13.1 nato roe and the Legal Basis for the Use of Force
13.1.1 General Comments
When nato forces use force, they must ensure that it complies with both the rules of engagement (roe) for the operation and applicable law. This requires them to fully understand the legal and operational framework for the use of force and how it applies to their present context. It would therefore be helpful if the ‘use of force’ concepts applied in roe and the applicable law could be communicated in a clear and concise manner. However, this requires further clarity concerning the scope of the operational categories, what the respective legal bases are, and how these legal bases apply in practice. The intention of this book was to shed further light on these questions, so that military forces may be given clearer guidance.
roe are political and operational restraints on the use of lawful force. Because the focus in the roe is on the tasks authorised rather than the legal basis for such acts, ‘use of force’ authorisations are formulated through operational categories and concepts. Fifteen years after the current nato roe doctrine, the MC 362/11, was approved, it is prudent to examine how these ‘use of force’ categories and concepts have developed and how they interrelate with their respective applicable legal bases. The last fifteen years have tested not only nato roe doctrine but also the law of armed conflict (loac). In particular, the concept of direct participation in hostilities has been the subject of extensive debate, and attempts have been made to clarify how it should be applied. As a result of the decision to focus this book on operations involving participation in an armed conflict, the concept of ‘direct participation’ is central to understanding when nato forces may use force against other persons.
The need to examine the relationship between the ‘use of force’ categories in nato roe and the applicable legal framework is also influenced by the approach nato has taken to roe for operations during armed conflict. While it was initially envisaged that roe for such operations would consist of limitations on the use of force that is lawful under loac but nonetheless not permitted, the nato operations in Afghanistan and Libya demonstrated that this approach did not work in practice. The approach taken has instead been to use the roe to define all circumstances when force and other provocative measures may be used, not merely when lawful force may not be used. As a result, during nato operations, all use of force or other provocative measures must be authorised by roe. The only exception to this rule is the use of force in self-defence.
As a result of the sole authority approach, self-defence has become both a legal exception for the use of force and an operational exception. The legal concept of self-defence is an exception to the prohibition on arbitrary deprivation of life. Operationally, self-defence is also being used to refer to the ability to use force beyond roe. In other words, it is used as an operational exception to the requirement that all use of force during nato operation must be authorised by roe. The legal basis for the use of force is not, however, determined by the terminology used to explain it, but rather by the facts and applicable law. Because loac applies to the use of force in offence and defence,2 the application by military forces of both the legal and operational self-defence concepts is intriguing. While the use of defensive force is generally not regulated by nato roe, the existence and importance of self-defence is referenced in the roe, and the study of the ‘use of force’ categories in nato operations would not be complete without including the legal and operational self-defence concepts.
The search for the legal basis for the use of force authorised by nato roe has revealed that there are fundamentally different approaches to the relationship between legal bases for the use of force and rules regulating the use of force. For some, they are the same; the rules providing exceptions to the prohibition on arbitrary deprivation of life serve as the legal basis for the use of force. In the current context, this means that when loac applies, the use of force authorised by roe will be legally based on loac, through the concept of ‘combatant immunity’ and ‘lawful acts of war’, or in law enforcement authorities. Others, however, view loac merely as the rules explaining how force may be used. For them, it does not provide legal authority, which, when a State is operating outside its own territory, must be found in the jus ad bellum. As a result, the use of force necessary to achieve the jus ad bellum goal will be presumed lawful unless expressly prohibited, even if there is no clear rule providing the authority to use force in the circumstances at hand. As explained in Section 4.3, this approach to the jus ad bellum results in an undesirable conflation of the two areas of law, and departs from the important lessons of the Nuremberg Tribunals that loac applies equally to all parties, regardless of who is perceived as having the lawful cause for being in an armed conflict.
Another area where States and commentators differ in their approach is with regards to the application of ihrl, especially in extraterritorial operations. In addition, nato States have ratified different loac treaties. As a result of these differing international legal obligations, States may have divergent approaches to, for instance, detention. However, in the context of the use of force against persons, the impact is less significant. First, the duty to respect the prohibition on arbitrary deprivation of life is so thoroughly entrenched in international law that it has arguably become a peremptory norm, reflected in both ihrl and loac. Second, States generally agree that the lex specialis of armed conflict is loac, and a large part of loac regulating the conduct of hostilities is now recognised as customary international law and therefore applies equally to all States.
As a result, the use of force by military forces during armed conflict, expected to affect the right to life of others, must first and foremost comply with loac. If loac is insufficient, self-defence will in extraordinary circumstances provide an alternative authority for the use of force. Legally, military forces may also use force in the context of law enforcement. However, nato forces are only exceptionally involved in such activities, and the topic has therefore not been examined in this book.
As mentioned in the introductory chapter, the research was also made challenging by the use of identical or similar terms to describe different concepts, a phenomenon that can conceal divergence in national approaches. This is so with regard to many of the central terms and concepts of the book, such as ‘hostile act’, ‘hostile intent’, ‘attack’, ‘self-defence’, ‘combatant’, and ‘direct participation in hostilities’ (dph), and is one of the reasons why the exact legal regulation of the respective ‘use of force’ concepts is unclear. As a result, there is a need for both enhanced awareness of these divergent views, and for further clarification.
The following subsections will summarise the findings concerning the relationship between the use of force authorised by nato roe, whether explicitly in the roe or implicitly as self-defence, and applicable law. The focus will be on the potential room for manoeuvre under the respective categories, and whether any changes should be made in how they are interpreted.
13.1.2 The Use of Force in Self-Defence during Armed Conflict
Although it is only indirectly a part of nato roe, the use of force in self-defence is the only exception to the requirement that all use of force must be authorised by roe. As such, it has played a particular role in nato ‘use of force’ doctrine. Contrary to the impression that the reliance on self-defence during military operations may give, the legal concept of self-defence only applies in exceptional circumstances. During armed conflict, this is defined by the respective troop-contributing nations’ domestic legislation, unless a third party enters the conflict and thereby potentially triggers State self-defence.
Central to all expressions of the legal concept of self-defence is the idea that otherwise unlawful acts are justified if they are taken as a necessary and proportionate response to an unlawful attack or one that is imminent. The reference to “otherwise unlawful acts” limits the potential application of self-defence during armed conflict because the use of force by military forces against opposing forces during an armed conflict is lawful, as a result of the loac concepts of ‘combatant immunity’ and ‘lawful acts of war’. In addition, if the opponent is a combatant, his or her attack is also a lawful act, with the result that there is no ‘unlawful attack’ to defend against with self-defence. Self-defence is therefore mostly relevant for military forces participating in armed conflict in relation to threats arising from persons known to be civilians, or where their status as lawful targets is not sufficiently clear and the presumption of civilian status applies.
During niacs, the opposing forces do not have the right to participate in hostilities. As a result, their attack on nato forces would most likely be considered unlawful under applicable domestic law, which means that self-defence requirement of an ‘unlawful attack’ is met. However, nato forces will most likely be authorised to use force, making their acts lawful rather than “otherwise unlawful”. Furthermore, as a result of the requirements of necessity and proportionality, force may only be used as a last resort; it cannot be incited, and, if possible, attempts must be made to try to avoid its use, at least if the threat is not imminent. Excessive use of force may prevent a finding of self-defence.
The use of force in self-defence must be reasonable, and if mistakes are made, for instance with regard to the severity of the situation or the degree of force required to fend off an attack, this will only be accepted if they are considered reasonable. In both regards, military forces are likely to be assessed to a higher standard than civilians in general, both due to their training and experience with the use of force and the effects of this, and because it should be less likely that they overreact due to fear or panic. They will, in other words, be assessed on the basis of what a ‘reasonable soldier’ would have done, not a reasonable civilian.
Although self-defence is most likely to be raised in connection with force being used against persons who are not lawful targets, it may also apply to the use of means and methods prohibited by loac. For instance, the use of tear gas or exploding ammunition under 400 grams is by most States considered to only be permissible for, respectively, law enforcement purposes and against vehicles. However, if they are the only means available to effectively act in self-defence, their use may be accepted as lawful if proportionate. Although the use of unlawful methods in self-defence is theoretically lawful, most prohibited methods, such as starvation, are unsuitable as individual acts in response to imminent threats. This is because, inter alia, they require planning, take time to carry out, or lack defensive intent. Self-defence may nonetheless be accepted for a violation of prohibitions such as the misuse of enemy uniform or feigning protected status, at least if it is done in order to escape the situation rather than attack.
Whereas causing civilian casualties may be acceptable when using force based on loac, harm to innocent third parties caused during acts of self-defence is more complicated. Because the innocent parties do not represent a threat, no right of self-defence arises in relation to them. Legally, such harm should therefore be treated as a question of necessity; however, the tendency seems to be to assess all consequences of the self-defence act together, under the proportionality principle. Still, the harm to third parties is subject to a stricter proportionality assessment, meaning that it must be clearly outweighed by the harm the defensive act is intended to prevent. While harm caused recklessly is likely to be deemed disproportionate, harm to others may be considered reasonable if it was unintentional or, as may be the case in a hostage situation, it is caused knowingly in order to prevent greater harm.
Defensive acts may be undertaken in order to protect oneself, others and to some extent property. For nato operations, the protection of others should be interpreted to include all forces participating in that operation, in furtherance of Article 5 of the North Atlantic Treaty.3 At the same time, it is preferable that the authority to defend others is provided for in the roe, making it unnecessary to argue self-defence other than in exceptional circumstances. The protection of coalition forces should also be regulated in roe. The use of force to protect persons who are not part of the nato mission or coalition forces, such as members of the civilian population or ngos, is more complicated. It will depend, inter alia, on the scope of the applicable domestic self-defence rule; whether the State has a duty to protect them; and the risk of harm to those carrying out the defence. Interference in situations unrelated to the operation may have negative implications for mission accomplishment because it interrupts an operation or, more importantly, is sometimes culturally unacceptable. It may therefore be limited, through roe, to situations of serious harm.
This ability to impose limitations on the use of force in self-defence by military forces is not limited to the defence of others unrelated to the mission. The authority to use force to protect property may also be limited. Furthermore, the use of force to defend other nato forces may be restricted if it is expected to cause unreasonable risk to one’s own forces. To the extent such limitations are not tantamount to ordering military forces to commit suicide, Commanders may also impose restrictions on the ability of subordinate personnel to defend themselves. However, such restrictions must be reasonable, and must be imposed, for instance, to gain a tactical advantage or to enhance the protection of the civilian population. Furthermore, limitations are more likely to be acceptable in response to imminent attacks than ongoing ones, and where the attack is expected to cause light injuries rather than death or serious injury.
If military forces use force in violation of a military order, the force itself may nonetheless be lawful either on the basis of self-defence or loac, but because military forces have a duty to comply with orders that are not manifestly unlawful, acting contrary to orders may itself be a criminal offence.
Whereas the legal concept of self-defence operates as an exception to the prohibition on arbitrary deprivation of life, the operational concept applies as the only exception to the roe limitations on the use of force. It permits the use of force not authorised by roe, which means that the legal authority for that use of force may be either loac or self-defence. As a result, it may be applied in a wider range of situations than the legal concept. To the extent the use force may be characterised as loac defensive force, the imposition of limitations in the form of necessity and proportionality in the self-defence meaning of those terms amount to operational or political rather than legal limitations.
In incidents where the operational concept of self-defence is applied, the main challenge to legality is whether sufficient precautions are being taken in order to avoid harm to civilians, and whether means or methods prohibited by loac are being used because the action is considered to be self-defence. As the discussion in Section 8.3 suggests, the risk of violating loac when applying the self-defence requirements for the use of force is minimal. There is nonetheless a concern that the conflation of two areas of law, with very different purposes, will have negative effects on both. The use of force accepted as self-defence in an armed conflict where the legal basis was loac may fail to meet the self-defence criteria in a different, peacetime, context. If not informed of the differences, military forces may unintentionally use force beyond what is permitted. Furthermore, the focus on self-defence even during armed conflict can cause military forces to become hesitant in defending themselves, thereby putting themselves and others at additional risk.
13.1.3 The Use of Force in Response to Attacks or Imminent Attacks by Opposing Forces
The use of force by nato forces in response to attacks or imminent attacks by opposing forces is a central aspect of warfighting and would as such be expected to be part of the roe rather than dependant on self-defence. However, it is not among the proposed roe in the MC 362/1, and is usually dealt with as ‘self-defence’, despite loac applying to the use of force in defence as well as offence. This is likely a result of the change in the approach to nato roe, whereby the use of force during all types of operations requires roe, even during armed conflict.
Because the use of force in such situations is doctrinally dealt with as a form of self-defence, military forces are given the impression that the use of force in such situations will be lawful when it is necessary and proportionate. Although these peacetime ‘use of force’ concepts do not require considerations akin to the loac rules on precautions in attack (and in particular the avoidance of collateral damage), they do limit the use of force to that which is absolutely necessary. Legally, regulating the use of force in contact situations in this manner should not cause nato forces to act unlawfully.
It would nonetheless be preferable to have explicit rather than implicit authorisation of the use of force in such situations. Furthermore, self-defence is by its nature an exception, and defending against attack from opposing forces does not fit with that characteristic. The challenge is to a find a way to formulate it so that all nato States will accept. Because it is closely connected to self-defence, it is a sensitive issue and, as previously explained, not all States consider loac to apply to defensive force. Furthermore, if such roe were to be included, it would be important to emphasise that it should only be subject to restrictions on its use or be retained in exceptional circumstances.
13.1.4 Use of Force to Achieve Designated Tasks
This category covers all use of force which does not qualify as an attack and is not defensive force under either self-defence or in combat engagement. It therefore includes a wide variety of situations, from searching and seizing vessels to defending third parties from serious crime. This category raises a different issue than the others: to what extent may force be used against individuals outside of attack? If the person harmed is a lawful target, the use of force against that person would most likely amount to a lawful attack under loac. The fact that the nato forces were tasked to carry out, inter alia, a search of that person’s property does not affect the legality of the use of force under loac. However, if the person is not a lawful target, the only way nato forces may expose that person to harm is if it is incidental to an attack on a lawful target and not deemed excessive. As a result, if there is no attack, force may not be used if it harms civilians. In any case, force may never be directed at a person not considered to be a lawful target (except in self-defence). Instead, the requirement to take constant care to spare the civilian population will apply. Although loac is not the only potentially applicable legal regime, other areas of law, whether national or international, are likely to only permit the use of force as a measure of last resort, and to only permit the use of lethal force in exceptional circumstances.
As a result, it is important not to vaguely or widely formulate or interpret the roe permitting the use of force to accomplish designated tasks. This means that roe authorising the use of force for ‘mission accomplishment’ or to secure ‘freedom of movement’ must be qualified in a manner that avoids the use of deadly force against individuals who are not lawful targets or causing harm that cannot qualify as lawful collateral damage.
The roe authorising the use of force to accomplish designated tasks may be formulated as either permitting ‘minimum use of force’, which includes deadly force, or lesser forms of force, usually expressed as non-deadly or non-lethal force. According to the MC 362/1, if the decision is made to only authorise non-deadly force, the term ‘minimum’ should be deleted. However, because these roe have the potential for authorising the use of force not authorised by loac and may therefore based on peacetime rules, the requirement to use no more force than necessary will still apply. The term ‘minimum’ implies that escalation of force procedures should be applied, which is a useful reminder even where only non-deadly force is authorised. It ensures that the force used is no more than necessary, and that the situation does not escalate into a self-defence situation where this could have been avoided. The term ‘minimum’ should therefore always be included.
13.1.5 The Attack Series
13.1.5.1 The nato Hostile Act and Hostile Intent roe
The nato roe authorising attack on persons carrying out a hostile act (not constituting actual attack) or demonstrating hostile intent (not constituting imminent attack) are ‘use of force’ authorities applicable to conduct-based targeting. They are therefore different from national ‘hostile act’ and ‘hostile intent’ concepts, which generally refer to self-defence. Because the roe authorise the use of force in response to threatening behaviour, they are particularly useful in operations where the opposing forces fail to distinguish themselves from the civilian population. The nato ‘hostile act’ concept enables nato forces to respond to any intentional acts that cause prejudice or pose a danger, while the nato ‘hostile intent’ concept authorises attack on persons having the intent to attack or otherwise inflicting damage before the threat becomes imminent. Because the roe authorise ‘attack’ rather than ‘minimum use of force’, there is no requirement to use graduated force. Escalation of force procedures may nonetheless be useful to ascertain intent.
The comparison of the nato hostile act and hostile intent roe and the most likely legal basis for their application during armed conflict, namely direct participation in hostilities, revealed that in most cases, the roe concepts are narrower than the legal concept. Whereas the nato ‘hostile intent’ concept requires ‘a substantial threat’, ‘adverse effects’ are sufficient under dph. The ‘causal link’ criterion of dph is met by the requirement of ‘capability, preparedness and intention to attack or otherwise inflict damage’ under the roe concept. Finally, although the belligerent nexus requirement for dph is not clearly reflected in the ‘hostile intent’ concept, it is likely that general training on lawful targets and mission specific training on what constitutes a likely threat in the area of operation will ensure that the roe is applied in compliance with this dph criterion.
Because the nato ‘hostile act’ concept includes not only acts preparatory to an attack, but also other acts causing serious prejudice, it is potentially wider than dph, particularly when the hostile act is committed against ‘persons with designated special status’ (pdss) rather than nato forces or partner forces. This is because the threshold for dph is higher when directed at civilians than opposing forces; it must be likely to cause death, injury or destruction. This aspect of the nato ‘hostile act’ concept therefore requires the person committing the hostile act to the lawful targets at all times in order for an attack in response to the hostile act to be lawful. Alternatively, the application of the concept may be reformulated to only authorise attack on “designated forces or individuals who commit or directly contribute to any intentional act causing serious prejudice or posing a serious danger to nato forces or posing a serious danger to pdss”.
While the second dph criterion, a causal link, is most likely met by the reference to ‘designated forces or individuals who commit or directly contribute to’ hostile acts, there is a potential for ‘acts causing serious prejudice’ to encompass acts that do not meet the belligerent nexus criterion. This is because the acts covered by the roe, such as impeding nato military operations, may also be carried out without any intention to adversely affect the military operation. If the act is not specifically designed to create such adverse effects, it is not dph. As suggested above in relation to hostile intent, compliance with the belligerent nexus criterion should most likely be ensured through training.
As a result of the reactive nature of the nato ‘hostile act’ and ‘hostile intent’ concepts, the ‘for such time’ criterion is most likely complied with when applying these roe. In fact, it is likely that acts preparatory to dph, which is considered part of the direct participation, are not covered by the nato hostile act and hostile intent roe. Furthermore, the nato hostile act and hostile intent roe will not authorise attack on persons considered to take a direct part in hostilities on a regular basis unless they commit a hostile act or demonstrate hostile intent at the time.
Finally, because both dph and the nato hostile act and hostile intent roe are highly contextual, it is important that military forces be provided with clear mission specific examples of how the roe should be applied. Mission specific examples of what constitutes a hostile act or hostile intent are also useful sources when analysing State practice on the application of dph.
13.1.5.2 Attack on Forces ‘Declared Hostile’
The use of force or an attack on forces ‘declared hostile’, sometimes referred to as ‘roe defined opposing forces’, is the least complicated ‘use of force’ category, at least when the opponent complies with the principle of distinction by distinguishing themselves from the civilian population. However, while combatants in an iac are likely to wear uniforms or other fixed distinctive signs, opposing forces in a niac are less likely to comply with the principle of distinction, with the potential exception of dissident armed forces. This makes this roe category more challenging to apply in niacs.
As explained in Chapter 12, the identification of opposing forces in a niac is usually done on the basis of their conduct; membership in an organised armed group will in most cases be difficult to discern when the members wear civilian attire. As a result, the most practical legal criteria for identifying opposing forces in a niac are those enabling a determination of direct participation in hostilities, not a theoretical test for membership. State practice suggests that the protection from direct attack may be lost for an extended period of time, either on the basis of membership in an organised armed group or other indications suggesting continuous direct participation in hostilities. Members of an organised armed group will therefore factually constitute a separate category, although the identification is likely based on their conduct and hence the criteria for dph.
Status-based targeting allows opposing forces to be attacked at any time, regardless of current activity, provided they have not become hors de combat. As a result, nato forces can be proactive in their use of force, rather than just reactive. Furthermore, status-based targeting enables detailed planning, which again makes it possible to apply targeting procedures designed, inter alia, to enhance loac compliance.
13.2 The Need for a Holistic and Uniform Approach to roe
In addition to setting out the nato ‘use of force’ categories and corresponding legal bases, the research has also given rise to more general conclusions on nato roe doctrine. The expectation that nato forces only use force authorised by roe (unless self-defence) affects the approach taken to nato roe with regard to drafting, implementation and interpretation. When the sole authority for using force beyond self-defence, the roe for an operation must cover all potential aspects of the operation; there can be no ‘gaps’ between roe series or rules where desired and lawful acts may fall. The roe must in other words not only be relevant and lawful, they also need to be sufficient to accomplish the mission. Furthermore, they must be sufficiently broad or flexible to enable their adaptation to changing situations. There is, in other words, a need to have a holistic approach to roe, and the nato roe doctrine should have templates covering the most likely potential aspects. This requires that the current version, the MC 362/1, be updated to include, among others, more modern cyber roe. Additionally, the authority to respond to threats or attacks by opposing forces on the basis of loac rather than self-defence should be explicit rather than implicit. Preferably, such roe should be introduced into the nato roe doctrine as this will help ensure a uniform and predictable approach to the issue. In the absence of a roe template authorising the use of defensive loac force, it should be created in a spare number.
To fight without orders is almost always, in the case of a soldier, to fight contrary to express orders. Accordingly there is hardly any other case than that of self-defense when a soldier or subordinate officers may act without orders. In that case orders may safely be presumed; or rather the right to defend oneself from attack belongs naturally to every man, and no permission is needed.4
The sole authority approach to roe for all operations has nonetheless been met with criticism, in particular in the context of armed conflict operations, and especially nato Article 5 operations. This appears to be the result of divergent command and control approaches. Those in favour of mission or de-centralised command are likely to prefer as few roe as possible, while detailed roe enable order based or centralised command. However, completely de-centralised command, to the extent that the military forces are authorised to use the force they consider necessary, provided it complies with loac, appears utopic, particularly during multinational operations requiring joint political approval. Because roe are currently one of the preferred tools for imposing political constraints on the use of force and enable operational command and control, it is difficult to envisage an operation without them. As a result, it seems likely that the sole authority approach for all operations will be continued.
The current tendency in nato to apply the same roe methodology to all operations is favourable for several reasons. It reflects the political demand for increased control over the use of military force, while the uniform roe methodology for all operations avoids a potential source for confusion. Due to the inherent flexibility in nato roe doctrine, both the contents, number and detail of roe may be adapted to any operation. During operations below the threshold of an armed conflict, the focus will be on the tasks at hand and the use of minimum force to achieve these. During armed conflicts, roe permitting attack on lawful targets will also be expected to be included. If the conflict is politically sensitive, the roe can be tailored to enable detailed control and limited use of force, for instance, by developing numerous and detailed roe that may be retained at higher levels of Command. If it is a nato Article 5 conflict, the roe may be broader and subject to less limitations, and as a result, more closely reflect the room for manoeuvre defined by loac.
Another benefit of the uniform approach to roe for operations both below and above the threshold of an armed conflict is that the North Atlantic Council (nac), which authorises the roe for all nato operations, will not need agree on the legal classification of the situation. As long as all nato States consent to the roe, the legal and political assessments justifying that consent is a national rather than multinational responsibility. Furthermore, a distinction in roe methodology for different types of armed conflict would be particularly confusing in operations where deterrence fails and the situation changes from a crisis to an armed conflict. Similarly, where there is a combination of traditional and hybrid warfare, and therefore a need for roe both dealing with armed conflict type scenario and authorities to use force in reaction to threats not sufficiently connected to the armed conflict, this uniform approach is clearly beneficial.
Another, and perhaps unintended, consequence of the emphasis on roe authorisation for all use of force is the resulting increased focus on the exception, namely self-defence. The use of force in self-defence is legally only permitted in exceptional circumstances. Even extensive applications of the ‘operational self-defence’ concept, legally founded on loac, is problematic because the use of force outside the roe framework would be the use of force beyond the Commander’s command and control. Integrating as many aspects of the use of force during military operations into the roe as possible enhances operational and political control because it makes it possible to subject the use of force to limitations and to retain the authorisation level. Even though military forces retain a right to defend their lives, Commanders are able to impose limitations on certain aspects of that right, such as by controlling the response to imminent but not yet occurring attacks, or the defence of others.
The comments by civilian investigators and military lawyers suggest that the existence of an alternate justification for the use of force that is considered to be inherent and inalienable (and thus, unquestionable) and is highly subjective makes it easier to avoid independent scrutiny. The greater prevalence of self-defense and hostile intent in the Afghanistan context means that this inscrutable defense would be applied to a greater number of incidents. At a macro level, this would deflect scrutiny over a wider proportion of use of force situations and would lead to overall weaker accountability for allegations civilian harm, and overall weaker ihl accountability.5
Reducing the scope of applying self-defence as an exception to roe, by ensuring that the roe for an operation better reflect both the offensive and defensive force required to accomplish the mission, could therefore also enhance loac accountability.
The overemphasis on self-defence is not only the result of restrictive roe and the sole authority approach; complicated ‘use of force’ categories, especially the nato hostile act and hostile intent roe, and insufficient guidance on how to apply complex concepts such as dph, are also to blame. As Corn points out, “in the absence of clear policy guidance, self-defense authorities have frequently been invoked at the operational and tactical levels as the basis for targeting presumptive civilians taking direct part in hostilities”.6 nato should therefore endeavour to provide further guidance on mission specific application of complex ‘use of force’ authorities, for instance as an appendix to the roe, or in a sop or frago.7 The current challenges should also be used as a lesson learned to not so easily apply existing approaches and solutions to new challenges without thoroughly analysing the appropriateness of doing so. In particular, national solutions and concepts do not always easily transfer into a multinational context.
The final comment that should be made on the overall application and interpretation of nato roe is the relationship between the roe and other ‘use of force’ guidance. While it would be beneficial for the sake of clarity to include most if not all ‘use of force’ guidance in one document, this would also be impractical. It would require nac approval to all details concerning use of force, which may be difficult and time-consuming to achieve, and it would leave the Commander with limited ability to further adapt the ‘use of force’ guidance to the changing realities of the operation. The better approach is therefore to emphasise that all ‘use of force’ guidance must comply with the roe, and to keep the number of alternative sources to a minimum.
13.3 The Relationship between Self-Defence, loac and roe
In that the initial interest in the research topic focused on the relevance and application of self-defence by military forces during armed conflict, and as a reaction to the surprisingly extensive reliance on self-defence to justify the use of force, it seems only fitting to end the book with some overall comments on the relationship between self-defence, loac and roe.
The relationship between self-defence and loac is commonly described by referring to the former as defensive and the latter as offensive. Because loac regulates the use of force in offence and defence, this is an oversimplification. In the context of armed conflicts, defensive force will primarily be regulated by loac. Self-defence will only apply in those cases in which the initial attack is unlawful, and is relevant where loac does not apply, thereby providing more robust authorisations to deal with the threat. As a result, loac and self-defence apply in parallel, and are not one the same scale. Furthermore, rather than stating that ‘roe begins where self-defence ends’, nato doctrine should make it clear that self-defence is not the starting point during armed conflict. Legally, self-defence is the fall back, and factually, self-defence happens later.
The area in which self-defence and loac most closely interact is in the context of conduct-based targeting and dph. If the belligerent nexus requirement is not met, the only way to lawfully use force in response to the threat is on the basis of self-defence. If the ‘for such time’ criterion of dph is narrowly construed, the military forces are required to catch the person ‘red handed’, which may escalate the situation into what the military forces perceive as self-defence. The application of self-defence during armed conflict is therefore influenced by the interpretation of dph. A wide interpretation of dph reduces the need for relying on self-defence. A corresponding broad interpretation of the nato ‘hostile act’ and ‘hostile intent’ concepts, authorising the use of force on the basis of dph, will also reduce the scope for relying on the ‘operational self-defence’ concept because the roe would be more inclusive and better reflect the military necessities for using force.
As a result, States must ensure that the interpretation and communication of the criteria for both dph and the nato hostile act and hostile intent roe are sufficiently clear to be able to apply at the operational and tactical levels. If not, self-defence will be applied instead.8 Complex operational realities may tempt Commanders to introduce further and more detailed regulation. However, as others also have emphasised, “[a]s conflict structures become more and more diffuse, legal certainty and clarity of humanitarian law prescriptions become ever more important.”9 loac should not become so complicated that it is impossible to apply.
In particular, it must be possible to identify the opposing forces beyond situations where they are about to attack nato forces. This requires that the criteria for membership in an organised armed group in particular become more practical. Further, the scope for repeatedly stepping in and out of civilian protection, and hence misusing the protection to gain operational advantage, should be reduced. It was argued in Chapter 5 that despite the attempts made to define whether someone is a member of an organised armed group, State practice suggests that the identification of such members is done on the basis of their conduct. Rather than developing complex, generally applicable legal criteria for membership in an organised armed group, the better approach is to develop clearer criteria for continued loss of protection. This would make it possible to treat all non-State actors as continuous dph. Phrased differently, the legal criteria for losing civilian protection will be the same for all non-combatants, both during iacs and niacs.
Being a member in an organised armed group would, however, be a valuable source of information for making the assessment of continuous dph. In operations where such membership and hence continuous dph is possible to discern, mission specific criteria for status-based targeting may be developed, thereby providing military forces with a clearly defined opponent. This enables the application of roe authorising attack on forces ‘declared hostile’, rather than having to rely on the reactive nato hostile act and hostile intent roe. The authority to attack persons considered to regularly take direct part in hostilities without a sufficiently clear connection to a group could be retained at a higher level, thereby enabling further assessment and control, or limited to situations where the person demonstrates a hostile intent or commits a hostile act. Due to the complexities of these roe concepts, it is important to provide mission specific examples of what amounts to ‘hostile act’ or ‘hostile intent’, for instance in the oplan.
In addition to providing military forces with clear mission-specific criteria for becoming a lawful target, it is also important to emphasise that absolute certainty is not required for target identification. Instead, an attacker is expected to act in honest belief and make a good faith assessment of the information reasonably available at the time. Rather than focusing on whether there is any doubt as to the status of the person attacked, the test should be whether a reasonable soldier in the same or similar situation would have made the same decision. If military forces are more comfortable in applying loac and roe, they are less likely to look to alternative authorities not tailored for the unique situations that arise in an armed conflict.
Although any potential expansion of dph may raise concerns that the protection of civilians is reduced, this is not necessarily the case. First, clear loac rules, including a more inclusive interpretation of dph, reduce the scope for resorting to self-defence. This in turn allows the Commander to exercise further control over the use of force, since most uses will require roe authorisation. This is beneficial both for mission accomplishment and the protection of civilians. Interpreting loac and dph to include members of organised armed groups at all times would also enhance the distinction between civilians and those participating in the armed conflict. Civilian protection is premised on civilians abstaining from participating in hostilities. This could arguably reduce the risk of harm to protected civilians. As Hays Parks explains, permitting those who nonetheless choose to directly participate in hostilities to keep returning to being a protected civilian is “to the benefit of a very few but to the danger of the ninety-five percent of the civilian population in any nation who do not take part in the hostilities”.10
When the opposing forces are lawful to attack at any time, it also becomes possible to plan the attack in more detail, and to apply targeting procedures such as weaponeering and collateral damage estimation aimed at reducing civilian casualties. Furthermore, dealing with situations before they escalate to the level of potential ‘self-defence’ enables the forces to deal with the situation in a way that provides greater tactical control.11 This reduces the risk of harm both to the military forces and to nearby civilians. The risk of civilians being mistaken for taking direct part in hostilities on a regular basis would be countered by setting the threshold for finding continuous dph sufficiently high.
Clarifying the scope of dph and the nato hostile act and hostile intent roe would also close some of the interoperability gap between States with expansive self-defence interpretation and reliance and those with a more restrictive approach; more precisely, between the U.S. and European States. Furthermore, reducing the scope for relying on the ‘operational self-defence’ concept also avoids the challenges discussed in Section 8.3, whereby the quasi-offensive operational concept, which includes defensive force regulated by loac, is confused with the criminal law concept. Finally, clarifying the possibility for something akin to status-based targeting during all types of armed conflict may also have the added value of refining the understanding of the relationship between loac and ihrl, since human rights courts would be less likely to automatically apply the default behaviour criteria.12
nato, Military Decision on MC 362/1 – NATO Rules of Engagement [hereinafter MC 362/1], 30 June 2003. The MC 362/1 is nato unclassified, however, permission to use parts of the document for the purposes of this research is granted by the nato Military Commission in document imstam(O&P)-0006-2018 (copy on file with the author).
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 Dietrich Schindler and Jiri Toman, The Laws of Armed Conflicts (Martinus Nijhoff Publisher, 1988) pp. 711ff, Article 49(1).
North Atlantic Treaty, Washington, D.C, 4 April 1949 (
Emer de Vattel, Le droit des gens; ou, Principes de la loi naturelle appliqués à la conduite et aux affaires des nations et des souverains (The Law of Nations or the Principles of Natural Law Applied to the Conduct and to the Affairs of Nations and of Sovereigns) (first published 1758, translated by Charles G. Fenwick, The Classics of International Law, No. 4 vol. iii. Carnegie Institution of Washington, Washington D.C., 1916) Book iii, Chapter xv, § 231, at p. 230.
Erica L. Gaston, ‘Reconceptualizing Individual or Unit Self-Defense as a Combatant Privilege’, 8(2) Harvard National Security Journal 283 (2017), p.327.
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. 212.
For instance, during the UN Operations in Somalia ii (unosom ii), a ‘frag order’ was issued as a supplement to the roe, declaring “armed militias, technical vehicles, and crew served weapons as threats which could be “engaged without provocation””. Todd C. Huntley, ‘Balancing Self-Defense and Mission Accomplishment in International Intervention: Challenges in Drafting and Implementing Rules of Engagement’, 29 Maryland Journal of International Law 83 (2014), p. 99.
As Corn explains, “in the absence of clear policy guidance, self-defense authorities have frequently been invoked at the operational and tactical levels as the basis for targeting presumptive civilians taking direct part in hostilities”. Corn, ‘Public Authority to Use Force in Military Operations’ (n 6) p. 45.
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), p. 46.
W. Hays Parks, ‘Air War and the Law of War’, 32 Air Force Law Review 1 (1990), p. 118.
See also Todd C. Huntley, ‘Balancing Self-Defence and Mission Accomplishment’ (n 7) p. 107.
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. 203–205. Note that in internal niacs, human rights may impose a requirement to capture rather than kill if possible or reasonable, especially when the person is not a threat at the time. By further developing and defining the scope for applying status-based targeting during niac, such considerations may be better taken into account.