopher Swinarski, and Bruno Zimmermann (eds.), Commentary on the additional protocols of 8 June 1977 to the Geneva Conventions of 12 August 1949 [hereinafter: AP Commentary] (icrc, Martinus Nijhoff Publishers, Geneva, 1987) para. 1389 and 2206. See also the St. Petersburg Declaration, which was intended to “fixed the technical limits at which the necessities of war ought to yield to the requirements of humanity”. St. Petersburg Declaration (Declaration Renouncing the Use, in Time of War, of Explosive Projectiles Under 400 Grammes Weight) [29 November / 11 December 1868] in Dietrich Schindler and Jiri Toman, The Laws of Armed Conflicts (Martinus Nijhoff Publisher, 1988) pp.102ff., Preamble. See further Program on Humanitarian Policy & Conflict Research at Harvard Univ. [hpcr], Manual on International Law Applicable to Air and Missile Warfare [hereinafter: AMW Manual] (2009, version 2.1 March 2010, hpcr, Harvard University) Rule 4, Commentary para. 2.
Michael N. Schmitt, ‘The Interpretive Guidance on the Notion of Direct Participation in Hostilities: A Critical Analysis’, 1 Harvard National Security Journal 5 (2010), p. 6.
Hans Boddens Hosang, Rules of Engagement (PhD thesis, University of Amsterdam 2017, available at
See discussion of the ‘Lotus principle’ in Section 4.1.
Sigrid Redse Johansen, On military necessity: And the commander’s assessment of military necessity under the international law of armed conflict during conduct of hostilities (PhD thesis, University of Oslo, 2017, copy on file with the author, forthcoming cup publication) pp. 38–41.
ibid, p. 65. See also discussion in Sandoz et al. (eds.), AP Commentary (n 4) para. 1390, and icty, Final Report to the Prosecutor by the Committee Established to review of the nato bombing campaign against the Federal Republic Yugoslavia [hereinafter: otp Report] (13 June 2000,
Protocol Additional to the Geneva Conventions of 12 August 1949, and relating to the Protection of victims of International Armed Conflicts (Protocol I) [1977], printed in Schindler and Toman, The Laws of Armed Conflicts (n 4) pp. 711ff, Article 49(1). See further Section 5.1.3.
In particular, the rules on attack, which is the focus of this Chapter, have become “nearly indistinguishable”: Michael N. Schmitt, ‘Targeting and International Humanitarian Law in Afghanistan’, 85 International Law Studies 307 (2009), p. 308. See also Prosecutor v. Tadic (Decision on the Defence Motion for Interlocutory Appeal on Jurisdiction), Case No. IT-94-1-A, icty, Appeals Chamber, 2 October 1995, para. 127. For an overview of relevant customary international law, see e.g. Jean-Marie Henckaerts and Louise Doswald-Beck, Customary International Humanitarian Law [hereinafter: icrc cil Study] (Cambridge University Press, Cambridge, 2005, available at
There are various approaches to the basic principles of loac. nato stanag 2449, for instance, include humanity instead of unnecessary suffering, with the latter being a part of the former. nato, stanag 2449: Training in the Law of Armed Conflict, Annex F to ATrainP-2, Edition A Version 1, March 2013 (available at
On the principle of necessity in particular, see Sandoz et al. (eds.), AP Commentary (n 4) para. 1389; Boothby, The Law of Targeting, ibid, p. 58; Boddens Hosang, Rules of Engagement (n 6) p. 209; Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 41; and Kleffner, ‘Section IX of the icrc Interpretive Guidance on Direct Participation in Hostilities’ (n 6) p. 41. For an in-depth discussion on balance between humanity and military necessity, see Michael N. Schmitt, ‘Military Necessity and Humanity in International Humanitarian Law: Preserving the Delicate Balance’ 50 Virginia Journal of International Law 796 (2010). See also Johansen, On military necessity (n 8) pp. 112–114.
Sandoz et al. (eds.), AP Commentary (n 4) para. 2206.
Kjetil Mujezinovic Larsen Camilla Guldahl Cooper and Gro Nystuen, ‘Introduction by the editors: is there a ‘principle of humanity’ in international humanitarian law?’, Chapter 1 in Larsen et al., Searching for a ‘principle of humanity’ in international humanitarian law (n 12) p. 1.
This principle is sometimes also referred to as the principle of discrimination or identification. UK Ministry of Defence, The Manual of the Law of Armed Conflict [hereinafter UK loac Manual] (Oxford University Press, Oxford, 2004) Section 2.5.1 (Distinction). For a more in-depth study of the principle of distinction, see John J. Merriam, ‘Affirmative Target Identification: Operationalizing the Principle of Distinction for U.S. Warfighters’, 56 Virginia Journal of International Law 83 (2016), pp. 86–107.
AP I (n 10) Article 48, considered to reflect customary law applicable in both iacs and niacs. See icrc cil Study (n 11) Rule 1. See also UK loac Manual, ibid, Section 2.5 (Distinction).
Kenneth Watkin, ‘Opportunity Lost: Organized Armed Groups and the icrc “Direct Participation in Hostilities” Interpretive Guidance’, 42 New York University Journal of International Law and Policy 641 (2009–2010), p. 646.
See e.g. Johansen, On military necessity (n 8); Nobuo Hayashi, ‘Contextualizing military necessity’, 27(1) Emory International Law Review 189 (2013); Nobuo Hayashi, ‘Military Necessity as Normative Indifference’, 44(2) Georgetown Journal of International Law 675 (2013); and Nobuo Hayashi, ‘Requirements of Military Necessity in International Humanitarian Law and International Criminal Law’, 28(1) Boston University International Law Journal 39 (2010).
UK loac Manual (n 16) Section 2.2 (Military Necessity), footnote omitted. See also Sandoz et al. (eds.), AP Commentary (n 4) para. 1389: “[m]ilitary necessity means the necessity for measures which are essential to attain the goals of war, and which are lawful in accordance with the laws and customs of war.”, and the St. Petersburg Declaration (n 4) where it is noted that “[t]he only legitimate object which States should endeavour to accomplish during war is to weaken the military forces of the enemy.”
UK loac Manual (n 16) Section 2.4 (Humanity).
AP I (n 10) Article 57(2)(a)(ii). See also UK loac Manual, ibid, Section 2.6 (Proportionality).
AP I, ibid, Article 51(4) and 51(5)(b), considered to reflect customary law applicable in both iacs and niacs. See icrc cil Study (n 11) Rules 11–12.
See also U.S. Department of Defense (DoD), Law of War Manual (December 2016 update, Office of the General Counsel of the Department of Defense, Washington, 2016, available at
Richard Baxter, ‘The Municipal and International Law Basis of Jurisdiction over War Crimes’, 28 British Yearbook of International Law 382 (1951), p. 385.
Kritsiotis, ‘War and Armed Conflict: Parameters’ (n 3) p. 7.
See also Ian Henderson and Bryan Cavanagh, ‘Military members claiming self-defence during armed conflict – Often misguided and unhelpful’ [hereinafter: ‘Claiming self-defence on the battlefield’], in Jadranka Petrovic (ed.), Accountability for Violations of International Humanitarian Law – Essays in honour of Tom McCormack (Routledge, London/New York, 2016) p. 77.
Hague Convention (IV) Respecting the Laws and Customs of War on Land of 1907 and its Annex: Regulations Respecting the Laws and Customs of War on Land (‘1907 Hague Regulations’), printed in Schindler and Toman, The Laws of Armed Conflicts (n 4) pp.66ff, Articles 1 and 2, and AP I (n 10) Article 43(3).
Geneva Convention (iii) relative to the Treatment of Prisoners of War (GC III) [1949], printed in ibid, pp. 507ff., Article 87. See also discussion in Henderson and Cavanagh, ‘Claiming self-defence on the battlefield’ (n 27) p. 78.
United States v. Lindh, 212 F. Supp. 2d 541, 553 (E.D.Va 2002): “Lawful combatant immunity, a doctrine rooted in the customary international law of war, forbids prosecution of soldiers for their lawful belligerent acts committed during the course of armed conflicts against legitimate military targets”. Cited in John J. Merriam, ‘Natural Law and Self-Defense’, 206 Military Law Review 43 (2010), p. 73.
Article 31(3) of the Rome Statute allows the Court to apply defences not expressly dealt with in the Article 31, provided they are found within the possible sources of law as listed in Article 21. Article 21(b) permits the Court to apply “applicable treaties and the principles and rules of international law, including the established principles of the international law of armed conflict”. Rome Statute of the International Criminal Court [hereinafter: Rome Statute], opened for signature 17 July 1998, 2187 UNTS 3 (entered into force 1 July 2002).
Boothby, The Law of Targeting (n 3) p. 433.
On the relationship between self-defence and loac/lawful acts of war, see further Sections 4.4 and 8.2.3.
Sasha Radin and Michael N. Schmitt, ‘Investigations Under International Humanitarian Law’, in Rain Livoja and Tim McCormack (eds.), Routledge Handbook of the Law of Armed Conflict (Routledge, London, 2016) p. 563, footnote omitted.
Gary P. Corn, ‘Should the Best Offense Ever be a Good Defense? The Public Authority to Use Force in Military Operations: Recalibrating the Use of Force rules in the Standing Rules of Engagement’, 49 Vanderbilt Journal of Transnational Law 1 (2016), p. 28 referring to U.S. Manual for Court-Martials Rule (rcm) 916(c), and fn. 118, referring to Albania, Australia, Brazil, Canada, Israel, Kosovo and Spain. He only makes reference to a public authority defence, as it the approach in the U.S., but he does not appear to consider the distinction set out above.
As mentioned in Section 1.4, the implications of monist or dualist approaches to how States give effect to international law in their respective domestic legislation is beyond the scope of this book.
Only the use of force by the police for law enforcement purposes is regulated, e.g. Lov om Kystvakten (re. The Coastguard) LOV-1997-06-13-42; Instruks om Forsvarets bistand til politiet (re. assistance to the Police) FOR-2012-06-22-581; Instruks om tildeling av begrenset politimyndighet til militære grensevakter ved den norsk-russiske grensen (re. limited police authority to military border guards on the Norwegian-Russian border) FOR-2005-08-05-851; and § 101(3) of the Norwegian Constitution of 1814, updated 2018, regarding the use of military force against Norwegians in a situation of riot.
See e.g. nou 1992: 23 Ny straffelov – alminnelige bestemmelser (Norwegian Green Paper 1992: 23 New penal code – general rules), p. 89, Section 5.6 which due to its context focuses on the police, but also provided general comments on State use of force. (“Offentlige myndigheter, særlig politiet, har i blant rett til å foreta handlinger som ellers er straffbare. Det er i dag ikke generelt regulert ved lov når i hvilken utstrekning politiet har rett til å bruke makt for å gjennomføre sine tjenestehandlinger”).
Rt 1946 s. 880.
See e.g. Ståle Eskeland, Strafferett (5. ed. by Alf Petter Høgberg, Cappelen Damm Akademiske, Oslo, 2017) p. 177; Johs. Andenæs, Alminnelig strafferett (5.ed, Universitetsforlaget, Oslo, 2004) p. 150; Arne Willy Dahl, Håndbok i militær folkerett (Cappelen Akademiske Forlag, Oslo, 2008) pp. 467–468.
nou 2002:4 Ny straffelov (Norwegian Green Paper 2002: 4 New penal code), Chapter 8, Subsection 8.3.1, pp. 209–210. See also Ot.prp.nr.90 (2003–2004) Om lov om straff (straffeloven) (Proposition Concerning the law on penalties (penal code)), Chapter 9, Subsection 9.2.6.
Lov om straff (Straffeloven) (Norwegian Penal Code of 1905), LOV-2005-05-20-28, 2005 update, Section 2 (previously Section 1.2), known as the ‘folkerettsreservasjon’, meaning international law reservation.
nou 2002:4 (n 41) Chapter 8, Subsection 8.3.1, pp. 209–210. See also Ot.prp.nr.42 (1995–1996): Om lov om endringer i straffeloven (forholdet til folkeretten – gjennomføring av havrettskonvensjonens krav til straffelovgivningen) (Proposition Concerning the law on changes to the Penal Code (relationship with international law – integration of the UN Convention on the Law of the Sea (unclos) into the criminal law)), Chapter 6, pp. 8–11, where the ‘folkerettsreservasjon’ or international law reservation was also deemed sufficient.
It should be noted that a recent change to the Constitution (Norwegian Constitution of 1814 as updated 2014, § 113), whereby any infringement by the authorities towards individuals must be based on law, may result in the customary rule authorising the Armed Forces to use force to be codified. However, this would also require the relationship between ihrl and loac to be spelled out, which is very complex, and no such law is expected to be developed in the near future.
Danish Penal Code of 1933 (lbk nr 977 af 09/08/2017), §12.
Article 25 of the Basic Law for the Federal Republic of Germany in the revised version published in the Federal Law Gazette Part iii, classification number 100–1, as last amended by Article 1 of the Act of 23 December 2014 (Federal Law Gazette I p. 2438) (translated by Professor Christian Tomuschat and Professor David P. Currie,
Frits Kalshoven and Thyla Fontein, ‘Some Reflections on Self-Defence as an Element in Rules of Engagement’ in Mariëlle Matthee etc. (eds.), Armed Conflict and International Law: In Search of the Human Face (tmc Asser Press, The Hague, 2013) p. 109, citing Article 38 of the Dutch Military Criminal Code read in conjunction with Article 71. Note that the cited Article 38 is now paragraph 1 of the updated Article 38. See also Boddens Hosang, Rules of Engagement (n 6) p. 385.
Article L4123-12 of the French Code of Defense (
American Law Institute, Model Penal Code: Official Draft and Explanatory Notes (Complete Text of Model Penal Code as Adopted at the 1962 Annual Meeting of the American Law Institute at Washington, D.C., May 24, 1962, The Institute, Philadelphia, 1985).
ibid, Section 3.03(2)(b).
ibid Section 3.03(1)(d). This is dealt with under the Section on execution of public duty, and has been interpreted by some to mean that all lawful conduct of war is dealt with as a public authority justification. See e.g. Henderson and Cavanagh, ‘Claiming self-defence on the battlefield’ (n 27) p. 85. This ignores the distinction made between rule 3.03(1) on justifications and 3.03(2) declaring lawful acts of war an exception.
United States Manual for Courts-Martial (mcm) (2016 Edition,
U.S. rcm 916(c), ibid. For a commentary on the rule, see e.g. Corn, ‘Public Authority to Use Force in Military Operations’ (n 35) pp. 27–31.
Merriam, ‘Natural Law and Self-Defense’ (n 30) p. 73. Note that in the current thesis a distinction is made between the criminal law concept of self-defence on the one hand and operational concepts of ‘self-defence’, whereby military forces defend themselves during combat in accordance with loac, on the other. This distinction is further dealt with in Chapter 8, in particular Section 8.3.
Sir Edward Coke, Institutes of the Laws of England: Concerning high treason and other pleas of the crown and criminal causes (Printed for E. and R. Brooke, Bell-Yard, near Temple-Bar, London, 1797) p. 47. Note that death occurring within a day and a year is no longer considered part of the definition.
See Michael Wood, Iraq, letter to ‘PS’ dated 15 October 2002, Iraq Inquiry, Written Evidence, January 26, 2010 (
David Ormerod et al., Smith and Hogan’s Criminal Law (Oxford University Press, Oxford, 2015) p. 562. See also Blackstone, Commentaries on the Laws of England in Four Books, Book IV, ibid, p. 118 (“However, the position itself, that it is at any time lawful to kill an enemy, is by no means tenable: it is only lawful, by the law of nature and nations, to kill him in the heat of battel, or for necessary self-defence.”); and Alexander Fraser Tytler, An essay on military law, and the practice of Courts-Martial (R.E. Mercier and Co, No. 31 Anglesea-Street, Dublin, 1800) p. 66 (“It is not common to put to death prisoners of war. They are detained in safe custody, till they are exchanged by cartel, for an equal number of our subjects who may be in the hands of the enemy, or till the conclusion of a peace restores them to their liberty: and in the mean time they are treated with humanity.”).
Wood, Iraq, letter to ‘PS’ dated 15 October 2002 (n 55) para. 5. It does not appear to be limited to the act of killing.
Radin and Schmitt, ‘Investigations under ihl’ (n 34) p. 562.
ibid, p. 563.
In the context of loac violations, this duty is set out in AP I (n 10) Article 87(3), considered to reflect customary law applicable in both iacs and niacs. See icrc cil Study (n 11) Rule 158. The procedures for dealing with suspected violations of domestic law will be set out in the respective domestic legislations.
According to Radin and Schmitt, in the absence of a purported violation, “Commanders, of course, may nevertheless choose to initiate investigations into any incident out of policy or operational concerns”. Radin and Schmitt, ‘Investigations under ihl’ (n 34) p. 563.
AP I (n 10) Article 43(2).
An early recognition of the idea that civilians directly participating in hostilities do not have the right to do so may be found in the 1863 Lieber Code Article 82: “such men, or squads of men, are not public enemies, and, therefore, if captured, are not entitled to the privileges of prisoners of war, but shall be treated summarily as highway robbers or pirates.” U.S., ‘Instructions for the Government of Armies of the United States in the Field’ (Lieber Code) [1863], printed in Schindler and Toman, The Laws of Armed Conflicts (n 4) pp. 3ff. See also W. Hays Parks ‘Air War and the Law of War’, 32 Air Force Law Review 1 (1990), p. 118.
Johs. Andenæs, ‘Okkupasjonstidens «likvidasjoner» i rettslig belysning’, Tidsskrift for rettsvitenskap (tfr) 1948, 1–31, p. 5, own translation.
ibid.
Protocol Additional to the Geneva Conventions of 12 August 1949, and relating to the Protection of Victims of Non-International Armed Conflicts (Protocol II) [1977], printed in Schindler and Toman, The Laws of Armed Conflicts (n 4) pp. 775ff., Article 6(5), encourages member States to provide amnesty to opposing forces where the Protocol applies, but they are not obliged to do so. Furthermore, this is only likely to be considered at the end of the conflict.
AP I (n 10) Article 49(1).
Schmitt, Michael N., Tallinn Manual 2.0 on the International Law Applicable to Cyber Operations [hereinafter Tallinn Manual 2.0] (2nd edition, Cambridge University Press, 2017) p. 416. See also amw Manual (n 4) Rule 1(e), Commentary para. 7.
Yoram Dinstein, The Conduct of Hostilities under the Law of International Armed Conflict (Cambridge University Press, Cambridge, 2016) p. 3.
As explained in Chapter 1, fn. 58, the U.S. DoD has made it clear that in their view, defining attack as the use of force in both offence and defence is etymologically inconsistent with its normal use, and that “the word “attack” or “attacks” historically has referred to and today refers to offensive operations only.” U.S. Department of Defense (DoD): Conduct of the Persian Gulf War, Final Report to Congress, 31(3) ILM 1992, available at
On the exceptional nature of self-defence and its limited scope of application by military forces during armed conflict, see Chapter 8.
Dinstein, Conduct of Hostilities (n 69) p. 3.
AP I (n 10) Article 49(3): “The provisions of this Section apply to any land, air or sea warfare which may affect the civilian population, individual civilians or civilian objects on land. They further apply to all attacks from the sea or from the air against objectives on land but do not otherwise affect the rules of international law applicable in armed conflict at sea or in the air.”
When the issue was discussed by a group of international experts in connection with the development of the amw Manual, they concluded that “as a general principle, the same legal regime applies equally in all domains of warfare”. AMW Manual (n 4) Commentary to Rule 30. See also Tallinn Manual 2.0 (n 68) p. 415, and Michael N. Schmitt, ‘Air Warfare’, in Andrew Clapham and Paola Gaeta (eds.), The Oxford Handbook of International Law in Armed Conflict (Oxford University Press, 2014) pp. 124–125.
AP I (n 10) Article 48 and AP II (n 66) Article 13(2).
See e.g. icrc cil Study (n 11) Rule 1 and Volume 1, Commentary Rule 1, p. 3, where this general rule is presented as applicable both during iacs and niacs. Other categories which are less relevant and hence will not be discussed here are spies and mercenaries. Mercenaries and combatants who are captured while engaged in espionage without wearing their uniform, i.e. spies, will be lawful targets but are not entitled to prisoner of war status or combatant immunity for the acts carried out in that capacity. AP I (n 10) Article 47 (mercenaries) and Article 46 (spies). For a useful article on this topic, see for instance William Baxter, ‘So-Called ‘Unprivileged Belligerency’: Spies, Guerrillas, and Saboteurs’, 28 British Yearbook of International Law 323 (1951).
Nils Melzer, ‘The Principle of Distinction between Civilians and Combatants’, in Andrew Clapham and Paola Gaeta (eds.), The Oxford Handbook of International Law in Armed Conflict (Oxford University Press, 2014) p. 296. See also Legality of the Threat or Use of Nuclear Weapons, Advisory Opinion, I.C.J. Reports [1996] 226, para 78; and Dinstein, Conduct of Hostilities (n 69) p. 41. For a useful historic overview of the principle in loac, see W. Hays Parks, ‘Part IX of the icrc “Direct Participation in Hostilities” Study: No Mandate, No Expertise, and Legally Incorrect’, 42 New York University Journal of International Law and Policy 769 (2009–2010), pp. 772–783.
Melzer, ‘The Principle of Distinction between Civilians and Combatants’, ibid, p. 299. Outside the context of targeting, the dichotomy also fails to take into account that participants in a levée en masse are recognised as combatants, but are neither members of the armed forces nor civilians. ibid, p. 306. Watkin also criticises the traditional dichotomy and argues that there are five classes of participants in AP I: “lawful combatants under article 43 of Additional Protocol I; otherwise lawful combatants who fall within article 44(4); members of organized armed groups who are not lawful combatants under article 43; civilians who take a direct part in hostilities; and uninvolved civilians.” In his view, the second, third and fourth categories should be viewed as unlawful combatants rather than civilians. Watkin, ‘Opportunity Lost’ (n 18) pp. 665–666. On ‘unlawful combatants’, see further Section 5.2.2.
Geneva Convention (I) for the Amelioration of the Condition of the Wounded and Sick in Armed Forces in the Field (GC I) [1949], Laws of Armed Conflict, printed in Schindler and Toman, The Laws of Armed Conflicts (n 4) pp. 459ff., Article 24; Geneva Convention (II) for the Amelioration of the Condition of the Wounded, Sick and Shipwrecked Members of Armed Forces at Sea (GC II) [1949], ibid, pp. 485ff., Article 36; AP I (n 10) Articles 12.1 and 15, and AP II (n 66) Article 9(1). The protection may be lost if the personnel carry out acts harmful to the enemy, cf. GC I, ibid, Article 21, and AP II, ibid, Article 11(2), which although they refer to medical units, is considered applicable by analogy to medical and religious personnel. See e.g. icrc cil Study (n 11) Rules 25 and 27; Commentary Rule 25, p. 85; and Commentary Rule 27, p. 91. See also Melzer, ‘The Principle of Distinction between Civilians and Combatants’ (n 77) pp. 300 and 309.
AP I (n 10) Article 67(1). The protection may be lost if the personnel carry out acts harmful to the enemy, cf. Article 67(1)(e).
AP I, ibid, Article 41(1–2); Common Article 3 to GC I-IV (n 79); and AP II (n 66) Articles 4(1) and 7(1).
AP I, ibid, Article 50(1). See also Prosecutor v. Blaskic (Judgement), IT-95-14-T, icty, Trial Chamber, 3 March 2000, §180: Civilians are “persons who are not, or no longer, members of the armed forces”.
AP I, ibid, Article 51(1) and AP II (n 66) Article 13(1).
AP I, ibid, Article 51(2) and AP II, ibid, Article 13(2).
AP I, ibid, Article 51(3) and AP II, ibid, Article 13(3).
1907 Hague Regulations (n 28) Articles 1 and 2. For further discussion of the definition of combatants, see Section 5.2.2.
See e.g. icrc, Interpretive Guidance on the Notion of Direct Participation in Hostilities under International Humanitarian Law [hereinafter icrc Interpretive Guidance] (May 2009, prepared by Nils Melzer, available at
For a brief distinction of conduct-based versus status-based targeting, see e.g. Eric D. Montalvo, ‘When Did Imminent Stop Meaning Immediate? Jus in bello Hostile Intent, Imminence and Self-Defense in Counterinsurgency’, Army Lawyer, August 2013, p. 24, pp. 29–31. See also Chapters 11 and 12 examining the relationship between loac and the conduct based nato hostile act and hostile intent roe and the status-based roe authorising attack on lawful targets regardless of current activity.
icrc Interpretive Guidance (n 87) pp. 25–26. See also Jody M. Prescott, ‘Tactical Implementation of Rules of Engagement in a Multinational Force Reality’, in Geoffrey S. Corn, Rachel E. VanLandingham and Shane R. Reeves (eds.), U.S. Military Operations – Law, Policy and Practice (Oxford University Press, Oxford, 2016) pp. 267–268. By contrast, see Dinstein, Conduct of Hostilities (n 69) p. 175: “a person directly participating in hostilities loses his civilian status and can be effectively assimilated to a combatant; and, for that matter, and unlawful combatant under customary international law”. See also Randall Bagwell and Molly Kovite, ‘It is not self-defence: Direct participation in hostilities authority at the tactical level’, 224(1) Military Law Review 1 (2016), p. 26, contending that “direct participation is status-based” targeting, but adds that the categorisation is “unhelpful and unnecessary”.
The U.S. military would considered such person to be “persons declared hostile”, and they would be targetable on the basis of that status. U.S. Army, Operational Law Handbook (n 70) p. 79. See also U.S. Chairman of the Joint Chiefs of Staff (cjcs), Standing Rules of Engagement (sroe)/Standing rules for the use of force (sruf) for U.S. Forces [hereinafter: sroe/sruf] (cjcs Instruction 3121.01B, 13 June 2005) p. A-2; Montalvo, ‘When Did Imminent Stop Meaning Immediate?’ (n 88) pp. 30–31, and Schmitt, ‘Targeting and IHL in Afghanistan’ (n 11) p. 314.
As will be explained below in Section 5.2.3, the picture is more nuanced than having to attack someone while in the middle of the act amounting to direct participation of hostilities, but it is simplified here to emphasise the distinction between conduct-based and status-based targeting.
Travis Normans, US v. El-Hanafi (cont), (loac blog, 19 June 2012)
For an example of the view that conduct-based targeting is regulated by self-defence, see Albert S. Janin, ‘Engaging Civilian-Belligerents Leads to Self-defense/Protocol I Marriage’, Army Lawyer (July 2007), DA PAM 27-50-410, p. 89.
See Section 8.3.2.3.
See also Melzer, ‘The Principle of Distinction between Civilians and Combatants’ (n 77) p. 299.
GC iii (n 29) Article 4(A). See also GC I (n 79) Article 13, and 1907 Hague Regulations (n 28) Articles 1 and 3.
Melzer, ‘The Principle of Distinction between Civilians and Combatants’ (n 77) p. 301–302.
AP I (n 10) Article 43(1).
ibid.
Their right to prisoner of war status upon capture is confirmed in GC iii (n 29) Article 4(A)(2) and 4(A)(6). See also 1907 Hague Regulations (n 28) Article 2.
AP I (n 10) Article 44.3 provides an exception for the requirement of distinction for situations where “owing to the nature of the hostilities an armed combatant cannot so distinguish himself”. The status as a combatant can nonetheless be retained, “provided that, in such situations, he carries his arms openly: (a) during each military engagement, and (b) during such time as he is visible to the adversary while he is engaged in a military deployment preceding the launching of an attack in which he is to participate.” This has been interpreted to only apply in extreme situations such as occupation, where complying with the principle of distinction at all times would make it impossible to carry out any operations. See e.g. Norwegian Armed Forces, Manual i krigens folkerett (Trans: Manual of the Law of Armed Conflict), issued under the authority of the Chief of the Defence, [2013], [hereinafter referred to as Norwegian loac Manual], p. 112, and UK, Reservations to the 1977 Additional Protocol 1 [hereinafter: Reservations to api], available at
AP I (n 10) Article 44.2–5.
Parks, ‘No mandate, no expertise, and legally incorrect’ (n 77) p. 778.
AP I (n 10) Article 41.2. icrc cil Study (n 11) Rule 47 and Commentary Rule 47, p. 164. This rule is equally applicable to niacs.
ibid.
AP I, ibid, Article 43(2). See also Dinstein, Conduct of Hostilities (n 69) p. 43.
See also Section 5.1.2 regarding national approaches to the concept of ‘lawful acts of war’.
GC iii (n 29) contains, inter alia, detailed rules pertaining the prisoners’ security (Article 23); quarters, food and clothing (Articles 25–28); penal and disciplinary sanctions (Articles 82–108); and release and repatriation (Article 118).
nato, STANAG 2449 (n 12) p. F-27.
The term ‘combatant’ is not found in treaties regulating niacs. See also Melzer, ‘The Principle of Distinction between Civilians and Combatants’ (n 77) p. 309.
See also ibid, p. 318.
ibid, p. 321. See for instance icrc cil Study (n 11) Rule 1, and UN General Assembly, Respect for Human Rights in Armed Conflict (9 December 1970) UN GA Res. 2676 (xxv) preamble and § 5, referring to “combatants in all armed conflicts”.
See e.g. Melzer, ‘The Principle of Distinction between Civilians and Combatants’ (n 77) p. 305 (‘privileged combatants’).
See e.g. Israel, Detention of Unlawful Combatants (Amendment and Temporary Provision) Amendment and Law, 5768-2008, (Mar. 19, 2008; July 30, 2008; Aug. 13, 2008). For an overview of the U.S. use of the term, see Knut Dörmann, ‘Unlawful combatants’, in Andrew Clapham and Paola Gaeta (eds.), The Oxford Handbook of International Law in Armed Conflict, Oxford University Press, 2014, pp. 620–622. See also Yoran Dinstein, ‘Unlawful combatancy’, in ‘International law and the war on terror’, 79 Naval War College International Law Study 151 (2003), and Kenneth Watkin, Warriors Without Rights?: Combatants, Unprivileged Belligerents, and the Struggle Over Legitimacy (Program on Humanitarian Policy and Conflict Research Occasional Paper Series, 2005) Volume 2.
See also icrc Interpretive Guidance (n 87) pp. 83–85 and Melzer, ‘The Principle of Distinction between Civilians and Combatants’ (n 77) p. 323.
See for instance Ex Parte Quirin, 371 U.S. 1 (1942), page 317 U.S. 31: “Unlawful combatants are likewise subject to capture and detention, but, in addition, they are subject to trial and punishment by military tribunals for acts which render their belligerency unlawful”. See also Boothby, The Law of Targeting (n 3) pp. 68–69. The question of appropriate terminology for this category of persons is not sufficiently relevant for the focus on this book and will therefore not be further examined. Useful articles on the topic include, Charles H.B. Garraway, ‘“Combatants” – Substance or Semantics?’, Chapter 12 in Michael N. Schmitt and Jelena Pejic (eds.), International Law and Armed Conflict: Exploring the Faultlines: Essays in Honour of Yoram Dinstein (brill, 2007); Baxter, ‘So-Called ‘Unprivileged Belligerency’ (n 76); and Melzer, ‘The Principle of Distinction between Civilians and Combatants’, ibid, p. 306.
UK loac Manual (n 16) para. 4.2.1, as amended 2010.
Boothby, The Law of Targeting (n 3) p. 66.
As mentioned in the introduction to this Chapter, there is some debate concerning the applicable rules on the treatment of detained or interned ‘unprivileged’ or ‘unlawful’ combatants which will not be examined in detail here. Because they are not entitled to the enhanced protections provided by GC iii, the question is whether they are civilians and therefore protected by GC IV, or if only the fundamental guarantees for humane treatment set out in Common Article 3 and customary international law including the protections set out in AP I (n 10) Article 75 apply. For a detailed presentation of this debate, see Jelena Pejic, ‘“Unlawful/enemy combatants”: Interpretations and Consequences’, Chapter 13 in Michael N. Schmitt and Jelena Pejic (eds.), International Law and Armed Conflict: Exploring the Faultlines: Essays in Honour of Yoram Dinstein (brill, 2007); Dörmann, Knut, ‘Unlawful combatants’ (n 114) pp. 607–618; and Knut Dörmann, ‘The legal situation of “unlawful/unprivileged combatants”’, 85 International Review of the Red Cross 45 (March 2003).
The topic has been the subject of extensive research, see e.g. Michael N. Schmitt, ‘The Status of Opposition Fighters in a Non-International Armed Conflict’, 88 International Law Studies 119 (2012); Jann K. Kleffner, ‘From ‘Belligerents’ to ‘Fighters’ and Civilians Directly Participating in Hostilities – on the Principle of Distinction in Non-international Armed Conflicts one Hundred Years after the Second Hague Peace Conference’, 54(2) Netherlands International Law Review 315 (2007); and Yoram Dinstein, Non-International Armed Conflicts in International Law [hereinafter: niac] (Cambridge University Press, Cambridge, 2014) pp. 58–60.
This expansive application is criticised by Dörmann: “While the notion of ‘unlawful combatant’ has been used by some also in the context of non-international armed conflicts, it has a place only within the context of the law applicable to international armed conflicts as defined in the GCs and AP I. The law applicable in non-international armed conflicts does not foresee a combatant’s privilege”. Dörmann, Knut, ‘Unlawful combatants’ (n 114) p. 617.
Common Article 3 to GC I-IV (n 79).
AP II (n 66) Article 1(1). For a discussion of the differences between Common Article 3 and AP II, see Melzer, ‘The Principle of Distinction between Civilians and Combatants’ (n 77) pp. 311–314.
Michael N. Schmitt, Yoram Dinstein, and Charles HB Garraway, The Manual on the Law of Non-International Armed Conflict, with Commentary (International Institute of Humanitarian Law, Martinus Nijhoff Publishers, Leiden/Boston, 2006) p. 4.
icrc cil Study (n 11) Commentary to Rule 3.
Melzer, ‘The Principle of Distinction between Civilians and Combatants’ (n 77) p. 312.
icrc Interpretive Guidance (n 87) pp. 83–85 and Melzer, ‘The Principle of Distinction between Civilians and Combatants’, ibid, pp. 323–324.
According to Bagwell and Kovite, there is significant disagreement in three areas: “(1) when direct participation begins; (2) when direct participation ends; and (3) what constitutes an act of direct participation”. Bagwell and Kovite, ‘It is not self-defence’ (n 89) p. 22.
AP I (n 10) Article 51(3) and AP II (n 66) Article 13(3).
icrc cil Study (n 11) rule 6. See also Dinstein, Conduct of Hostilities (n 69) p. 175; and Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 13. Note, however, that the application of dph is likely to be different in naval operations. See Wolff Heintschel von Heinegg, ‘The Development of the Law of Naval Warfare from the Nineteenth to the Twenty-First Century – Some Select Issues’, 17 Yearbook of International Humanitarian Law 69 (2014), p 86, where he explains that “Article 51(3) Additional Protocol I (1977) is inapplicable to “sea warfare”, which does not “affect the civilian population, individual civilians or civilian objects on land””. On the application of Article 51(3) to air warfare, see amw Manual (n 4) Section F.
See also Dinstein, NIAC (n 119) p. 61, footnote 207, where he states that the assessment of what amounts to direct participation in hostilities is the same for both forms of conflicts.
icrc Interpretive Guidance (n 87).
New York University Journal of International Law and Politics, Volume 42, Number 3, Spring 2010: Forum: ‘Direct Participation in Hostilities: Perspectives on the icrc Interpretive Guidance’. Articles: Goodman, Ryan and Derek Jinks, ‘The icrc Interpretive Guidance on the Notion of Direct Participation in Hostilities under International Humanitarian Law: An Introduction to the Forum’, p. 637; Watkin, Kenneth, ‘Opportunity Lost: Organized Armed Groups and the icrc “Direct Participation in Hostilities” Interpretive Guidance’, p. 641; Schmitt, Michael N., ‘Deconstructing Direct Participation in Hostilities: The Constitutive Elements’, p. 697; Boothby, Bill, ‘“And for such time as”: The Time Dimension to Direct Participation in Hostilities’, p. 741; Parks, W. Hays, ‘Part IX of the icrc “Direct Participation in Hostilities” Study: No mandate, no expertise, and legally incorrect’, p. 769; Melzer, Nils, ‘Keeping the Balance between Military Necessity and Humanity: A Response to Four Critiques of the icrc’s Interpretive Guidance on the Notion of Direct Participation in Hostilities’, p. 831.
Boddens Hosang, Rules of Engagement (n 6) pp. 205–206.
See also Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 43: “the constitutive elements of direct participation, although not bereft of flaws, represent a useful step forward in understanding the notion”.
The notion of ‘direct participation in hostilities’ is also relevant in the context of civilian contractors in the battlefield and the discussions regarding the tasks they should be permitted to carry out. This is, however, less relevant here where the focus is on who nato forces may be authorised to use force against. For a presentation of challenges concerning the use of civilian contractors, see Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) pp. 7–10; and W. Hays Parks, Evolution of Policy and Law Concerning the Role of Civilians and Civilian Contractors Accompanying the Armed Forces (Paper presented at the Third Expert Meeting on the Notion of Direct Participation in Hostilities Geneva, October 2005,
Schmitt, ‘Targeting and ihl in Afghanistan’ (n 11) p. 315. The lit standard was intended to reduce the risk of collateral damage and accidental attacks on friendly forces, and was therefore so restrictive is was referred to as “self-defence plus”. ibid, citing e-mail to author from a “key participant in its development at US Central Command (US centcom)”.
See also Bagwell and Kovite: “direct participation authority is firmly rooted in international law and provides substantial benefits as a bridging authority between self-defense and declared hostile force authority”. Bagwell and Kovite, ‘It is not self-defence’ (n 89) p. 6, voicing frustration over the fact that despite its central role, this authority “is currently ignored in U.S. use of force policy at the tactical level”. ibid.
ibid, p. 8.
The phrase ‘active part in hostilities’ may be found in Common Article 3 to GC I-IV (n 79). See also icrc Interpretive Guidance (n 87) pp. 43–44, and Dinstein, Conduct of Hostilities (n 69) pp. 174–175. In the case Prosecutor v. Lubanga, Case No. ICC-01/04-01/06, icc, Trial Chamber, 2012, it was suggested that ‘to participate actively in hostilities’ should be interpreted wider than ‘direct participation in hostilities’ (51 ilm 1021 at p. 1068), however, this is criticised by Dinstein who contends that “there is not even slender evidence in practice for such an assertion”. Dinstein, niac (n 119) pp. 59–60. See also Boothby, The Law of Targeting (n 3) pp. 144–145.
Melzer, ‘The Principle of Distinction between Civilians and Combatants’ (n 77) p. 325.
ibid, p. 325, citing the following articles as references: 1907 Hague Regulations (n 28) Article 22; and AP I (n 10) Articles 35(1), 43(2), 45(1) and (3), and 51(3).
ibid. See also Sandoz et al. (eds.), AP Commentary (n 4) para. 1944: “‘direct’ participation means acts of war which by their nature or purpose are likely to cause actual harm to the personnel and equipment of the enemy armed forces”.
icrc Interpretive Guidance (n 87) p. 46.
ibid, p. 47.
ibid.
ibid, pp. 47 and 49.
ibid, p. 47.
ibid, p. 48.
ibid, p. 49.
ibid, p. 50.
ibid, p. 51.
ibid. See also the Commentary to Article 51(3), “There should be a clear distinction between direct participation in hostilities and participation in the war effort. The latter is often required from the population as a whole to various degrees. Without such a distinction the efforts made to reaffirm and develop international humanitarian law could become meaningless. In fact, in modern conflicts, many activities of the nation contribute to the conduct of hostilities, directly or indirectly; even the morale of the population plays a role in this context.” Sandoz et al. (eds.), AP Commentary (n 4) 1945.
icrc Interpretive Guidance (n 87) p. 53.
ibid.
ibid, p. 54.
ibid, p. 53.
ibid, pp. 54–55.
ibid, p. 55.
ibid, p. 58.
ibid, pp. 59 and 60–61.
ibid, p. 61. See also Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 34.
icrc Interpretive Guidance (n 87) p. 59.
ibid, p. 64.
Dinstein, niac (n 119) p. 61, citing Prosecutor v. Dragomir Milosevic (Judgement), Case No. IT-98-29/1-T, icty, Trial Chamber, 12 December 2007, para. 947.
Dinstein, Conduct of Hostilities (n 69) p. 179.
ibid.
As Hans Boddens Hosang points out, it is unlikely that the icrc rules for determining direct participation in hostilities “can easily be translated into roe that would still meet the criteria of simplicity and ease of understanding”. Boddens Hosang, Rules of Engagement (n 6) p. 193. See also Prescott, ‘Tactical Implementation of roe’ (n 89) p. 252.
Boddens Hosang, Rules of engagement, ibid, p. 194. Other areas that have been criticised include the requirement that the direct participation in hostilities must impose harm on one side, thereby excluding acts that strengthens the enemy’s capacity without causing direct harm. See Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) pp. 27–28.
Michael N. Schmitt and Eric Widmar, ‘The Law of Targeting’, in Paul AL Ducheine, Michael N. Schmitt; and Frans Osinga (eds.), Targeting: The Challenges of Modern Warfare (Asser Press, 2016) p. 127.
Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 30.
ibid, p. 30. Boothby, however, criticises the “integral part” requirement for “inappropriately restricting the scope of dph”. Bill Boothby, ‘“And for such time as”’ (n 132) pp. 751–752.
Schmitt and Widmar, ‘The Law of Targeting’ (n 169) p. 127. See also Bagwell and Kovite, ‘It is not self-defence’ (n 89) p. 28.
Prosecutor v. Tadic (Opinion and Judgement), Case No. IT-94-1-T, icty, Trial Chamber, 7 May 1997, para. 616.
U.S. DoD, Law of War Manual (n 24) p. 229, footnotes omitted.
For a useful and extensive list of considerations that may be relevant to the determination of direct participation in hostilities, see ibid, pp. 230–231. The Manual also provide several examples of acts considered to amount to direct participation in hostilities and acts that are not. See ibid, pp. 231–233. The U.S. differs from the icrc proposal with regard to “acts of the civilian population [that] are not part of the immediate actual fighting, but instead perform combat support functions that may be temporally or geographically remote from the fighting.” Bagwell and Kovite, ‘It is not self-defence’ (n 89) p. 31.
icrc Interpretive Guidance (n 87) p. 48.
ibid.
ibid.
ibid, p. 49, footnote omitted.
ibid, p. 54.
ibid, p. 56.
Boddens Hosang, Rules of Engagement (n 6) p. 196.
ibid, p. 201.
icrc Interpretive Guidance (n 87) p. 56, footnote 141. See also Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 31.
icrc Interpretive Guidance, ibid, pp. 56–57.
Boddens Hosang, Rules of Engagement (n 6) p. 197.
ibid. See also Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 32, and Dinstein, Conduct of Hostilities (n 69) p. 184.
Schmitt and Widmar, ‘The Law of Targeting’ (n 169) p. 127.
Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 33. The presence of civilians near military objectives would could also cause the required approval level for carrying out the attack to be elevated, if the policy of zero civilian casualties is in force. Schmitt, ‘Targeting and ihl in Afghanistan’ (n 11) p. 322. For a different solution to this difficult question, see Boothby, The Law of Targeting (n 3) p. 139, where he argues that voluntary human shields should be regarded as protected and therefore count in the proportionality analysis, but that their weight is reduced due to their acceptance of the risk involved. This solution introduces the idea that civilians can have varying weight or importance in the proportionality analysis, which seems to be worryingly open to abuse.
Boddens Hosang, Rules of Engagement (n 6) p. 196.
Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 30. See also Boothby, ‘“And for such time as”’ (n 132 p. 749).
See also Boddens Hosang, Rules of Engagement (n 6) p. 196: “practice has shown that the geographical and temporal proximity between their manufacture and their actual use is comparable to the delivery of ammunition to troops fighting on the frontline”.
Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 30. See also Watkin, ‘Opportunity Lost’ (n 18) pp. 658–659.
Watkin, Kenneth, ‘Opportunity Lost’, ibid, p. 569. See also Henderson, The Contemporary Law of Targeting (n 6) p. 102, where he argues that tactical intelligence gathering amounts to dph, whilst strategic intelligence gathering does not.
Watkin, Kenneth, ‘Opportunity Lost’, ibid, p. 660.
ibid, p. 690, footnote omitted.
ibid., p. 660.
Schmitt and Widmar, ‘The Law of Targeting’ (n 169) p. 128. See also The Public Committee against Torture in Israel v. The Government of Israel, Israel hcj 769/02, Judgment of 13 December 2006, Israeli Supreme Court 57(6) 285, para. 38; and Prosecutor v. Blaskic (Judgement), Case No. IT-95-14-A, icty, Appeals Chamber, 29 July 2004, para. 157. Other controversial parts of the icrc Interpretive Guidance are the requirement of a ‘continuous combat function’ within an organised armed group and its Section IX on ‘Restraints on the use of force in direct attack’, where it is proposed that there may sometimes be a duty to capture rather than kill. These issues are examined further in Sections 5.2.3.4 and 5.3.
icrc Interpretive Guidance (n 87) p. 65.
ibid.
ibid.
ibid.
ibid.
ibid, p. 67.
Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) pp. 36–37. See also Boothby, ‘“And for such time as”’ (n 132) criticising the icrc approach.
Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) pp. 36–37. See also Dinstein, Conduct of Hostilities (n 69) p. 177.
Boothby, ‘“And for such time as”’ (n 132) p. 750.
ibid, pp. 751 and 765.
See also Boddens Hosang, Rules of Engagement (n 6) pp. 202–203.
ibid, p. 204.
See e.g. Parks, ‘Air War’ (n 63) p. 119. See also Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 37, and Watkin, ‘Opportunity Lost’ (n 18) pp. 686–687.
icrc Interpretive Guidance (n 87) p. 70.
Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 38. See also Boothby, ‘“And for such time as”’ (n 132) pp. 756–757, and Bagwell and Kovite, ‘It is not self-defence’ (n 89) p. 26.
Dinstein, Conduct of Hostilities (n 69) pp. 177–178. See also Boddens Hosang, Rules of Engagement (n 6) p. 203.
See further Section 5.2.3.4.
Watkin, ‘Opportunity Lost’ (n 18) p. 686.
icrc Interpretive Guidance (n 87) p. 71, fn. 192.
Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 38.
Boddens Hosang, Rules of Engagement (n 6) p. 204. See also Boothby, ‘“And for such time as”’ (n 132) p. 768.
Sandoz et al. (eds.), AP Commentary (n 4) para. 4789.
Boothby, ‘“And for such time as”’ (n 132) p. 756. Boothby refers to the AP I Commentary, but the quote is taken from the commentary to Article 13.3 of AP II.
Bagwell and Kovite, ‘It is not self-defence’ (n 89) p. 32. By contrast, the question when direct participation ends is “significantly more problematic for implementation at the tactical level”, although it is not made clear whether the authors considers the challenge to be more significant at the tactical level than other levels of operation. ibid, p. 32.
ibid, p. 35.
Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 38. See, however, Henderson, The Contemporary Law of Targeting (n 6) p. 96, where he argues that Schmitt’s view does not reflect current law. See also Robin Geiss and Michael Siegrist, ‘Has the armed conflict in Afghanistan affected the rules on the conduct of hostilities?’, 93(881) International Review of the Red Cross 11 (March 2011), pp. 22–26.
See also David Kretzmer, ‘Targeted Killing of Suspected Terrorists’, 16(2) European Journal of International Law 171 (2005), p. 199.
icrc Interpretive Guidance (n 87) p. 74. As will be discussed below in Section 5.4, the U.S. disagrees with this being customary law.
Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 39. See also Boothby, ‘“And for such time as”’ (n 132) pp. 759–761.
icrc Interpretive Guidance (n 87) p. 22.
Kretzmer, ‘Targeted Killing of Suspected Terrorists’ (n 225) p. 198.
icrc Interpretive Guidance (n 87) p. p. 33.
ibid, p. 71. Rather than not being applicable, it is perhaps more accurate to state that “insofar as they are members of a group that exists for the very purpose of engaging in hostilities, the ‘for such time’ criterion must be interpreted as extending throughout the duration of their membership”. Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) pp. 16–17.
Schmitt, ‘The Interpretive Guidance: A Critical Analysis’, ibid, pp. 21–22.
icrc Interpretive Guidance (n 87) p. 31.
ibid, p. 32.
ibid, p. 33.
ibid, p. 35.
ibid, p. 34.
ibid, p. 32 (re. niacs). See also ibid, pp. 23–24 (re. iacs).
See also Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) pp. 17–20.
Francoise J. Hampson, ‘Direct Participation in Hostilities and the Interoperability of the Law of Armed Conflict and Human Rights Law’, 87 International Law Studies 187 (2011), pp. 199–200.
Boddens Hosang goes as far as arguing that “while the icrc’s Guidance contains a number of valuable and valid contributions to the overall understanding of the notion of dph, its theoretical and at times idealistic approach may render many elements and subtleties of the Guidance nearly impossible to implement in the actual practice of modern military operations”. Boddens Hosang, Rules of Engagement (n 6) p. 206.
See also Wolff Heintschel von Heinegg and Peter Dreist, ‘The 2009 Kunduz Air Attack: The Decision of the Federal Prosecutor-General on the Dismissal of Criminal Proceedings Against Members of the German Armed Forces’, 53 German Yearbook of International Law 833 (2010), p. 858, referring to it as a “double standard (...) that would not be operable in real life”.
Watkin, ‘Opportunity Lost’ (n 18) p. 675.
icrc Interpretive Guidance (n 87) p. 34.
Watkin, ‘Opportunity Lost’ (n 18) p. 675.
ibid, p. 680. See also Dinstein, niac (n 119) pp. 58–59.
See also Boothby, ‘“And for such time as”’ (n 132) p. 758, footnote omitted. This argument is further developed below.
Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 24.
See e.g. ibid. See also Watkin, ‘Opportunity Lost’ (n 18) pp. 690–691.
See also U.S. DoD, Law of War Manual (n 24) pp. 100–101.
The cook is commonly used as an example of a function that will not be considered a lawful target if direct participation in hostilities is a requirement, but who may be considered a combatant and lawful target if a member of the State armed forces. See e.g. Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 23, and Watkin, ‘Opportunity Lost’ (n 18) p. 676.
On the loac standards for identifying a lawful target and precautions in attack in general, see Section 5.4.
See also Dinstein, niac (n 119) p. 63, where he argues that a person who dph in a recurrent manner may be considered to dph on a continuous basis. See also Dinstein, Conduct of Hostilities (n 69) p. 177.
Boothby, ‘“And for such time as”’ (n 132) p. 758, footnote omitted. See also the AP Commentary, which refers to “a civilian who takes part in armed combat, either individually or as part of a group, thereby becomes a legitimate target”, suggesting that it is possible to view members of organised armed groups as civilian directly participating in hostilities. Sandoz et al. (eds.), AP Commentary (n 4) para. 1942.
Watkin criticises the icrc for creating a “third category with criteria for membership that are unique and are not found in existing treaty or customary law”. Watkin, ‘Opportunity Lost’ (n 18) p. 643.
icrc Interpretive Guidance (n 87) p. 45.
See Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 23, where he argues that: “it makes little sense to justify the continuous combat function criterion on the basis of concern about an inability to distinguish members of an organized armed group from civilians or civilian affiliates of the armed group, as ihl already deals with doubt through a presumption of civilian status”. On the issue of doubt, see further Section 5.4 below.
Boothby, The Law of Targeting (n 3) p. 152.
Watkin, ‘Opportunity Lost’ (n 18) p. 688. See also Dinstein, NIAC (n 119) p. 62.
icrc Interpretive Guidance (n 87) p. 35.
Prescott, ‘Tactical Implementation of ROE’ (n 89) p. 252.
ibid.
Boothby, ‘“And for such time as”’ (n 132) p. 766.
ibid.
Parks, ‘Air War’ (n 63) p. 118. See also Melzer, ‘The Principle of Distinction between Civilians and Combatants’ (n 77) p. 312.
UK loac Manual (n 16) para. 5.3.3.
icrc Interpretive Guidance (n 87) p. 72.
Watkin, ‘Opportunity Lost’ (n 18) p. 693.
See also Bagwell and Kovite, ‘It is not self-defence’ (n 89) p. 36. In Bagwell’s experience from Afghanistan, “To account for the fact that a person can indicate he is no longer a direct participant by not committing any acts of direct participation over a period of time, the person’s inclusion on the list [of verified targets] was not permanent. United States forces were required to refresh his status with new intelligence evidencing a person’s continued direct participation within a given timeframe”. (Footnote omitted).
Jean S. Pictet (ed.), The Geneva Conventions of 12 August 1949: Geneva Convention relative to the protection of civilian persons in time of war, Vol. 4 (icrc, Geneva, 1958) Article 5, p. 53.
Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 17, referring to differing views expressed in the development of the icrc Interpretive Guidance.
Dinstein, Conduct of Hostilities (n 69) p. 42.
See also Corn, ‘Public Authority to Use Force in Military Operations’ (n 35) p. 42.
AP I (n 10) Article 35(2).
icrc Interpretive Guidance (n 87) p. 82.
ibid. The icrc position is influenced by the decision of the Israeli High Court of Justice in The Public Committee against Torture in Israel v. The Government of Israel (n 198) para. 40. The reliance on and application of this decision outside its unique context is criticised, inter alia, by Hays Parks, see Parks, ‘No mandate, no expertise, and legally incorrect’ (n 77) pp. 788–793.
This is set out clearly in both Article 22 of the 1907 Hague Regulations (n 28) and AP I (n 10) Article 35(1).
Dinstein, Conduct of Hostilities (n 69) p. 42. See also Dinstein, NIAC (n 119) p. 59. The notion is criticised by several international experts, see e.g. Parks, ‘No mandate, no expertise, and legally incorrect’ (n 77); Geoffrey S. Corn, Laurie R. Blank, Chris Jenks, and Eric Talbot Jensen, ‘Belligerent Targeting and the Invalidity of a Least Harmful Means Rule’, 89 International Law Studies 536 (2013); Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) pp. 41–43; Kleffner, ‘Section IX of the ICRC Interpretive Guidance on Direct Participation in Hostilities’ (n 6); and Boddens Hosang, Rules of Engagement (n 6) pp. 206–214.
Corn et al., ‘Belligerent Targeting and the Invalidity of a Least Harmful Means Rule’, ibid, pp. 540–541 and pp. 613–618.
Boddens Hosang, Rules of Engagement (n 6) p. 209. See also Section 5.1.1, note 13 and accompanying text.
Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 41.
Kleffner, ‘Section IX of the ICRC Interpretive Guidance on Direct Participation in Hostilities’ (n 6) p. 41.
See also Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) p. 43, and Michael N. Schmitt, ‘Wound, Capture, or Kill: A Reply to Ryan Goodman’s ‘The Power to Kill or Capture Enemy Combatants’, 24(3) European Journal of International Law 855 (2013), p. 861. However, in internal niacs, where the State’s human rights obligations will apply in full, subject to derogation, States may also consider such limitations to arise from their human rights obligations, in particular where the members of an organised armed group “pose no immediate danger and it might be feasible to apprehend them and place them on trial.” Kretzmer, ‘Targeted Killing of Suspected Terrorists’ (n 225) p. 202.
See also discussion in Section 2.4.1.
Boddens Hosang, Rules of Engagement (n 6) p. 209.
Ryan Goodman, ‘The Power to Kill or Capture Enemy Combatants’, 24(3) European Journal of International Law 819 (2013), pp. 830–836. According to Hessbruegge, there is increasing support for this view. Jan Arnold Hessbruegge, Human Rights and Personal Self-Defense in International Law (Oxford University Press, New York, 2017) pp. 220–221. See also Kleffner, ‘Section IX of the icrc Interpretive Guidance on Direct Participation in Hostilities’ (n 6) p. 36.
Schmitt, ‘Wound, Capture, or Kill’ (n 283) p. 861. See also Goodman, ‘The Power to Kill or Capture Enemy Combatants’, ibid, and Ryan Goodman, ‘The Power to Kill or Capture Enemy Combatants: A Rejoinder to Michael N. Schmitt’, 24(3) European Journal of International Law 863 (2013).
AP I (n 10) Article 35(2).
This double effect of being a lawful target is confirmed by the AP Commentary on direct participation in hostilities: “Once he ceases to participate, the civilian regains his right to the protection under this Section, i.e., against the effects of hostilities, and he may no longer be attacked”. Sandoz et al. (eds.), AP Commentary (n 4) para. 1944. See also Schmitt, ‘The Interpretive Guidance: A Critical Analysis’ (n 5) pp. 13–14; and International Law Association Study Group on the Conduct of Hostilities in the 21st Century (ila), ‘The Conduct of Hostilities and International Humanitarian Law: Challenges of 21st Century Warfare’, 93 International Law Studies 322 (2017), p. 357.
AP I (n 10) Article 57(2)(b): “an attack shall be cancelled or suspended if it becomes apparent that the objective is not a military one or is subject to special protection or that the attack may be expected to cause incidental loss of civilian life, injury to civilians, damage to civilian objects, or a combination thereof, which would be excessive in relation to the concrete and direct military advantage anticipated”. See further Section 5.4.
See also the discussion above in Section 5.2.3.2., and the comment on this issue by Schmitt and Widmar in Schmitt and Widmar, ‘The Law of Targeting’ (n 169) p. 127.
For instance, the UK position is that “even where human shields are being used, the proportionality rule must be considered.” UK loac Manual (n 16) para. 2.7.2.
AP I (n 10) Article 57(1), considered to reflect customary law applicable in both iacs and niacs. See icrc cil Study (n 11) Rule 15.
UK loac Manual (n 16) para. 5.32, fn. 187.
ila, ‘Challenges of 21st Century Warfare’ (n 289) pp. 380–381. Note, however, that although Article 57(2)-(5) only relate to the attack aspect of military operations, the distinction between the wider general obligation and the more narrow focus on attack is not always made. See e.g. Schmitt and Widmar, ‘The Law of Targeting’ (n 169) p. 137; Sandoz et al. (eds.), AP Commentary (n 4) para. 2191, and Boothby, The Law of Targeting (n 3) p. 119.
AP I (n 10) Article 58, and icrc cil Study (n 11) Rules 22–24. See also amw Manual (n 4) Section H. Dinstein makes the distinction between active and passive precautions, see Dinstein, Conduct of Hostilities (n 69) pp. 164–174, and amw Manual, ibid, Section G, Commentary para. 1. See also U.S. DoD, Law of War Manual (n 24) pp. 271–276, and Boothby, The Law of Targeting (n 3) pp. 74–76.
UK, Reservations to AP I (n 101) §b. Similar or identical declarations or statements were made by several States, e.g. Canada, Germany, Netherlands, Algeria, Austria, Italy, Belgium, Ireland, and Spain. All reservations and declarations are available at
icty, otp Report (n 9) para. 29.
U.S. DoD, Law of War Manual (n 24) p. 193, and Schmitt and Widmar, ‘The Law of Targeting’ (n 169) p. 137.
See e.g. U.S. DoD, Law of war manual, ibid, p. 193.
ila, ‘Challenges of 21st Century Warfare’(n 289) pp. 377–378.
Schmitt and Widmar, ‘The Law of Targeting’ (n 169) p. 137. See also amw Manual (n 4) Definition (q) and Commentary, pp. 28–29, and Merriam, ‘Affirmative Target Identification’ (n 16) pp. 118–123. On the relationship between feasible and reasonable, see ila, ‘Challenges of 21st Century Warfare’, ibid, p. 375.
Sandoz et al. (eds.), AP Commentary (n 4) para. 2198. See also amw Manual, ibid, Commentary to Definition (q), para.6, p. 39, and U.S. DoD, Law of War Manual (n 24) pp. 195–197.
See e.g. Merriam, ‘Affirmative Target Identification’ (n 16) pp. 109–111. See also Parks, ‘Air war’ (n 63) p. 3, and Brian J. Bill, ‘The Rendulic ‘Rule’: Military Necessity, Commander’s Knowledge, and Methods of Warfare’, 12 Yearbook of International Humanitarian Law 136 (2009).
United States vs. List et al. (‘The Hostages Trial’) (Nuremberg, 1948) 11 nmt 1230, p. 1296. The focus of the case was on military necessity rather than distinction, but its notion that battlefield acts must be evaluated based on honesty judgement and objective reasonableness, based on the information available at the time, has been applied to all battlefield acts. See Merriam, ‘Affirmative Target Identification’ (n 16) p. 109. For a current example of a similar formulation, see U.S. DoD, Law of War Manual (n 24) p. 194: “In assessing whether the obligation to take feasible precautions has been satisfied after the fact, it will be important to assess the situation that the commander confronted at the time of the decision and not to rely on hindsight”.
United States vs. List et al (‘The Hostages Trial’), ibid, p. 1297.
AP I (n 10) Article 57(2)(a)(i). See also icrc cil Study (n 11) Rule 16. Article 57(2)(a)(i) also refers to objects, but this will not be examined further here.
AP I, ibid, Article 57(2)(b). See also icrc cil Study, ibid, Rule 19.
ila, ‘Challenges of 21st Century Warfare’ (n 289) p. 382.
AP I (n 10) Article 50(1). According to the Commentary to AP I, the rule “concerns persons who have not committed hostile acts, but whose status seems doubtful because of the circumstances. They should be considered to be civilians until further information is available, and should therefore not be attacked.” Sandoz et al. (eds.), AP Commentary (n 4) para. 1920. See also icrc Interpretive Guidance (n 87) p. 35, where it is emphasised that the determination whether someone is a lawful target or not is “subject to all feasible precautions and to the presumption of protection in case of doubt”.
icrc Interpretive Guidance, ibid, pp. 75–76. Note that this presumption was met with some criticism during the expert meetings. Boothby points out that there is no such rule in loac, only a presumption of civilian status. Boothby, The Law of Targeting (n 3) p. 149. However, the presumption of civilian status only makes sense if it refers to protected civilians, that is, civilians who are not directly participating in hostilities.
Melzer, ‘The Principle of Distinction between Civilians and Combatants’ (n 77) pp. 297–298. See also Sandoz et al. (eds.), AP Commentary (n 4) para. 1921, and Laurie R. Blank, ‘Extending Positive Identification from Persons to Places: Terrorism, Armed Conflict, and the Identification of Military Objectives’, 2013 Utah Law Review 1227 (2013), pp. 1237–1239.
See discussion above in Section 5.2.3. See also Corn, ‘Public Authority to Use Force in Military Operations’ (n 35) p. 44.
Schmitt and Widmar, ‘The Law of Targeting’ (n 169) pp. 128–129. See also Henderson, The Contemporary Law of Targeting (n 6) p. 163.
See e.g. amw Manual (n 4) Rule 12(a), Commentary paras. 4 and 5; Tallinn Manual 2.0 (n 68) p. 424; and Boothby, The Law of Targeting (n 3) p. 121.
UK loac Manual (n 16) para. 5.3.4. See also UK, Reservations to AP I (n 101) §h.
Norwegian loac Manual (n 101) p. 29. The requirement in Norwegian is “overveiende sannsynlig”.
U.S. DoD, Law of War Manual (n 24) p. 200.
ibid, p. 201.
ibid.
Sandoz et al. (eds.), AP Commentary (n 4) para. 2195.
Henderson, The Contemporary Law of Targeting (n 6) p. 165.
Merriam, ‘Affirmative Target Identification’ (n 16) p. 126. See also Boothby, The Law of Targeting (n 3) p. 122. Despite the low threshold for the degree of doubt accepted in AP I, the icrc Commentary recognises that the test for doing every feasible to verify the nature of the target is one of “common sense and good faith”. Sandoz et al. (eds.), AP Commentary (n 4) para. 2198.
Merriam, ‘Affirmative Target Identification’ (n 16) p. 124.
Schmitt and Widmar, ‘The Law of Targeting’ (n 169) p. 138.
Boothby, The Law of Targeting (n 3) p. 172. See also Merriam, ‘Affirmative Target Identification’ (n 16) pp. 124–125.
Merriam, ‘Affirmative Target Identification’, ibid, pp. 127–128, citing Prosecutor v. Galic (Judgement and Opinion), Case No. IT-98-25-T, icty, Trial Chamber, 5 December 2003, para. 55. See also amw Manual (n 4) Rule 12(a), Commentary para 4: “The degree of doubt necessary to preclude an attack is that which would cause a reasonable attacker in the same or similar circumstances to abstain from ordering or executing an attack.”
Merriam, ‘Affirmative Target Identification’, ibid, p. 127.
See also Johansen, On military necessity (n 8) pp. 134–135.
See also amw Manual (n 4) Rules 1(q), Commentary para. 5.
UK loac Manual (n 16) para. 5.32.2.
See also Schmitt, ‘Targeting and IHL in Afghanistan’ (n 11) p. 313.
Norwegian loac Manual (n 101) p. 29. See also Merriam, ‘Affirmative Target Identification’ (n 16) pp. 129–130.
Merriam, ‘Affirmative Target Identification’, ibid, p. 130, and Henderson, The Contemporary Law of Targeting (n 6) p. 164.
Johansen, On military necessity (n 8) p. 80.
AP I (n 10) Article 57(2)(a).
Tallinn Manual 2.0 (n 68) pp. 478–479. See also Merriam, ‘Affirmative Target Identification’ (n 16) pp. 103–104.
AP I (n 10) Article 57(2)(b). See also See also icrc cil Study (n 11) Rule 19.
Tallinn Manual 2.0 (n 68) p. 478.
ibid, p. 484.
On manifestly unlawful orders, see also Section 8.2.10.3.
See also Merriam, ‘Affirmative Target Identification’ (n 16) p. 131.
Andenæs, ‘Okkupasjonstidens «likvidasjoner» i rettslig belysning’ (n 64) p. 15, own translation.
AP I (n 10) Article 57(2)(b).
Boothby, The Law of Targeting (n 3) p. 176.
See UK, Reservations to AP I (n 101) §o. See also Boothby, The Law of Targeting, ibid, pp. 119–120, discussing the level of responsibility in the context of AP I Article 57(2) in general.
See also Merriam, ‘Affirmative Target Identification’ (n 16) p. 85: “[Positive identification] is, in effect, the phrase by which the U.S. military ‘operationalizes’ the law of armed conflict (loac) principle of distinction”.
On the background of the concept, see ibid, pp. 132–134; Schmitt, ‘Targeting and IHL in Afghanistan’ (n 11) p. 316; and U.S. Army, Legal Lessons Learned from Afghanistan and Iraq, Volume 1 (U.S. Army Judge Advocate General’s Legal Center and School, Center for Law and Military Operations, Charlottesville, Virginia, 2004,
Merriam, ‘Affirmative Target Identification’, ibid, p. 134, and Schmitt, ‘Targeting and IHL in Afghanistan’, ibid, p. 316.
Schmitt, ‘Targeting and IHL in Afghanistan’, ibid, p. 319.
U.S. Army, Legal Lessons Learned from Afghanistan and Iraq, Vol. 1 (n 350) pp. 96–98. For a criticism of the concept, see Merriam, ‘Affirmative Target Identification’ (n 16) especially pp. 132–140. He suggests that the better concept would be one of “Affirmative Target Identification”, which would require “an honest and reasonable belief – based on such affirmative evidence as is reasonable available at the time – that the object of attack is a lawful military target”. ibid, p. 85, and pp. 140–145. However, rather than defining pid to require too high a standard and then criticise it on that basis, it appears more useful to recognise its potential utility and work to ensure it is interpreted to set the correct standard.
See also Aaron Pennekamp, ‘Standards of Engagement: Rethinking Rules of Engagement to More Effectively Fight Counterinsurgency Campaigns’, 101 The Georgetown Law Journal 1619 (2013), p. 1641. For a similar view, see Montalvo, ‘When Did Imminent Stop Meaning Immediate?’ (n 88) p. 31. He explains that “the presence of a hostile act and/or hostile intent satisfies the military necessity principle of loac, and the requirement to have pid of the enemy satisfies the distinction principle under loac”. He goes on to explain that in short, pid is about recognising the military necessity to use force (in the form of hostile act and/or hostile intent). Note that he uses the terms ‘hostile act’ and ‘hostile intent’ as reference to conduct-based targeting.
U.S. Army, Legal Lessons Learned from Afghanistan and Iraq, Vol. 1 (n 350) p. 98, referring both to status-based and conduct-based use of force.
U.S. cjcs, No-Strike and the Collateral Damage Estimation Methodology (cjcsi 3160.01A, 13 February 2009,
Burkart, ‘Deadly Advice: Judge Advocates and Joint Targeting’, ibid, p. 16: “target validation determines whether a potential target fits within the commander’s targeting guidance and complies with the loac and roe”. Target validation may also include additional considerations, such as the effect of the attack on the operation, whether the target is politically or culturally ‘sensitive’, and the risk and likely consequences of collateral damage. See Schmitt, ‘Targeting and ihl in Afghanistan’ (n 11) p. 310, citing U.S. Air Force Doctrine (afdd 2-1.9 (2006), Targeting, p. 35).
Others view ‘positive identification’ as relating to the first aspect only, that is, the identification of someone as a lawful target. See e.g. Johansen, On military necessity (n 8) p. 327, where she argues that “[t]he “feasible to verify-rule” is often applied by the military as “positive identification”, so-called pid. The jargon is a linguistic pleonasm and as such not ideal to describe a rule”. (Footnote omitted).
Burkart, ‘Deadly Advice: Judge Advocates and Joint Targeting’ (n 356) p. 16: “Rules of engagement will set forth standards for the quantity, quality, timeliness, and duration of the intelligence necessary to establish positive identification”.
This may, however, also be viewed as a separate roe issue, rather than forming part of pid.
Merriam voices concern that if pid is intended to reflect not only law but also policy, continued application of that higher standard may make it difficult to retreat from. Merriam, ‘Affirmative Target Identification’ (n 16) pp. 137–139. Although this is a valid argument, it is not a challenge that is unique to pid. In fact, it is arguably equally valid for other ‘use of force’ concepts, such as the zero civilian casualty policy. However, this is not a reason to abandon the operational concepts or tools, but rather to better educate Commanders, military forces, and their legal advisers in the meaning of such concepts and tools.
See also Montalvo, ‘When Did Imminent Stop Meaning Immediate?’ (n 88) p. 33.
Targeting is further examined in Sections 3.1 and 12.2.
Pennekamp, ‘Standards of Engagement’ (n 354) p. 1641.
Burkart, ‘Deadly Advice: Judge Advocates and Joint Targeting’ (n 356) p. 19.
See Section 3.1.
On cde, see further Section 12.2.
AP I (n 10) Article 35(1). This principle is also reflected in more historic loac texts, such as the 1861 Lieber Code (n 63) Article 14 (“Military necessity, as understood by modern civilized nations, consists in the necessity of those measures which are indispensable for securing the ends of the war, and which are lawful according to the modern law and usages of war”); implicitly in the St. Petersburg Declaration 1868 (n 4) (“That the only legitimate object which States should endeavour to accomplish during war is to weaken the military forces of the enemy”); and the 1907 Hague Regulations (n 28) Article 22 (“The right of belligerents to adopt means of injuring the enemy is not unlimited”).
AP I (n 10), Article 35(2)-(3).
ibid, Article 57(2)(a)(ii-iii).
ibid, Article 51(5)(b). In addition to the general prohibitions on indiscriminate attacks and attacks causing unnecessary suffering or superfluous injury or which are intended, or may be expected, to cause widespread, long-term and severe damage to the natural environment, there are specific treaty restrictions, such as the 1980 Convention on Prohibitions or Restrictions on the use of Certain Conventional Weapons which may be deemed to be Excessively injurious or to have Indiscriminate Effects [1980], printed in Schindler and Toman, The Laws of Armed Conflicts (n 4) pp. 184ff, with various Protocols (see e.g. note 298), applicable to States party.
AP I, ibid, Article 57(2)(b).
ibid, Article 57(2)(c).
ibid, Article 57(3).
Schmitt and Widmar, ‘The Law of Targeting’ (n 169) p. 140: “Proportionality does not require a strict mathematical comparison, nor does it, as is often mistakenly believed, call for a balancing test”. See also Laurent Gisel, ‘Can the incidental killing of military doctors never be excessive?’, 95(889) International Review of the Red Cross 215 (2013), p. 218.
AP I (n 10) Article 57(2)(a)(iii).
Schmitt and Widmar, ‘The Law of Targeting’ (n 169) p. 140. See also amw Manual (n 4) Rule 1(l), Commentary para. 2. Although the focus in this book is on persons, it should be pointed out that objects are to some extent considered to weigh differently than persons in the proportionality count. See for instance U.S. DoD, Law of War Manual (n 24) p. 261, para. 5.12.1.1.: “the expected loss of civilian life and injury to civilians should be given greater consideration than the expected damage to civilian objects”.
For a useful overview of instances that may cause collateral damage, see Dinstein, Conduct of Hostilities (n 69) pp. 150–152.
Schmitt and Widmar, ‘The Law of Targeting’ (n 169) p. 138.
ibid. See also Dinstein, Conduct of Hostilities (n 69) pp. 169–170, and ila, ‘Challenges of 21st Century Warfare’ (n 289) pp. 383–384.
Boothby, The Law of Targeting (n 3) p. 93 and pp. 124–125.
For a similar example, see Schmitt and Widmar, ‘The Law of Targeting’ (n 169) p. 139.
Dinstein, Conduct of Hostilities (n 69) p. 155.
ibid, pp. 156–157, criticising the position taken in the Commentary to AP I: “The Protocol does not provide any justification for attacks which cause extensive civilian losses and damages. Incidental losses and damages should never be extensive.” Sandoz et al. (eds.), AP Commentary (n 4) para. 1980.
AP I (n 10) Article 57(2)(a)(iii) and (b).
See Section 5.4.2.
Dinstein, Conduct of Hostilities (n 69) p. 157.
Schmitt and Widmar, ‘The Law of Targeting’ (n 169) pp. 140–140.
Prosecutor v. Galic (Judgement and Opinion) (n 328) para. 55. See also Sandoz et al. (eds.), AP Commentary (n 4) para. 2208; amw Manual (n 4) Rule 14, Commentary para. 6; Dinstein, Conduct of Hostilities (n 69) p. 159; and Section 5.4.2 above.
See also ila, ‘Challenges of 21st Century Warfare’ (n 289) p. 369.
amw Manual (n 4) Rule 14, Commentary para 9. See also Boothby, The Law of Targeting (n 3) p. 94, and ila, ‘Challenges of 21st Century Warfare’, ibid, p. 354.
Dinstein, Conduct of Hostilities (n 69) pp. 160–161, criticising the Commentary to AP I for making reference to substantial military advantage. Sandoz et al. (eds.), AP Commentary (n 4) para. 2209: “The expression “concrete and direct” was intended to show that the advantage concerned should be substantial and relatively close, and that advantages which are hardly perceptible and those which would only appear in the long term should be disregarded.” The term ‘substantial’ was included in the icrc proposed draft of the provision but did not make it into the final version. See Draft Article 46, para. 3 (b), included in fn. 31 in Sandoz et al. (eds.), AP Commentary, ibid, para. 1976.
Several States made declarations upon the ratification of AP I similar to the UK’s declaration that: “the military advantage anticipated from an attack is intended to refer to the advantage anticipated from the attack considered as a whole and not only from isolated or particular parts of the attack.” United Kingdom, Declarations made upon signature of the 1977 Additional Protocol 1, §i. See also declarations made by Austria, Belgium, Canada, France, Germany, Italy, Netherlands, New Zealand, and Spain (available at
icty, OTP Report (n 9) para. 78. See also discussion in Johansen, On Military Necessity, ibid, pp. 289–291.
The U.S. Law of War Manual, however, stipulates that the strategic concerns may also be taken into account. U.S. DoD, Law of War Manual (n 24) para. 15.12.2.1, p. 264.
Dinstein, Conduct of Hostilities (n 69) pp. 108–109.
Boothby, The Law of Targeting (n 3) p. 370.
U.S. DoD, Law of War Manual (n 24) para. 15.12.1.2, p. 261.
For a useful overview of the debate, see Johansen, On military necessity (n 8) pp. 257–262.
ila, ‘Challenges of 21st Century Warfare’ (n 289) pp. 352–355. For an example of the contrary view, see Dinstein, Conduct of Hostilities (n 69) p. 159, stressing that “the only consequences that count are those that occur directly: remote effects need not be counted.” See also U.S. DoD, Law of War Manual (n 24) p. 261, where it is stated that collateral damage is limited to “such immediate or direct harms foreseeably resulting from the attack. Remote harms that could result from the attack do not need to be considered in applying this prohibition.” However, the manual goes on to provide the following example: “if the destruction of a power plant would be expected to cause the loss of civilian life or injury to civilians very soon after the attack due to the loss of power at a connected hospital, then such harm should be considered in assessing whether an attack is expected to cause excessive harm.” It is difficult to see how this harm would be “immediate or direct”. In fact, the Allied bombing of the Iraqi electricity grid during the First Gulf War (1991), resulting amongst others in devastating effects on the hospital in Bagdad, is the classic example of why remote harms must be taken into account when foreseeable. See e.g. Johansen, On military necessity (n 8) pp. 259–262.
ila, ‘Challenges of 21st Century Warfare’ (n 289) p. 354. See also Johansen, On military necessity, ibid, pp. 260–261.
See e.g. icrc, Explosive Weapons in Populated Areas: Humanitarian, Legal, Technical and Military Aspects (Report from Expert Meeting, Chavannes-de-Bogis, 24–25 February 2015, available at
The debate is reflected in the revised Commentary to the First Geneva Convention: “It would be unreasonable to consider that direct or indiscriminate attacks against the wounded and sick would be strictly prohibited and would amount to a grave breach, while incidental harm and even excessive incidental casualties would not be prohibited.” icrc, Commentary on the First Geneva Convention – Convention (I) for the Amelioration of the Condition of the Wounded and Sick in Armed Forces in the Field [hereinafter: GC I Commentary] (second edition, 2016, available at
See for instance Jann Kleffner, ‘Transatlantic Workshop on International Law And Armed Conflict: Wounded and Sick and the Proportionality Assessment’ (icrc Intercross blog, 12 October 2017,
Dinstein, Conduct of Hostilities (n 69) p. 155.
Gisel, ‘Can the incidental killing of military doctors never be excessive?’ (n 375) pp. 220–221.
This is the argument made by Gisel. See ibid, p. 230: “the protection against incidental harm granted to civilians and civilian objects by the principles of proportionality and precautions must actually be understood as extending to other persons and objects protected against direct attack”. See also ila, ‘Challenges of 21st Century Warfare’, (n 289) p. 358.
Geoff Corn, ‘Transatlantic Workshop on International Law and Armed Conflict Wounded and Sick, Proportionality, and Armaments’ (Lawfare, 10 October 2017,
Rome Statute (n 31) Article 8(b)(iv).
The principle is commonly referred to as either nullum crimen, nulla poena sine praevia lege poenali (no crime can be committed, nor punishment imposed without a pre-existing penal law), nulla poena sine lege (no penalty without law) or nullum crimen sine lege (no crime without law). It is reflected in the Rome Statute, ibid, Article 22(1).
Rome Statute (n 31) Article 22(2) states that: “The definition of a crime shall be strictly construed and shall not be extended by analogy. In case of ambiguity, the definition shall be interpreted in favour of the person being investigated, prosecuted or convicted.” According to Cassese, this reflects customary law and is a result of the principle of strict legality. Antonio Cassese et al., International Criminal law (Oxford University Press, Oxford, 2013) p. 33. See also Dinstein, Conduct of Hostilities (n 69) p. 155.
The Marten’s clause first appeared in the Preamble to Hague Convention (II) with Respect to the Laws and Customs of War on Land and its annex: Regulations concerning the Laws and Customs of War on Land, The Hague [29 July 1899], printed in Schindler and Toman, The Laws of Armed Conflicts (n 4) pp. 69ff: “In cases not included in the Regulations adopted by them, populations and belligerents remain under the protection and empire of the principles of international law, as they result from the usages established between civilized nations, from the laws of humanity and the requirements of the public conscience”. It has reappeared in various but similar versions in later treaties regulating armed conflicts, see e.g. 1907 Hague Convention (IV) (n 28) Preamble; GC I (n 79) Article 63; GC II (n 79) Article 62; GC iii (n 29) Article 142; Geneva Convention (IV) Relative to the Protection of Civilian Persons in Time of War (GC IV) [1949], printed in Schindler and Toman, The Laws of Armed Conflicts (n 4) pp. 575ff., Article 158; AP I (n 10) Article 1(2); AP II (n 66) Preamble; and 1980 Convention Certain Conventional Weapons (n 371) Preamble.
Corn and Culliver, ‘Wounded Combatants, Military Medical Personnel, and the Dilemma of Collateral Risk’ (n 404) pp. 14–16.
Schmitt, ‘Targeting and IHL in Afghanistan’ (n 11) p. 322.
See Section 2.4.1. From 2007–2011, various isaf commanders issued Tactical Directives on the avoidance of civilian casualties (so-called civcas directives). Some of these are available to the public, see e.g.:2008:
Johansen, On military necessity (n 8) pp. 332–333.