6.2.1 Self-defence and rights of assistance

Non-state agents engaged in armed resistance sometimes claim their use of violence to be justified directly by individual defensive rights. The Irish Republican Army, for instance, made claims of this kind regarding its defence of Catholics from violence by Protestant gangs in Belfast in the late 1960s and early 1970s. I discussed in Chapter 4 a similar justification offered by the Haganah for its code of ‘restraint’ in the face of attacks on Jews in Palestine during the Arab Revolt. In these circumstances, the evil they combated was identified as individual aggression.Footnote ⁸ The justificatory claim they made was one of rightful defensive assistance as third parties.

It is necessary to begin with the right of self-defence in order to identify important additional considerations that apply to rights of defensive assistance. In the account I discussed above in Chapter 3, David Rodin argues that the right of self-defence is best understood as a simple ‘full’ liberty possessed by the innocent victim of an unjustified attack.Footnote ⁹ As a full liberty, it gives to the victim a right (a ‘half liberty’), on the one hand, to defend herself forcefully when such an action is necessary to prevent an injury of equivalent magnitude from an attacker. But equally she has a right (a ‘half liberty’) to waive her defensive right and face an unimpeded attack. By contrast, the right to defend another person is not a ‘full’ but a ‘half liberty’ or ‘duty’. Rodin argues that the duty to assist in defence is not derived from the victim's right of self-defence but from ‘more general considerations concerning the duty to protect the good and the valuable’. In a similar vein to the Rawlsian account I drew on earlier in this book, he argues that the duty to defend others is strongest when the value at stake is high (a human life) and the risks associated with doing are comparatively low; but ‘when the risks to the subject [of the duty] are high, the chances of success doubtful, and the relationship between the subject and the end of the defensive action tenuous, the duty to act may become diminished until it is indistinguishable from a full liberty’ (though even without a specific duty of care towards the victim, the duty to help them may be ‘extremely strong’).Footnote ¹⁰ Following Rodin, then, we could argue that the duty to assist those under attack claimed by some armed resistance groups flows from a general duty of justice and bears on anyone in a position to help, subject to limits arising from reasonable partiality towards themselves which might prevent them having to sacrifice themselves for the sake of the victims.

I am generally sympathetic to this account but want to qualify it somewhat in light of some further possibilities affecting the relationship between the agent using violence and its beneficiary. Unless we nuance it further, the account so far would seem to oblige someone to engage in defensive attack on behalf of a victim of aggression even if they didn't want them to. So even were someone to forbear from defending themselves from unjust aggression based on deeply felt personal preferences (based on pacifist religious beliefs, for example), a suitably positioned third party might still simply be obliged – and therefore permitted – to engage in a defensive action on their behalf. We therefore arrive at the following: any suitably positioned individual or group could claim the right to intervene using armed force to assist individual members of a community that was threatened with wrongful violence. No consultation with the victims would seem to be required as the duty to assist occurs independently of the defensive rights of the victim.

It seems to me, however, that looking at things in this way would take insufficient account of the full range of relevant intuitions. One scenario that it does not consider is the possibility that the victims of aggression might sometimes claim the right to refuse assistance. If members of a particular community, for instance, did not wish to be represented by an armed group – a non-state organization wedded to a particular ideology, for instance – would they not have a right of refusal? Would a non-state organization not exceed its rights if it persisted in killing aggressors on behalf of individuals who had explicitly expressed a desire not to have such violence used in their name?Footnote ¹¹ I think we should sometimes be prepared to say yes to these questions based on a commitment both to human dignity and to democratic politics. The right to assist using defensive force, I argue, properly rests in part on the consent of its intended beneficiaries or, minimally, on a consultative input in some circumstances at least.Footnote ¹² I will therefore argue that the legitimacy of assistance does sometimes depend on the rights of victims and on whether or not they choose to exercise them.

For present purposes, then, a key problem on the account given so far is that rights of assistance seem to occur independently of victims and their preferences. The effect of this idea becomes clear when we try to account for them using Rodin's ‘three-legged’ analytical framework. Rodin analyzes the right in terms of three relations: first, the relation between the subject of the right and its end (i.e. respectively, the victim of aggression who defends herself and her life); second, the relation between the right's content (the action used) and its end; and finally that between the subject of the right and its object (the aggressor against whom force will be used). In the case of assistance, the subject of the right is a third party defender; the end, as before, is the victim's (right to) life. The perspective of the victim as an agent, however, has no independent term and the question of what choices she or he might wish to make is therefore entirely effaced. In effect, using this framework to analyze rights of assistance reduces a scenario involving three agents (a victim, a defender, and an aggressor) to a relation between only two of them (the defender and the aggressor). The victim appears only in a passive, objective light – present implicitly as the owner of a value constituting the end of someone else's moral action – and not as a moral agent in their own right. My suggestion, therefore, is that the theory of defensive assistance needs a fourth leg to register the relationship between the subject of the right of assistance (the defender) and the subject in whose name the action is taken (the victim of aggressive attack).

Taking this fourth relationship seriously reinserts the moral agency of the victim back into our account of rights of assistance. It demands that we ask whether the preferences of the victim of an aggressive threat can place any limit on the justifiability of defensive assistance. Respect for the victims, I will argue, requires that autonomous choice and expressed preferences be taken into account in cases where victims of aggression retain the ability to reflect and deliberate on their options.

6.2.2 Defensive assistance and consent

To develop an account of the role that victim choice might play, it is necessary to elaborate on three different ways in which a third party may be related to the victim of aggression. First, the two parties may be bonded already by a specific duty of care, such as that of parents for their children.Footnote ¹³ Thus, independently of a child's right to defend herself, she may claim the right to be protected by her mother, a claim mirrored in the mother's duty to protect her child. The second possibility is that of ‘rescue’. It is this kind of relationship, I take it, that obliged the Samaritan to rescue the man who had been attacked by thieves and, as Paul Ramsey once argued, the altruistic duty owed by the Samaritan might have extended beyond the nonviolent assistance described by Jesus had he met the victim during the attack.Footnote ¹⁴ To these two I will add a third, contrasting kind called ‘authorization’. Care and rescue properly obtain, I argue, only where for some reason the agency of the victim has been negated or is unavailable for consultation; the third connection is required in cases where a victim retains their ability to deliberate and communicate.

Neither of the first two cases necessarily requires any consultation with the victim: a parent, I presume, is entitled to defend her daughter even if the child had asked her not to. The duty is unaffected by questions of beneficiary consent or consultation, and one would feel morally ambivalent at best about parents who allowed their children to be killed out of respect for their current convictions. This type of relationship arises where the responsibility of the intervening third party is such that it overrides the autonomy of the victim, and, conversely, when the latter possesses insufficient force (due to infancy, for instance) to override the duty to assist. It is not a relationship that ordinarily occurs between two adults with full moral competence, unless one has a special, legitimate, professional duty of care (as a member of the police in a legitimate state, for instance).

I will restrict the term ‘rescue’ to denote circumstances in which victims are capable neither of self-defence nor of reflecting, deliberating, or communicating adequately concerning the question of assistance. Such would be the case where an attacker had overwhelmed her victim, perhaps knocking her unconscious, for instance. In such circumstances any right to decline an offer of assistance that the victim might ordinarily be able to claim would have been rendered meaningless since the relevant subjective preconditions would manifestly not obtain. There may also be room here to include those whose preferences have been radically adjusted, perhaps due to psychological violence or the constant fear of physical attack.Footnote ¹⁵ Differentiating between true and false preferences will prove difficult in practice. But situations undoubtedly do occur – in abusive personal relationships and under terroristic regimes, for instance – where victims are likely to refuse assistance, not based on an assertion of autonomy, but because they have already been rendered incapable of autonomous decision. In these situations, to withhold assistance on the basis of a refusal of consent would be wrong as the relation of rescue would obtain. I will say something about the peculiar moral problems that these kinds of situation present in Chapter 10.

Clarifying the nature of rescue and specific duties of care helps, in turn, to distinguish the contrasting case of ‘authorization’, signifying situations in which the victim is in a position to make an informed choice and to communicate it to a would-be defender. Not all personally or politically violent confrontations deprive their victims of the capacity for autonomous moral choice, and it is not appropriate to assume that victims who remain competent would inevitably be prepared to accept any available assistance, especially when it is offered in the form of violence. There are political as well as moral reasons why, as members of a community, individuals might sometimes prefer to accept increased risks of violent attack rather than endorse the interventions of a ‘friendly’ armed non-state organization to which I return in Section 6.4. But at this point it is important to say something about the reasons why an individual might reasonably refuse violent assistance in her own right in the face of potentially lethal aggression. This may help clarify – though it cannot, I think, decide in a thoroughly determinate way – the weighting that a third party should give to victim preferences.

The possibility for which I am arguing is that third parties might sometimes be obliged to value – or at least to consider valuing – a victim's autonomous choice over their right to life even if, without assistance, they are likely to die. It could be objected that this stance is self-contradictory: it would entail valuing one exercise of autonomous choice over the victim's total autonomy, which would be annihilated in a lethal attack. In response to this line of argument, it is necessary to compare briefly some different ways of grounding the value of autonomy. I will consider three approaches: the moral-rationalist, the hedonist or eudemonistic, and the pluralist. The first is unlikely to support the right to refuse assistance but the other two sometimes will.

On a moral-rationalist account, one's autonomy could be valued as the basis for freely choosing to uphold right and prevent injustice. The value of autonomy itself would therefore require that, where one had a reasonable opportunity to do so, one should try to prevent an unjustified attack from annihilating the autonomy of another person. To allow the killing to proceed would constitute a failure to make choices in conformity with the moral imperatives grounding the value of autonomy itself and it would most likely be wrong, therefore, to withhold defensive assistance even if the victim asked one to. The request itself could be regarded as inconsistent with autonomy as a value and it would therefore have insufficient force to prevent a third party from intervening to defend the autonomy of the victim as a whole. On a hedonistic or eudemonistic account, autonomy – or more properly, ‘freedom’ – is valued as the basis for choosing sources of pleasure and happiness over those of pain or unhappiness. On this view, it would be irrational to refuse assistance where it offered the only hope for continued enjoyment of life. Having said that, if one were predisposed to debilitating feelings of guilt and believed one could not live with the emotional consequences of the aggressor's death, then refusal could still be rational. Finally, the kind of pluralist view I have in mind sees autonomy as valuable above all for the opportunities it affords to try to realize particular – and plural – visions of the good. It is on this view that a third party would be most strongly compelled to consult the wishes of the victim in deciding whether or not to act. For on this view, a person's life and their autonomy itself are not valued on opportunities for pleasure maximization, but on the moral, religious, or other principles they have chosen to try to realize. If, say, a person had adopted Quaker doctrines and valued their being and their capacity for moral choice based on the ability to realize the pacifist commitments these entail, then it might well be that to save them from unjust aggression using lethal violence would be self-contradictory. Extending their life through violence might negate the value of their autonomy instead of helping to realize it.

I do not want to suggest, on this basis, that third parties must always choose to respect the victims’ wishes but to argue that some consultation is appropriate.Footnote ¹⁶ This is well seen in an illustration suggested by Jeff McMahan: if a mother were under potentially lethal attack from her own child and a third party could save her life only by killing the attacker, what should they do? And what weight should be given to the preferences of the mother?Footnote ¹⁷ A case such as this is probably too ‘hard’ to legislate across the board, but it underlines the importance that victim input could have as a part of decision-making where circumstances permit. One could certainly argue here for victim-consultation under either the hedonist–eudemonistic or the pluralist framework, and possibly even on the moral–rationalist view; and the very indeterminateness of the choice between moral frameworks themselves reinforces the point. Deciding which is the relevant moral evaluation of the scenario is a matter that rightfully concerns both the defender and the victim of aggression. It is ultimately because of this fundamental indeterminacy in deciding which moral values and which perspective to prioritize that it is so important to take account of the victim's preferences. If they have strong objections, it means that violent actions taken on their behalf will be less clearly justified. Whether the objections of the victim entirely negate the justification, however, depends on the moral theory we adopt and the conception of the good followed by the victim.

The actions of third parties, on this view, who proceed with a violent intervention in the face of a refusal of authorization lack legitimacy in relation to the victim-beneficiary (that is, the victim of aggression and intended beneficiary of defensive actions). Thus, even though their use of force may not do any wrong to the aggressor, it might constitute an injustice to its supposed beneficiary.Footnote ¹⁸

6.2.3 Paramilitaries and legitimate defence

The armed campaigns of the Provisional IRA in Northern Ireland provide a useful example with which to illustrate the criteria we can now apply. During 1969 and 1970, sectarian attacks on Catholic housing estates in Belfast by ‘loyalist’ gangs met with armed response from IRA volunteers.Footnote ¹⁹ These events have since been presented as the basis for arguing that a vital raison d’être for the Provisional IRA was to defend Northern Irish Catholics from armed attack.

In what circumstances could the justificatory possibilities opened up by an argument from defensive rights cover these aims? Armed groups could claim rights of assistance in two ways. The first is on the basis of ‘rescue’. If we imagine a scenario like the one in Belfast where a local community – the people living on a particular street, for instance – came under sudden and unprovoked attack by sectarian gangs or terrorists from another area, they might initially be in no position either to deliberate collectively or to articulate a shared view on the desirability of armed assistance. If an armed group acting in good faith were to step in and defend civilians in the area from such attacks, it would seem excessive to ask that it demonstrate explicit endorsement. The second possibility would arise in situations where the victims of aggression were in a position to deliberate. If they authorized defensive force in these circumstances, or at least allowed it to take place without objecting, then it would positively validate the legitimacy of armed action and the representativeness of the group. Though, of course, the ability to make a choice may be adversely affected by the presence of the ‘friendly’ armed group if it appears prepared to coerce ‘its own people’, so to speak.Footnote ²⁰

It is necessary to state that, as it stands, this justification – of purely defensive violence – could not be made to stretch beyond actions responding to an immediate threat, as the limiting clauses of necessity, proportionality, and imminence make clear.Footnote ²¹ For the Provisional IRA to justify its sustained use of ‘armed struggle’ during subsequent decades it was necessary to make a larger and more difficult justificatory claim than one of rescue, namely, that it represented a particular community (variously Ireland as a whole, the Catholics of Northern Ireland, or Irish nationalists as a whole) and was entitled to wage war on its behalf. This would take it from individual self- and other-defence to collectively organized, offensive war (and from Purely Defensive Violence to the rules governing Organized Offensive Violence). I will turn to this kind of justificatory framework in Section 6.3 below.

The principle of authorization presents an important limit to the justification of defensive violence by non-state entities. Where members of an armed force present themselves as the defenders of competent individual victims who have expressly rejected their authority to do so, we can question their right to resort to arms on their behalf. We can assert this in the first instance on the basis of human dignity and the victims’ entitlement to be involved in important moral decisions taken in the name of their rights. Were IRA gunmen to continue armed action in the face of opposition, then, we could say that while the aggressors from loyalist gangs had no grounds for complaint, having, as it were, made themselves liable by their own actions, the Catholic citizens who were the unwilling ‘beneficiaries’ of defensive violence would have grounds to complain against the IRA as well as against the loyalists. Where a group like the IRA used its success in civilian defence as a basis for claiming representative legitimacy, it is clear that the complaint could have important political ramifications for the communities on whose behalf it claims to act.Footnote ²²

6.3.1 Armed resistance and unjustified criminal violence

If armed robbers emerged from a bank to find themselves confronted by members of the security forces and then tried to shoot their way to freedom, what combat rights could they claim and what rights and duties would be claimed by their adversaries? The answer, presumably, is that the security forces could claim targeting rights against the robbers, that is, the right to use proportionate force up to and including lethal force, until the threat was eliminated. This means that a principle of discrimination applies to their actions (though not, of course, that of the Standard JIB): the security forces are, on the one hand, prohibited from directing their force at innocent parties, that is, non-participants in the combat; on the other hand, they are permitted to direct force against the robbers as participants. By contrast, the robbers have no such permission. While also bound by the prohibition on targeting bystanders, they are equally bound by a moral and legal prohibition on targeting the security forces. In this sense, we might say, they are subject to the usual rules of peacetime morality where all individuals – civilians, police, and soldiers alike – have moral immunity from attack unless they engage in unjustified aggression against others, as the robbers have done in this illustration. This renders the culprits morally non-immune to defensive actions taken against them as long as they persist in posing a threat.

There are two things to note about the conflict between robbers and security forces. First, the distribution of rights, duties, and prohibitions does not depend on a convention such as the Standard JIB. The right to kill, in the scenario described, arises temporarily on the basis of rights of defence against an immediate threat and lasts only as long as the threat does. Second, there is no ‘moral equality’, as just war theorists call it, between participants in the combat. In fact, we must resist calling them ‘combatants’ here, since that word carries a technical meaning in just war theory, the Law of Armed Conflict (LOAC), and the Standard JIB, that implies a right to use force as well as a liability to attack in the context of war. There are no ‘combatants’ as such in this situation, strictly speaking, since the security forces are not liable to attack and the robbers lack the permissions that combatant status implies.

Non-state entities that claim their fight is justified as a legitimate form of resistance seek a different distribution of rights, duties, and prohibitions from the one arising for the bank robbers. They can either try to claim for their members full combatant status in the technical sense and ascribe to all participants in the conflict the right to use military force against their armed opposite numbers (the Standard JIB); or they can claim that the right to use force rests exclusively with members of the non-state entity while the use of force by the official military is illegitimate (the Partisan JIB, the subject of Chapter 8 below). In the first case, soldiers of both the rebel forces and those of the state possess targeting rights; by the same token, as combatants they can all legitimately be targeted. In the second, only the rebels have the right to use violence and only against combatants and some others who contribute in important ways to regime violence.

One way or the other, then, just rebels claim to have a very different status from armed bank robbers, one that has an important salience in determining, first, their rights as combatants, and second, the moral status of their actions and armed campaigns. To gain greater clarity on the way rebels could justify a claim to this status, it is necessary next to ask how just war theory tries to justify the killing of soldiers in the context of a war initiated by states.

6.3.2 Legitimate authority and the War Convention

I have already discussed in Section 4.3.1 the difficulty that a traditional account of the Standard JIB faces in trying to account for its central principles by direct appeal to ordinary morality. If we turn to the alternative, conventionalist justification for these principles, I now want to argue, we will find that a link is silently made, in this view, between the jus in bello and the jus ad bellum, grounding the liability of combatants. This link implies a particular notion of legitimate authority that is important to a full understanding of the application of the Standard JIB in wars of resistance.

Let's say a state declares and initiates war against its neighbour to seize territory or natural resources without due title. We would say of this war, first, that it lacked a just cause and, in fact, pursues an unjust cause, legally and morally. Conceivably, another way to express the same point would be to say that the aggressive state did not ‘have legitimate authority’ to wage a war of this kind. In doing so, we would be invoking a notion of legitimate authority that would only come into effect once a state has established just cause for war. On this usage, in fact, the legitimate authority criterion does little or no work of its own. Where we say that an entity possesses legitimate authority for war in a particular situation, it reflects a judgement that it has fulfilled the other jus ad bellum criteria. ‘Legitimate authority’ on this reading is, therefore, more or less synonymous with the jus ad bellum as a whole.

Traditionally, however, the principle of legitimate authority does some extra work based on the substantive doctrine that ascribes the jus belli only to rulers or states. In this role, the legitimate authority criterion is largely negative, excluding non-states from the right to declare war even if they fulfil the requirements of just cause, Proportionality, last resort, and so on. This substantive doctrine is, of course, no longer widely accepted in its pure form either in just war theory or international law.Footnote ²⁴ Conventionalist accounts of just war theory, however, implicitly invoke an authority criterion in a related but distinct way. The initiation of hostilities by an aggressive state does entail some adjustment in the distribution of individual rights which, once we reject the traditional account, cannot entirely be due to moral forfeiture. In particular, it makes it possible for individual soldiers on the aggressive side to use military violence against soldiers defending the victim state without fear of criminal punishment. Despite the fact that it was not entitled to declare the war it has initiated, once it has done so, the aggressive state is implicitly empowered to invoke the terms of the law of war, that is, to create a state of war and bring to bear the rules governing such a condition, the Standard JIB. Within this set of rules, their combatants then have the same privileges, duties, and liabilities as their opponents. States, therefore, appear to have two forms of legitimate authority: first, the moral right to declare a just war justly (i.e. when the other terms of the jus ad bellum are fulfilled); second, the purely conventional right to create a state of war even, apparently, in the absence of a just cause or despite a failure to fulfil the other terms of the jus ad bellum. Legitimate authority in this latter sense – which I will call ‘Lesser Authority’ – thus appears as an intermediate right: it arises implicitly in relation to the jus in bello but is also relevant to the jus ad bellum (and thus connects the two).

Let me add three points of clarification regarding states in particular. First, the principle of Lesser Authority does not affect the permissiveness of just war theory as regards the number or kinds of wars that can justifiably be declared. Justification for each war as a whole is still limited by the scope of just cause and the other jus ad bellum principles as usually understood. What it does is underpin the legal privilege that all combatants can claim to use force against opposing combatants, while reinforcing thereby the ban on directing violence intentionally against non-combatants. Second, this kind of authority does not diminish the burden of responsibility weighing on the leaders of states. On the contrary, it increases it, underlining their responsibility for the killings carried out by soldiers acting under their orders. If the war is just, fulfilling all the terms of the jus ad bellum, then it makes little difference since much of the violence deployed by the soldiers will be morally justified whether we regard it as their violence or as violence for which their leaders are responsible. But if a war is unjust – if, say, it served only the private interests of a ruling caste within the state – then the soldiers’ violence wouldn't be justified by the ends of the war. In that case, Lesser Authority effectively places the total burden of legal responsibility on the leaders themselves while treating the soldiers as if they are not guilty of any crime (provided they observe the constraints of the jus in bello).Footnote ²⁵ Finally, many wars are initiated by political leaders whose ability to rule is based on violence and usurpation rather than on a valid claim of legitimate political authority. In such cases, attributing Lesser Authority does not affect the usual implications of the leaders’ illegitimacy: it does not grant them a right to act on behalf of the state and its people; and it does not diminish their liability to the use of such means as are available to remove them from power, either by the people they rule or by external powers where the domestic violation of human rights is grievous enough to justify humanitarian intervention. It merely reflects their ability de facto to order soldiers into war and manages some of its consequences through the legal principles of discrimination and combatant equality.

The salience of Lesser Authority in the LOAC may be quite dramatically in tension with the morality of conflict at its deepest levels. But, as we have seen, there is a morally pragmatic case for maintaining combatant equality as a principle generally governing the conduct of hostilities between states. A conventionalist approach therefore supports the tendency of international law insofar as the latter enshrines the principle of combatant equality and, with it, the rights of combatants to target opposing combatants regardless of the validity of their own state's claims about just cause and the other jus ad bellum criteria. The question is whether the same logic applies to non-state actors, that is, should there be a general presumption of Lesser Authority for politically motivated, armed non-state parties, just as there is for states?Footnote ²⁶

6.3.3 Non-state entities and moral equality

My view is that, though they might sometimes be able to justify a claim to possess the authority to initiate war, non-state entities do not possess Lesser Authority by default as states conventionally do.

Imagine how it would be if a legitimate state were bound to recognize as a belligerent entity with the corresponding privileges, any group of persons under its jurisdiction that declared itself at war with the government or some other group. Some such group declares war one week, killing a number of soldiers (or members of rival groups) the next, and then, a week after that, declares their war to be complete. Had any of them been captured, they would have been entitled to prisoner-of-war status for the weeks of the war and to be released afterwards without charge provided their actions during wartime were consistent with the Standard JIB. And those that remained at large by the end of the conflict would be immune from arrest for any part they played in the discriminate prosecution of their ‘war’. Such an image of law and order within the jurisdiction of the state is clearly unattractive and troubling, and I doubt anyone would wish to defend it. Part of its peculiarity is the fact that, regardless of whether there was any justice in its cause, any private party taking up arms and declaring itself a political organization would be empowered by such an arrangement to render state soldiers who are otherwise at peace to be legally liable to attack or, at least, it would render the attackers impunible under law. Yet such is the actual arrangement in the international society of states. If we substitute for the non-state group above a state and imagine it invading its neighbour without any moral justification, its combatants do generally gain the privileges of war, and any discriminate violence used within the period of conflict is treated as permissible or at least impunible. If this is so counter-intuitive within states, then why is it so widely accepted between them and are there sufficient reasons to justify maintaining the arrangement between states while rejecting it as the default arrangement for sub-state organizations?

Let's begin by looking at how things operate within states. I presume it is reasonable to suppose that deliberate attempts by private individuals to cause severe harms to their fellow citizens ought to be prohibited by law and subject to punishment except in cases of legitimate defence against wrongful attack (including justified armed resistance, per Chapter 3 above). States, as I argue in Chapter 2, have a duty to protect all citizens’ Life and Limb Rights and, relatedly, to punish and deter violations thereof. The first-best way to address such threats, presumably, is through policing and criminal justice. Some non-state political movements commit acts of violence without justification (that is, either they are entirely unjustified throughout the entire range of their activities or so much of their activity is unjustified that it vitiates any claim they have to legitimacy or justification). Sometimes they unjustifiably attack and severely harm members of the armed forces or police. And some states have sufficiently legitimate police forces to be able to manage such threats in a discriminating and effective way and good enough criminal justice systems to be able to impose suitable punishments. Where these features are present, then, it seems desirable that those engaging in unjustified violence be subject to criminal justice and policing as far as possible.

In the affairs of states in international society too it might also be true that an approach to inter-state aggression analogous to the domestic criminal justice approach should be regarded as the ideal, first-best solution. Not only should states and their political leaders be held accountable under international criminal law for international aggression, war crimes, and so on, but ideally so should those who culpably perpetrate even ‘discriminate’ but morally unjustified acts of violence on their behalf, that is, soldiers involved in aggressive wars. So, at least in ideal circumstances, the first-best framework for responding to violent aggression in both domestic and international affairs might be law-enforcement, police action, and criminal law.

This approach is likely to cause problems, however, once we take into account the sorts of incentive it might generate and the consequences these could give rise to. At both levels – domestically and internationally – there are ideally reasons to fear that, by treating all unjustified killings equally as criminal acts of murder whether perpetrated against soldiers or civilians, the criminal justice approach may fail to provide sufficient incentive for discrimination in the use of violence. By contrast, a second-best approach in which the immunity of those tasked with combating with armed force the wrongful violence of aggressors is traded off to reinforce the immunity of non-combatants is likely to incentivize discrimination and help manage and mitigate the destructiveness of war.

So, on the face of things, it might seem that the same approaches are ideally appropriate, whichever the level we operate at, international or domestic, but there are reasons to think the case for a second-best approach is more widely and generally applicable in the international realm than in the domestic. First, whereas some states have reasonably just institutions, providing an authoritative basis for condemning the violence of illegitimate non-state political organizations, similar institutions do not (yet) exist in international society. Some elements of a criminal justice system do exist at the international level, but they are not sufficient; in particular, they cannot yet be enforced fully and adequately in the way a stable state enforces law. This lack is part of the incentives problem. Unjustly oppressive states also lack an authority of this sort, of course, but it can't be described as a regular feature of all states so the asymmetry remains between domestic and international societies. Secondly, there is also a matter of scale. Although states sometimes initiate violent actions that fall short of the scale of war (and which might be dealt with through criminal justice if agents are captured in foreign states), they nevertheless have a fairly consistent capacity, from one to another, to initiate armed conflict on a warlike scale. Non-state parties, by contrast, sometimes do and sometimes don't. They range from lone actors like the Unabomber through small organizations with limited membership and capability like the German RAF to belligerent sides in civil wars (even across more than one state in the case of ISIS) with large, well-organized armies. So there are differences, I think, between states and non-state groups that could explain why a uniformly second-best approach is the most reasonable choice at the international level but not at a domestic level.

By contrast, establishing a convention giving Lesser Authority to any and all non-state groups with declared political aims would give rise to unattractive consequences sufficient, I think, to counter-balance any likely benefits. First and foremost, while the ability to invoke the Standard JIB unilaterally might encourage discrimination in favour of civilians by non-state actors, it would also give any individual or gang with political motives, however spurious, a blanket mandate to attack and kill members of state security forces with legal impunity.Footnote ²⁷ Second, if arrested, individuals with a claim to political status as members of non-state entities would have to be treated as prisoners of war while their ‘war’ persisted and they would have to be released if their organization came to terms of peace with its enemy. In effect, the two factors would combine to make it impossible to bring a criminal charge of murder against anyone who had killed a member of the security forces as long as they could demonstrate political motivation.Footnote ²⁸ Killing of this kind would cease to be a criminal offence. A presumption of legitimate authority for all politically motivated non-state entities would thus efface the distinction we need to make, between groups whose killing of military personnel or police belongs in the criminal category of ‘murder’, and those which can justifiably describe otherwise similar actions as the discriminate targeting of ‘combatants’.

International law does, however, recognize some types of case where non-state groups are granted recognition as belligerent sides. Full-scale civil wars, for instance, and wars of national liberation historically raised the question of whether applying a principle of discrimination and combatant equality could offset the impact of sub-state conflicts on civilians. The resulting First and Second Protocols to the Geneva Conventions (1977) provide a basis for regulating such conflicts once members of the non-state forces are recognized as combatants fighting as part of a belligerent side. This reflects the idea that, once a non-state group achieves the level of organization, territorial control, and capacity for effective armed action that is commonly presumed with states, its status might thus be recognized for prudential reasons similar to those applying to states.Footnote ²⁹

However, having the power to force enemies or the international community into recognizing a state of war isn't the same as having a moral right to claim such recognition and the privileges of combatants. This still leaves open the question of which conditions need to be fulfilled before, say, a non-state group fighting for national self-determination could oblige others to interpret and evaluate its soldiers’ actions within the framework of the Standard JIB. Of particular concern for present purposes are non-state entities whose success, power, or prospects fall short of those that could force other parties to recognize belligerency but which might in some cases be justified in using armed force. For these, there can be no a priori presumption of a Lesser Authority applying to particular cases since there is neither a legal nor a prudential basis for doing so in general. The consequence is that if particular non-state groups are sometimes to be able to invoke a state of war and claim legitimate right to denominate and target ‘enemy combatants’, then they will have to follow the pattern of the first, more stringent sense of ‘legitimate authority’ as it applies to states: that is, to claim any kind of legitimate authority to declare war, non-state groups would first of all have to fulfil the same or equivalent conditions as would be necessary for a state to declare a just war justly. Where a non-state organization initiated hostilities without fulfilment of these conditions, then its war would be unjust not only as a whole, but also in each of the actions carried out in its prosecution. Whereas states that initiate unjust wars of this kind can still effectively authorize their soldiers to use discriminate force under conventional and legal rules (however problematic it may appear to be from a deeper moral perspective), non-state entities cannot. The latter is not only true as a matter of political fact, but my view is that it is right as a matter of morality: non-state groups should not be granted this power indiscriminately. Where a non-state entity initiates war without fulfilment of the jus ad bellum, those individuals who fight on its behalf cannot, therefore, distinguish their combat from that of a purely private interest such as the bank robbers considered earlier. Unless they exceed a large scale, they cannot force others to grant them recognition; and unless they have justice on their side, they cannot claim it as a moral right. As a result, all individuals whom they target – whether soldiers or civilians – will qualify as morally immune and ‘non-combatant’ both legally and morally.

6.4.1 Legitimacy and just cause

I think the legitimate authority of a non-state organization requires some form of representative legitimacy in the community on whose behalf a campaign of offensive armed force is proposed. There are two sorts of reason for this. One is substantive and relates directly to the ‘just cause’ criterion as well as to legitimate authority. It occurs particularly in relation to political causes justified in terms of self-determination, in relation, for instance, to the IRA, in its various forms and phases, the Algerian FLN, the PLO, ETA, the Tamil Tigers, and so on, all of which have waged ‘war’ at least partly on behalf of claims about the self-determination of ethno-national communities. All such justifications for armed struggle presuppose an identity between self-determination and the desire for national independence. But this claim would be falsified if most members of the community in question preferred to remain part of a larger political entity. The force of the claim would also be diminished if the community were divided on the question of independence.Footnote ³⁰

As it stands, this is a matter, not of legitimate authority, but of just cause. It would be meaningless for a non-state organization to claim that national autonomy for a group, for instance, was a just cause on the basis of self-determination if the community on whose behalf the claim were made did not endorse it. Without endorsement, such causes might actually negate self-determination. The matter of legitimate authority arises, here, in two senses. First, we might use the term to describe the non-state organization's jus belli based on the presence of a just cause, that is, ‘legitimate authority’ might express in truncated form the proposition that the organization in question ‘has legitimate authority to wage war based on such-and-such a cause’. This, however, would add nothing to just cause as a criterion and any question of legitimacy would be reduced entirely to a relationship between the just cause itself and the community on whose behalf it is pursued (i.e. a question of whether the cause had legitimacy in the community). There is, however, a second sense in which legitimate authority may have something to add here. This would arise on the basis of an empirical supposition that adequate appraisal of a community's will and of the strength of its desire for national independence, hence the justice of the ‘cause’, is more likely to occur when the organization making the judgement is able to monitor – and perhaps shape and lead – opinion in the community. ‘Legitimate authority’ would then be useful as an independent criterion measuring the representativeness of the organization vis-à-vis the community. In this form, the ‘authority’ or ‘legitimacy’ of a non-state organization would be an index of its relationship, with the cause mediated through the will of the community on whose behalf it claims to act.Footnote ³¹

There are, however, some further qualifications that must be added to this framework, and these arise through consideration of the second set of reasons why an independent criterion of ‘legitimate authority’ is necessary.

6.4.2 Contingency and the right not to resist

The second set of reasons arises from important empirical concerns about contingency and unpredictability. There are two different ways in which the matter of representative legitimacy and legitimate authority arise here. One argument is supplied by John Stuart Mill's influential essay, ‘A Few Words on Non-Intervention’. Mill's concern was partly analogous to the present one: he asked in what circumstances it was permissible for a third party to the struggle between a people and its government to intervene and assist in the pursuit of ‘liberty’. Mill's third parties are other states rather than domestic political actors, but the risks they face in intervening are similar and may arise from the same cause. Mill's assumption is that representative self-government as a political goal cannot be realized without both a sufficient desire for liberty and certain well-established virtues and habitual practices. The intended beneficiaries of liberation, in other words, must be prepared in important respects if it is really to benefit them. If the desire for self-government or the virtues needed to succeed in it are not present in sufficient force, then either an attempted liberation will fail in its immediate military objectives due to a lack of popular support during the war or, if its war succeeds, the political outcomes it sought to achieve will fail due to the lack of popular support or participation after the war.Footnote ³² Thus even where a cause has some justice in principle – even, in fact, if the population were at least not unwilling to see it achieved, though not enthusiastic either – the claim to possess the jus belli would fail on grounds of reasonable prospect of Success and, hence, Proportionality. Once again, then, there would be an important independent role for a criterion of legitimate authority in evaluating the jus belli as a whole: a non-state organization genuinely representative of popular values and desires and capable of an adequate evaluation of the condition of the population in general would be well placed to judge its ability to sustain a war of liberation and support the institutions at which it aims.

Mill's argument thus emphasizes one kind of contingency that could affect the rightfulness of a decision to engage in warfare. Perhaps the most influential authority on violence and contingency, however, is Hannah Arendt, and it is particularly some of her observations that suggest the last and probably the most compelling reason why legitimate authority must remain important as an independent criterion for evaluating the jus belli as a whole of armed, political, non-state organizations. Arendt emphasizes two features of violence in its relations with politics that are important for present purposes. First she says it is inherently unpredictable. For this reason, she maintains, only the most proximate, short-term goals can sensibly be pursued by violent means because any more distant aims would probably be thwarted by the means employed.Footnote ³³ It is hard to control violence and therefore impossible to ensure that it will perform as intended. Second, Arendt warns that, far from simply serving ends slavishly as an instrument, violence tends to alter in more or less radical ways the political situations in which it was initiated and may thus displace or vitiate the purposes for which it was initially intended as a means. Patricia Owens calls this the ‘generative’ tendency of war.Footnote ³⁴ As such, resorting to violence in the service of one cause may generate further causes that in turn justify – or appear to justify – further violence. Arendt's warnings are salutary for any careful consideration of the jus belli, in particular, concrete contexts. Expressed in the terms of just war theory, their importance can be seen in relation to applying the principles of reasonable prospect of Success and Necessity in judging the occasions for war both of which form important parts of the evaluation of Proportionality analyzed in Chapter 5.

The element of contingency and the degree of uncertainty inevitable in making estimates of Success, Necessity, and Proportionality in concrete contexts establish a degree of indeterminacy in the application of moral principles within which consultation with those who are supposed to benefit from an armed campaign is essential.Footnote ³⁵ While a case might be made at a particular point in time during the development of a resistance movement that the resort to Organized Offensive Violence is the most proportionate alternative available, some of its intended beneficiaries may still feel that it is better to wait and see if the prospects for peaceful transition improve later on. Given the kinds of escalating risk that are associated with engagement in Organized Offensive Violence, they might therefore insist on a particularly stringent version of ‘Last Resort’ as it has traditionally been understood, that is, that violence should be undertaken only once all nonviolent alternatives have been exhausted. While the traditional understanding of Last Resort tries to draw a line defining cases where a belligerent has initiated war too early, they could argue, it does not prohibit waiting till later. Again, therefore, as there is no way to legislate in advance in a thoroughly determinate way concerning the point at which particular causes need to give up on nonviolent means, here is a contingency which would require inputs from all parties concerned. Given the indeterminacy of the principles and the unpredictable nature both of nonviolent politics and violent means, the possibility that a given community or a majority of its members might choose to await a peaceful opportunity for progress is in many cases unlikely to appear irrational. Where their preference is for nonviolence, then, the jus belli of a non-state entity initiating violence without authorization can be challenged as a whole on a criterion of legitimate authority grounded in representative legitimacy.

When we consider these principles together, what they point towards is the fact that few if any concrete political situations in which a resort to Organized Offensive Violence might be justifiable offer clear strategic signposting. In cases of aggression but a fortiori in cases of violent political oppression, it has to be assumed that indeterminacy, contingency, unpredictability, and choice in deciding how to respond will be the rule rather than the exception. That being the case, it ought to be up to those whose interests are most directly at stake to judge in each particular case when reasonable hope of avoiding full-scale war can be abandoned and when the enormous risks involved in recourse to Organized Offensive Violence ought to be taken.Footnote ³⁶ The dilemmas faced by participants cannot be answered in advance but require the conscientious engagement of the intended beneficiaries themselves. It is ultimately on this basis more than any other that the importance of legitimacy rests in evaluating the authority of non-state entities in the decision to wage war. The agent in question claims to act on behalf of a community and its members, but it can seldom provide demonstrable assurance to those members either that all hope is to be abandoned of future nonviolent amelioration or that recourse to violence will achieve a satisfactory result, that is, one that is successful in serving the cause and proportionate in the harms it inflicts. A commitment to dignity would dictate, in such circumstances, that all or as many as possible of those whose rights and interests are involved in the decision should participate in making it.Footnote ³⁷

Non-state groups considering recourse to violence should therefore consult where possible with those they claim to represent and should seek wide endorsement in order to legitimate their programmes for action. (I will have more to say about the content of such consultation below in Section 7.3.2.) In contexts where democratic institutions are available, this will mean that organizations – like the IRA and its counterpart, Sinn Fein, in Northern Ireland – should be able to demonstrate widespread electoral support in the relevant population. Where other parties offering nonviolent political alternatives are able to enlist significantly greater support, however, the legitimacy of resorting to violence will be greatly diminished if not negated entirely.Footnote ³⁸ Where elections are not available then, as McPherson suggests, other ‘credible measures of approval’ may provide indications of the representativeness of a non-state entity and its actions: ‘mass demonstrations, general strikes, and polling’, can provide some indication of support for a group's policies, though onlookers should remain wary of engineering by the group itself.Footnote ³⁹

6.5.1 Legitimate authority and armed groups without just cause

It is highly likely that some groups will be found to have wide support in the communities they claim to represent while pursuing a cause that is unjust from the liberal-republican perspective that informs the present argument. Such groups may be said to have legitimate authority of a kind while lacking just cause for war. The first thing to say about them, of course, is that they lack moral justification for resorting to arms regardless of authorization. Killing for the sake of whatever other goals they may pursue is wrong with or without the authorization of its beneficiaries. If wide support in a community is often sufficient to undergird a widely destructive capability in such a group, however, to such an extent that it can create a de facto state of war unilaterally, it may force opposing states to trade off the right to punish perpetrators of wrongful killing against the restraint that could be purchased by recognizing belligerency and invoking a principle of combatant/non-combatant discrimination. In such cases second-best rules would apply: fighters for the non-state force would be granted the war privilege in exchange for discriminate fighting. In principle, however, it remains a matter for the discretion of the enemy state and the international community whether to recognize such groups as belligerents in armed conflict.

Non-state groups with neither just cause nor authority have no right or ability to invoke the Standard JIB unilaterally, but in cases where they become large and powerful enough, there may yet be a prudential case for recognizing them.

6.5.2 Legitimate authority and groups with just cause

If a non-state resistance movement has both just cause and legitimate authority, then it is not necessary to gain control of territory or otherwise force the state into recognition by virtue of strength and war-making capacity in order to have a moral claim to such recognition.Footnote ⁴⁰ Groups with both are able to justify a unilateral moral claim to recognition against the enemy and against the international community.

The most complex cases are those where there is a just cause for armed resistance and, a fortiori, there is a case for resorting to Organized Offensive Violence based on considerations of Proportionality and Necessity, but where those seeking to lead the resistance lack support among the beneficiaries. There are at least three variants within this category.

1. (1) Rescue: if the group had just cause based on the need to combat a violently oppressive regime (i.e. the type considered in Section 3.5.1, where the regime engaged widely in Life and Limb Rights₁ violations prior to resistance) and if the conditions necessary for a relationship of ‘rescue’ as outlined in this chapter were present, then it could justify initiating proportionate offensive, armed resistance even without support. As such, it has a moral claim to recognition as a legitimate belligerent. I think that a possibility of this kind is imaginable, but only in some severe circumstances. In cases where the terroristic nature of political rule compromised subjects’ ability to engage in meaningful deliberation about their interests and political prospects, then authorization would be less important (though not entirely irrelevant). A predicament of this kind might be suffered by those living under totalitarian regimes as characterized by Arendt, for instance, and more recently by Samuel Scheffler and Jeremy Waldron. Waldron in particular seeks to characterize the mental state that arises when ‘terrorization’ removes the capacity for deliberation upon which even coercion normally relies.Footnote ⁴¹ Individuals suffering from this kind of abuse will be incapable of the kinds of input through which normal legitimacy requirements are fulfilled. The Arendtian analysis on which Scheffler draws offers a way of imagining too how the communal or social bonds necessary for a community to be able to achieve solidarity in the face of oppression and to authorize leaders to act on its behalf could be rendered impossible by the ubiquitous intrusions of the state through informers or the secret police.Footnote ⁴² In such circumstances, then, we can accept that if a non-state group were available that could defend the fundamental rights of victims through discriminate violence then it would have a freer hand in doing so, so far as victim consent is concerned. The situations in which this could occur would be analogous to domestic relationships in which the psychological damage inflicted by severe or persistent abuse was such that its victims were incapable of leaving the situation even if offered the chance to do so. As with the victim of domestic abuse, a wider population subject to terroristic rule would have an interest in liberation clear at least to outsiders, albeit one that its members might disavow owing to fear of reprisals. Persistent threat might even lead victims to adapt their preferences. Initiating violence in such circumstances, however, is likely to involve strategic difficulties, raising further moral questions that I will address in greater detail in Chapter 10.

2. (2) Where Authorization Is Possible but Withheld: in the second variant, the resistance movement has just cause of the same sort as in the first, but the conditions of rescue don't apply. In such cases, if there is a compelling evidence-relative case for armed resistance, that is, if it has a high enough chance of achieving a sufficient degree of success to counter-balance the expected risks and costs of pursuing it, then it might be that the leadership could overrule popular dissent from the plan. However, I think it is unlikely that this would happen often since (a) such matters are generally too hazardous to be so certain; (b) for this reason, popular misgivings about whether the risks are worth facing (given, especially, the chance of making things worse by resisting) must be taken seriously; and (c) lack of support is likely, in any case, to reduce chance of success. Reason (b) reflects the concern that aspiring leaders of an emerging non-oppressive democratic state ought to have with respecting the preferences and concerns of the beneficiaries. On the other hand, if the lack of support for war is based on reasonable concerns that beneficiaries have with prospects of Success or Necessity, that is, indeterminacies in the application of these principles that require judgement, then it would diminish the scale of permissible armed resistance that the leaders could justifiably initiate. A narrower war might be justifiable on this basis, its breadth matching the breadth of support and the risks and costs undertaken being scaled down according to the proportion of the population willing to bear them. But such a war is likely to be less effective in resisting and therefore armed resistance is less likely to be proportionate overall. In some cases, therefore, the leaders might be deprived of a jus belli altogether. Although they might be justified prima facie in fighting on their own behalf, the fact that doing so will impose heavy burdens on others that those others aren't under a duty to bear (given the uncertainties about victory etc.) means they won't have an all-things-considered justification.

3. (3) Authorization and Wars for Political Rights: finally, in Chapter 3 (Section 3.5.4) I discussed regimes that are not ‘violently oppressive’ but that are likely to engage in ‘repressive’ violence if resisted. Such regimes might be oppressive in some respects without being grievously so, or they might be resisted by groups representing aspirations for national self-determination or other causes whose legitimacy is based in widespread popular support. Where there is sufficient support for pursuing Political Rights and a willingness to face the risks, resistance may be justifiable; and if it faces unjust repressive violence, so too – in turn – is armed force. But by the same token a lack of authorization would have greatest effect in cases of this kind where there is no prior, compelling reason to initiate a form of resistance that will inevitably cost lives and limbs among the beneficiaries themselves. They should surely have a veto on this.

In cases like (3), as well as in those discussed in Section 6.5.1, where the non-state side has not been recognized as a belligerent, it is arguable that attacks waged even against soldiers can be described as ‘terrorist’ on one version of the just war definition of the word. If, that is, ‘terrorism’ refers to attacks on non-combatants, and if ‘combatant’ status is valid only in a legitimate state of war, then both soldiers and civilians in such cases are ‘non-combatants’, strictly speaking. Soldiers might, of course, use force defensively in such cases, but this would be no different from the use of force by police officers in the bank robber example earlier in the chapter: defending yourself does not make you a combatant in the legal sense. Of course, guerrillas fighting on the non-state side might be able to justify violence in defence of the innocent on some occasions (and it wouldn't be ‘terrorist’ on any usual understanding of the standard definition). Organized Offensive Violence, however, encompasses various categories of target in its variant forms who are not presently or imminently engaged in aggressive harming. Future threateners from the armed forces, political leaders, and others who support the army or oppressive government in various ways may be legitimate targets, as I will argue more fully in the coming chapters. But attacking them is justifiable only in the context of a legitimate collective armed struggle against a wrongful, collective threat of Life and Limb Rights violations.

There is therefore an important difference between the effects of a lack of legitimate authority on individual acts of pure self- and other- defence as compared with Organized Offensive Violence by non-state political groups: in the former case, a violent act may be illegitimate as an act of defence even if it is justified in a narrow sense with regard to its target;Footnote ⁴³ in the latter case, the lack of legitimacy appears to mean that, sometimes, the act is also unjustified with regard to its target. This is because in the various forms of Organized Offensive Violence the status of targets is determined by their role in a broader political conflict.

Finally, while my emphasis has been on the way a criterion of legitimate authority can be used to question the right claimed by some rebels to initiate and engage in war, the same principle is also an important component of the argument that legitimate rebels can make in support of a claim to recognition. Legitimate authority has been criticized for excluding legitimate non-state actors from the cover of just war theory; on the understanding I offer, it is a vital part of how they may demand access and claim the privileges of war in circumstances where it is necessary as the means of defending the innocent while pursuing their moral and political rights. The ability to make such claims will be crucial not only in Conventional War but also, as I argue, in Partisan War since this is justifiable particularly against enemies that refuse to recognize rebels with a moral right to engage in just war. The claim to legitimate authority is also important as part of the justification that rebels may offer for the use of irregular methods that may be necessary in both Partisan War and Conventional War, which is the matter to which I now turn.

7.2.1 From a two-tiered to a three-tiered theory

To substantiate this argument, let's begin by asking what the just war principle of ‘discrimination’ requires of participants in war. However interpreted, it is clear that discrimination condenses two propositions. It is often defined with a much stronger emphasis on the first, the one that prohibits soldiers from intentionally attacking those who are not legitimate targets; this is its restrictive force. But discrimination also expresses a permission, allowing at least some soldiers to attack legitimate targets directly and intentionally.Footnote ¹¹ In the theory of the just war, the default position must be that everyone is immune from attack until a valid reason can be given for harming them.Footnote ¹² The major burden in any account of the principle of discrimination, therefore, is to explain the occurrence of liability rather than immunity: we have to ask, who is a ‘legitimate target’ in war and why.Footnote ¹³

Broadly speaking, there are two approaches. Were we to answer the question based on considerations purely of the innocence or moral responsibility of individuals and their liability to harm, we might arrive at a result that is both complex and, in principle, highly determinate.Footnote ¹⁴ Those responsible ‘for an objectively unjust threat of harm’ are liable to be attacked when doing so is a necessary and proportionate means of preventing or remedying that harm.Footnote ¹⁵ Those who are not responsible for an injustice are not liable to be harmed intentionally in addressing its effects. Discrimination does not, in this perspective, distinguish between combatants and non-combatants but, in a qualified sense, between the guilty and the innocent. But while this approach to the jus in bello would be highly determinate in principle, it would face significant practical obstacles such that few think it likely that it could be applied generally in war. It would face significant epistemic difficulties due to the fact that those who are morally (as distinct from legally) liable to attack are often indistinguishable amongst the wider population, whether of soldiers or of civilians (though I shall argue in Chapter 8 that these difficulties are not always entirely insurmountable).Footnote ¹⁶ But more important is the problem indicated in Section 4.3.3 that specifying discrimination in this way in just wars generally is unlikely to help limit the scale of armed conflicts and incentivize restraint in targeting.

The second approach is also the dominant one in international law: it distinguishes between those who are and those who are not engaged in combat, irrespective of the side for which they fight, and allocates the same privileges and liabilities to both. Not only is this view supported in the traditional view on the morality of the just war but, as we have seen, even amongst revisionists, some see it as the most suitable framework for the law of war at the present time on grounds of moral pragmatism.Footnote ¹⁷ McMahan, for instance, argues that in the absence of a global judge competent to make timely and authoritative pronouncements under the jus ad bellum, the law of war cannot presently be altered in such a way as to reflect the philosophical insights of the deeper moral view directly. The ethics of war therefore have to be analyzed on two normative levels, the purely moral and the legal.Footnote ¹⁸

Whether one arrives at an endorsement of the Standard JIB through a two-level revisionist route like McMahan's – one that regards some elements as conventional and some not – or on the basis of the traditional view, the same problem arises that I want to identify regarding discrimination as an action-guiding principle. In circumstances where strict moral purism is impracticable, none of the remaining approaches gives a complete account of how the principle of discrimination can provide action-guidance because none of them reflects on the important normative dimensions of the problem of selecting and marking out a group of individuals to put forward as legitimate targets in war. To offer meaningful practical guidance to soldiers or resistance fighters, the principle needs a determinate content; that is, we need to be able to specify not only that some people are and some are not legitimate targets, but also which particular individuals will be put into which category on a given side and for the purposes of a given conflict. Analysis of the ‘deep’ morality of war is incapable of supplying this content to the Standard JIB directly. So, there must be a third level at which abstract, formal principles are given concrete content and hence made practically meaningful.Footnote ¹⁹ My claim is that the abstract nature of the legal principle of discrimination therefore leaves those who lead each warring side with an active role to play in giving it concrete meaning and practical significance. By deciding who will have combatant status and how the resulting class of combatants will conduct its relations with civilian life, they indicate to soldiers on the opposing side how to apply the principle of discrimination during the fighting and thus influence a factor important in determining how risk is distributed among their own citizens.

7.2.2 Just warriors and the loss of immunity

The ‘problem of in bello justice’ refers to the issues arising from this need to decide who will bear the risks and burdens of war and how they will do so. Before turning to the question of how political leaders should address it, it is necessary to say something further about why the problem arises specifically for sides with a reasonable claim to be fighting a just war. This, in turn, helps clarify exactly what is demanded of the leadership of a just side in formulating an appropriate conception of in bello justice.

On McMahan's two-tiered analysis, the fact that unjust warriors lose their normal immunities under the LOAC is relatively unproblematic since, by threatening unjust harms, most of them simultaneously forfeit their moral right not to be attacked. However, forfeiture cannot generally explain the same loss of rights by just warriors since the harms they threaten are presumed to be justified and therefore do not give rise to liability. Yet, like their unjust opponents, just warriors who fight for a state under the terms of the Standard JIB no longer enjoy the protection that they can normally claim against it by right. In fact, by putting them forward and designating them ‘legitimate targets’, the state actively aids and abets in their killing. Moreover, because the LOAC treats their deaths as impunible, just warriors suffer infringement of a second moral right, the right to see their killers pursued by legitimate legal authorities and punished (usually by their own state).Footnote ²⁰ As Benbaji writes, ‘[s]tates equalize the legal status of soldiers by immunizing enemy soldiers from post bellum legal prosecution’.Footnote ²¹ Just as importantly, if non-combatants are killed or otherwise harmed as the permissible side effect of actions aimed by unjust warriors at military targets on the just side, their losses are treated as if they were permissible.

If the just warriors’ liabilities under the LOAC cannot be explained in terms of rights forfeiture on moral grounds, then we have to explain them in some other way. On Joel Feinberg's account, there are three different ways in which one can lose rights: ‘they can be (a) voluntarily given away or exchanged; (b) lost involuntarily through negligence or wrongdoing; or (c) taken away by some other party’.Footnote ²² Option (b) corresponds to forfeiture while (a) corresponds to waiving a right or, more strongly, its ‘alienation’. The third possibility, (c), could take a number of different forms: rights might be taken away permanently; or they might be infringed, in which case they are recognized as valid in principle even when they are overridden by considerations of the lesser evil in practice; finally, they might be suspended, that is, taken away temporarily and for certain purposes, but restored once the occasion for suspending them has passed.

Where citizens are recruited by means of conscription, then many of their usual peacetime rights may be seen as having been infringed, for example, their rights of free movement, or free expression, or privacy. Insofar as the LOAC requires, furthermore, that the right of the just warrior to be protected from harm is taken away or overridden only for the duration of the war, we might consider using the term ‘suspension’. But the right to justice for those who are killed or harmed as a result of legally permitted actions by unjust warriors is lifted with permanent effect. However, for simplicity's sake, I will refer below to this bundle of requirements imposed on individuals on the just side by the LOAC as a ‘suspension’ of rights. With forfeiture ruled out, the loss or truncation of the just warrior's rights must be the result of either alienation or suspension.

Citizens might volunteer for military service in the face of an unjust attack on their country or an unjust occupation and thereby waive or alienate their normal rights and protections. Indeed this is likely to be the basis of recruitment to many non-state resistance groups.Footnote ²³ I want to set alienation aside for the moment, however, for three reasons: first, because if fully realized it would not pose the problem of in bello justice in its most acute form. If a sufficient number of competent individuals were to volunteer freely to fight in a given war, then the distribution of combatant liabilities could be regarded as legitimate on the basis of consent alone or at least as significantly less troubling from the point of justice than the liabilities suffered by conscripts (though leaders would still need to determine which and how many of the available volunteers to put forward for combat). Notice, however, that even if a just side recruited an army entirely of volunteers it would not make the killing of its (just) warriors by their enemies justified. Thomas Hurka has argued that by volunteering to fight, just warriors ‘freely gave up their right not to be killed in certain circumstances and so made their killing in those circumstances not unjust’.Footnote ²⁴ My point, by contrast, is that it is the distribution of legal liabilities to (morally unjust) harm under the Standard JIB in such cases that is ‘not unjust’ rather than the unjust warriors’ attempts to kill just warriors. Under such a distributive scheme, truly voluntary just warriors would still be wronged by their enemies when attacked even if they suffered no injustice from their own side. The problem of in bello justice, by contrast, can be seen most clearly in circumstances where the recruitment of a sufficient number of able volunteers is impossible.

Second, it is also very rare to find an army recruited exclusively on the basis of voluntary alienation. Since the end of the Vietnam War, for instance, the US has drawn its combatants from a professional, volunteer army. But, as many critics emphasize, its success in recruiting enough soldiers partly relies on a combination of material inducements and the relative poverty of those to whom they are offered.Footnote ²⁵ It seems likely, therefore, that insofar as they rely on forms of structural coercion to press individuals into military service, such cases run closer to suspension than to alienation and it is likely that most so-called volunteer armies involve elements of both.Footnote ²⁶ Third and probably most important, alienation cannot account for the fact that the killing of civilians is impunible when it occurs as a side effect of legally permissible action by unjust warriors.

I will therefore concentrate primarily on the third way in which individuals might lose their rights as just combatants or as permissible civilian collateral damage, that is, through the suspension of their rights by the state or, in a war of national liberation, by the provisional government or political leadership. A conception of in bello justice based on suspending rights would see political leaders trying to decide how best to meet the demands of a war by seeking the most just distribution of risks and responsibilities within the community they represent and on this basis ordering individuals to enlist in a justified infringement or more or less durable suspension of ordinary rights.Footnote ²⁷ The use of non-uniformed irregulars by non-state forces in liberation wars should, I think, be interpreted in this light. Where political leaders consider the use of irregular methods of warfare, forgoing the use of uniformed soldiers in favour of unmarked guerrillas and deploying them in close proximity to civilians, the normative proposition they would have to justify is that an expanded suspension of rights can legitimately be imposed on their non-combatants too. Thus, in addition to seeing their right to punishment for (proportionate collateral) harm overridden, non-combatants would suffer a partial infringement of their normal right to protection from the harms of war as the decision to use irregulars caused an increase in the rate of collateral damage. I will discuss this in more detail in Section 7.4.

I now turn to the question of which normative principles ought to guide the leaders of a particular warring side in trying to determine the most suitable distribution of liabilities and risks in a just war.

7.3.1 Justified determinations of in bello justice

We assume, ex hypothesi, that the need to engage in any just war is a matter of grave urgency and that doing so is the only or the most proportionate means of trying to prevent or remedy injustice. Engagement in the war then necessarily imposes certain injustices on the just side: it is forced to send its citizens out to fight, sometimes coercively, knowing that some of them will die as a result, the victims of morally unjustified killings; and, to subject itself to the Standard JIB, it has to default on its duty both to protect them and to pursue justice for their deaths and for those of some civilians harmed collaterally. The Standard JIB treats unjust warriors fighting on the other side who inflict unjust harms as if doing so were morally permissible and, by the same token, obliges those they attack to treat them as if they did no wrong.

My claim in Section 7.2 was that a community forced to engage in just war under the terms of the Standard JIB contributes to specifying the application of discrimination by delineating combatant and non-combatant groups within its own population. I will refer to a particular way of trying to construct these categories fairly and effectively as a ‘determination of in bello justice’. I now want to argue that, faced with a particular conflict, a just side might have to select from a set of several alternative but individually defensible determinations, all of them compatible with adherence to the Standard JIB, and that these might include options for the non-state sides in asymmetric wars that involve allowing higher levels of exposure to risk for non-combatants than are usual in regular wars between states. Indeed, it might be that options of this kind are the only ones that could justifiably be implemented in some cases. Any of these might be capable of realizing in its own way the abstract concept of discrimination and establishing a basis for operating within the Standard JIB. Leaving aside for now the question of fairness towards enemy soldiers (I turn to this in Section 7.4), each alternative, I argue, must be evaluated according to its ability to achieve a suitable balance between three distinct considerations: (1) fairness domestically; (2) the value of survival which I will call the goal of bare peace; and (3) the value of justice as a goal of war, which we can call the goal of just peace. I will outline each in turn.

The first consideration demands that those burdens of war that are deemed necessary on reasonable grounds to win and to realize or defend the cause should be distributed within the society as equitably as possible. The workings of a particular determination of in bello justice have, to use Allen Buchanan's terms, ‘profound and enduring effects on individuals and groups’ but these ‘effects are for the most part neither chosen nor consented to by those affected…’ Intuitively, therefore, in cases where a side has to recruit many of its soldiers or take decisions affecting the risk of collateral harm to civilians on a non-voluntary basis, it must seek to achieve ‘fairness of distributions of benefits and harms’ as far as humanly possible.Footnote ²⁸ Fairness in this sense would primarily rule out selecting combatants based on arbitrary criteria that are not relevant to their ability to fight (such as social privilege or education would usually beFootnote ²⁹). If soldiers are recruited on a voluntary basis, this might to some extent obviate the question of fairness in the distribution of the burden of combatancy. But in a state that did so it would be necessary to ensure that the use of financial inducements was not taking unfair advantage of social inequalities. And since variations in the visual and spatial separation of combatants from non-combatants will affect the degree and kind of risk to which civilians are exposed, the question of distributing the burdens of war fairly is also relevant to the non-combatant population.

Fairness in this sense is not, however, the only consideration, and a defensible determination of in bello justice must also weigh important prudential considerations in the balance.Footnote ³⁰ On the one hand, what I call the ‘goal of bare peace’ (the second consideration) concerns the need to ensure that as many of the people in whose interests the war is fought survive as possible. This reflects the aim of the discrimination principle as usually understood: it provides each side with a means of trying to protect some of its population, structures, and resources from the conflict, which it does partly by designating as many people ‘non-combatant’ as it can afford to. For the purposes of applying a viable determination of in bello justice, this will mean that the alternative arrangements should be evaluated according to the nature and degree of exposure to military hazards that each is likely to entail for the side as a whole.Footnote ³¹ Wars generally entail at least some harm to civilian lives and the structures and spaces within which they are lived. A people engaging in war can try to vary the degree of harm by its decisions both about which civilians to enlist into the armed forces and how many, and about how to distance its combatants from civilians and civilian zones during the fighting.

The third value determining the range of available determinations in bello justice arises from considerations of the cause justifying war. The presumption that the side in question has engaged in a liberation war under the Standard JIB that meets the Proportionality requirements analyzed in Chapter 5 implies, ex hypothesi, that the goals it seeks to achieve or the values it tries to protect are of sufficient value to justify resistance, given the violence it is likely to provoke and the costs of defending people from it. Thus, the value of the cause is itself a key variable in the formula by which political leaders should decide how far to expose their citizens to danger. Insofar as the primary cause of resistance is defence against the Life and Limb Rights violations of a violently oppressive regime, the goals of bare and just peace overlap since pursuing just peace is a means of survival. But to the extent that resistance aims at securing non-Life and Limb Rights and faces violent reaction from the state as a result, the goals are in tension since the threat to innocent lives from the enemy arises as a result of pursuing just peace.

In his analysis of the contract by which he thinks the basic principles of the jus in bello would be agreed between states, Benbaji writes that ‘the objective of the contracting parties is minimizing the harm inflicted on morally innocent people within wars, without limiting the right states have to use force in protecting their just claims’.Footnote ³² My argument is that the same considerations should guide states in their uptake of the jus in bello as might guide them in its original construction. In this perspective, the more important the cause is, the greater the degree and type of exposure of civilian life and lives a people might find it reasonable to contemplate. Together, the second and third considerations present themselves to a warring side as a gamble which it has to approach as prudently as possible, all things considered. This takes us back to the idea of the Proportionality of Resistance I introduced in Chapter 3. It has to ask itself what exactly and how much it is able or prepared to place on the table. What losses is it prepared to risk suffering? Are the forces at its disposal such that it can hope to avoid defeat even with only a relatively small exposure of people and goods to the fighting? Or is a high level of exposure for its citizens necessary to offer serious hope of victory, for instance through mass conscription? The balance of forces between different sides in the war and the tactical question of how to meet hostile forces successfully are variables that will affect the decisions a side needs to make as it considers how best to absorb the demands of moral, political, and military necessity.

Hence, in bello justice requires the just side to specify the most just way to secure individuals in a community from Life and Limb Rights violations while taking account of sacrifices they might be willing to make in pursuit of important Political Rights. Sometimes, the best arrangement will require concentration of risks and burdens on the shoulders of relatively few who will serve as combatants. In Chapter 2, I indicated how a Natural Duty of Justice required people to make reasonable contributions and sacrifices for the sake of securing justice when a proportionate opportunity for doing so presented itself. This duty, I now submit, may obligate those best able to do so to volunteer for combat roles where war is necessary. It might be objected that this goes beyond ‘reasonable sacrifice’. My response is that the sacrifices it demands are not excessive in circumstances where greater numbers will face similar or worse risks unless the volunteers defend them. There are factors here that may lessen the force of the duty to volunteer: if more eligible persons exist than are needed, then the duty is imperfect; and reasonable partiality might lessen the force of the duty further. But if the need for common defence is urgent and grave, it seems likely that a competent leadership could justify enforcing it nevertheless. In circumstances where an authoritative resistance leadership exists, one with the wherewithal and opportunity for a proportionate attempt to defend against Life and Limb Rights violations from a violently oppressive regime, then it is likely that it will need to undertake coordination and enforcement of the duty to serve. This it would do by securing enough recruits to mount an adequate resistance by consent if possible and by conscription if necessary. However, insofar as the goals of resistance are chiefly Political Rights, such as national self-determination, it lessens the likelihood that coercion will be justifiable. Conscripting someone to defend against a potentially lethal threat that they face in any case along with others but that they may be able to prevent can be justified in terms addressed to the interests and moral duties of the conscript. But if the cause of resistance is a more narrowly political one, whether s/he has a duty to serve in a defensive capacity made necessary by pursuing it depends on the extent to which the conscript is himself or herself committed to it.

The range of possible determinations of in bello justice that a people and its leaders might contemplate in the face of a particular conflict is thus based on considerations of both fairness and prudence in seeking to secure the rights and interests of their citizens. The rights of enemy soldiers form a further constraint when it comes to the use of irregular combatants within the terms of the Standard JIB, which I will discuss in Section 7.4. But first, the possibility that there might be more than one potentially valid approach at the same time becomes clearer when we differentiate between three distinct ways in which a particular determination of in bello justice may be adjudged ‘legitimate’.

7.3.2 Legitimacy

A determination of in bello justice may be deemed ‘legitimate’ in one sense according to its fairness and probable outcomes, that is, if it is justified in terms of the balance it achieves between the three social goals outlined in Section 3.1: distributive fairness domestically, the goal of bare peace, and the goal of just peace. Legitimacy in this sense runs close to ‘justification’ as characterized by A. John Simmons, but the usage is compatible with that of various theorists, Buchanan in particular.Footnote ³³ However, the first form of legitimacy does not alone complete the political dimension of in bello justice.

The second sense in which a determination may be adjudged legitimate concerns the proper basis for determining its content and for ensuring that its outputs are justified. Following the argument of Chapter 6, I assume that to achieve a meaningful understanding of what would count as an appropriate output, you need consultation with those affected by the policies in question; you need to know something about their interests and preferences. Justifying a particular determination relies on difficult and uncertain judgements about various imponderables: first, about the likely consequences of different strategic alternatives for military recruitment and deployment both in terms of the risks they pose to civilians and the hope of victory or defeat that each offers; second, about the kind of weight that should be given to the protection of civilian life from the war itself (bare peace); and third, the value that should be ascribed to the goals to be secured through victory (just peace). The first set of issues are subject to contingencies that make any attempt at calculation highly uncertain. The second and third have an irreducibly subjective component insofar as the beneficiaries of resistance are entitled to choose a high exposure to risks necessary for securing Political Rights if they deem them sufficiently valuable. All three, therefore, require consultation as widely as possible with the intended beneficiaries of the just war. On the other side of the calculation, where the prospects of success are uncertain and there is a danger of large numbers of innocent casualties, consultation is necessary to find out how much additional risk those affected are prepared to endure for a given increase in the chance of success. As Rawls puts it, ‘[t]here is no avoiding […] having to reach a complex judgment weighing many imponderables, about which reasonable persons are bound to differ’.Footnote ³⁴ Therefore, dialogue with those on whose behalf the war is undertaken is necessary for a proper calibration both of the goals of the war and its potential costs and a second form of legitimacy turns out to be the necessary prerequisite of the first. Whereas established democracies are likely to have a range of instruments and institutional channels available to them through which to seek indications of popular concerns and preferences, non-state groups have much more varied opportunities for consultation and will have to rely more often on informal forms of communication.

The question of legitimacy arises in a third form, where there is more than one possible determination of in bello justice and where each has a contested claim to justification (or legitimacy in the first sense). This is most apt to occur where there is widely felt uncertainty about the weight that ought to be given to different variables. For instance, the value assigned to the justifying cause of the war might be calibrated in different ways on subjective grounds, as some individuals may be happier to face the risks associated with defeat than others. Similarly, there may be reasonable differences of opinion about the chance that a policy will be effective (say, an increase in the numbers of civilians recruited as combatants) or people might feel more or less strongly about the risks associated with deploying irregular rather than uniformed troops. If there are effective channels through which to measure, influence, or negotiate with popular will, then it is necessary to use them, seeking indications of support for a particular policy in the wider population. The greater the exposure of beneficiaries to risk, the greater the urgency of consultation. Where there are indeterminacies of this sort, that is, where there is not simply one policy with clear advantages over all others, then it is necessary to assess the degree to which those most affected are likely to consent to them. The ‘burdens of judgement’ ought to be shared, therefore, as widely as is practicable in the circumstances.Footnote ³⁵

Some arrangements are unjust no matter how widely they are accepted (e.g. if a state conscripted only the members of an ethnic minority or if it sought to recruit only the economically vulnerable as ‘volunteers’). Even if wide support could reasonably be expected in a particular context and so they were legitimate de facto (i.e. in sense 3), they could be challenged from the point of view of justice (legitimacy in sense 1). So while fulfilment of the third legitimacy criterion (or at least the absence of strong evidence of popular opposition) may be necessary, where feasible, to validate a particular determination of justice, it is not sufficient. But equally, where channels are available for consultation and authorization, it will sometimes be the case that prima facie fulfilment of legitimacy in the first sense is not sufficient without fulfilment of the second and third senses.

7.4.1 Four types of irregular tactic

There are several quite different ways in which irregular combatants can involve the civilian population on whose behalf they fight. I will distinguish four.

With the first tactic (1: Civilian Camouflage) irregulars retain the appearance of civilians but not for the purposes of military deployment or ambush, rather to avoid elimination prior to combat. The guerrillas’ decision is simply not to take the additional step of wearing a uniform or insignia, remaining dressed in their normal civilian garb and concealing their weapons except during combat. The consequences of this decision are greatest where they operate in densely populated civilian areas.

Tactic (1) is different from cases where soldiers adopt a civilian disguise in order to mislead enemy soldiers.Footnote ³⁶ As an instance of this second type of tactic (2: Civilian Disguise), consider the French partisans discussed in Walzer's Just and Unjust Wars who dressed as peasants in order to ambush passing German soldiers or those FLN guerrillas depicted in Gill Pontecorvo's film, The Battle of Algiers (1966), who used civilian guise to get close to targets before shooting them when their defences were down.Footnote ³⁷ The difference compared with (1) is that civilian identity in (2) is actively and deliberately deployed as a means of lowering the enemy regulars’ guard in order to diminish their ability to defend themselves while they adhere to the discrimination principle. The distinction is made in Protocol I, which permits (1) but expressly prohibits (2): non-uniformed combatants are permitted as long as the irregulars carry their arms openly ‘(a) during each military engagement, and (b) During such time as [they are] visible to the adversary while […] engaged in a military deployment preceding the launching of an attack in which [they are] to participate’ (Art. 44 (3)). But the Protocol prohibits the use of civilian disguise where it is used perfidiously: Article 37 declares that, ‘[i]t is prohibited to kill, injure or capture an adversary by resort to perfidy. Acts inviting the confidence of an adversary to believe that he is entitled to, or is obliged to accord, protection under the rules of international law applicable in armed conflict, with intent to betray that confidence, shall constitute perfidy. The following [exemplifies] perfidy: […] (c) The feigning of civilian, non-combatant status […]’.Footnote ³⁸

Tactic (2) is different again from tactic (3) (which I'll call Human Shields), which involves placing people clearly identifiable as civilians in a visible position close to military targets in order to deter attack.Footnote ³⁹ The extent to which this tactic actually increases the risks facing civilians will depend on the enemy's tolerance of making sacrifices in military effectiveness for the sake of discrimination and its willingness to inflict collateral damage. As I intend it, tactic (3) refers specifically to the coercive use of civilians as human shields. It is conceivable that civilians might sometimes volunteer to aid their own soldiers, but I leave such cases aside. I presume they raise questions about discrimination rather than Proportionality since the civilians thereby arguably take on a combat role.Footnote ⁴⁰

Finally, a fourth possibility (4: Induced Civilian Casualties) is that a group might conceal non-combatants in a position close to a visible military target to increase the chance of collateral harm and fuel feelings of outrage against the group's enemies. This is a form of terrorism – in the standard sense of the term – since it involves intentionally causing the death of civilians as a military or political instrument. While civilian casualties arising as a result of tactic (3) implicate enemy forces morally when they continue to attack regardless of the evident collateral risks, tactic (4) is designed to diminish the agency of the enemy in order to blame them after the fact.

The only tactic that I believe to be defensible in the context of the Standard JIB is number (1). In some more desperate circumstances, tactics (2) and even (3) might be justifiable in terms of the first three normative considerations discussed above in Section 7.3 (something I will return to in Chapter 8), but they raise problems of fairness towards enemy regulars under the framework of symmetrical combatant rights codified in the Standard JIB as I will argue in Section 7.4.3. By contrast with tactic (1), tactic (4) aims directly at the death of innocent civilians for political advantage which both violates the central principle of the Standard JIB and is pro tanto wrong independently of it; tactics (2) and (3) both involve a breach of the trust necessary to the Standard JIB by exploiting the willingness of enemies to comply in good faith with the principle of non-combatant immunity (NCI) within the terms of the LOAC as a means of killing them. I will expand on this point in Section 7.4.3 where I defend tactic (1) from the charge that it necessarily imposes unfair disadvantages on enemy combatants. But first, in Section 7.4.2, I challenge the argument that it necessarily imposes unfair burdens on non-combatants.

7.4.2 Irregular war and the rights of non-combatants

Some critics of Protocol 1 argue that using non-uniformed soldiers wrongs civilians on their own side by increasing the rate of collateral harm they suffer from enemy attacks, particularly when irregulars fight from within densely populated areas.Footnote ⁴¹ On the account of in bello justice I offer, irregular tactic (1) can be defended from this objection.

I have argued that the question of whether a particular determination of the combatant–non-combatant distinction within a society is justified cannot be answered by reference to whether the people whom it places at risk are morally liable to harm, since none is (unless and until they render themselves liable to attack by wrongful acts of their own such as war crimes). Instead, it must be decided on the basis of whether its distribution of risks adequately balances fairness and the goals of just peace and bare peace. The claim I wish to make on this basis regarding irregular warfare is as follows: if these three considerations can sometimes justify a legitimate wartime leadership in exposing some of its morally innocent citizens to very high levels of risk by conscripting them as combatants and ordering them out to fight, then, a fortiori, it could surely also justify in some circumstances the exposure of other equally innocent citizens to lesser risks of harm collaterally by permitting the soldiers to fight without visible insignia. Both decisions involve the same kind of rights ‘suspension’, as I have called it, though the truncation of citizens’ rights in the second case is less drastic. Based on my account of in bello justice, irregular warfare would therefore be justified where wartime leaders determined in consultation with their people that the ends pursued were of sufficient value, the chances of victory using conventional methods were sufficiently slight, and the likely damage to civilian life resulting from the use of non-uniformed troops was likely to be both sufficiently fair in distribution and acceptable in scale to the population generally as the price of increasing the chances of victory.Footnote ⁴²

Based on the third legitimacy criterion, a non-state movement should not initiate irregular war in the face of widely articulated popular opposition. But the greater the support, the greater the legitimacy of using a strategy by which people agree (to whatever extent possible) to spread the risk more widely across the population than in regular war. The ability of a non-state group to conduct an irregular war justly is therefore likely to be strongly affected by the legitimacy of the organization itself, in terms both of its support amongst the people it claims to represent and the degree to which it reflects their interests and preferences in its decisions.

Setting aside the question of the rights of enemy regulars, the policy of narrowing the visual and spatial distance separating combatants from non-combatants by deploying non-uniformed soldiers in urban settings could be justified on the basis that the distribution of risks and potential benefits it offers is fair overall. Insofar as the additional harms civilians suffer as a result of their increased vulnerability are the result of unjust war by their enemies, they are grave injustices. But where the three legitimacy conditions (in Section 7.3.2) are fulfilled, they do not constitute an unfair infringement of the civilians’ rights by the irregulars themselves or by their political leaders.

7.4.3 Irregular war and the rights of regulars

To complete the argument, I turn now to the claim that the use of irregulars within the Standard JIB as applied to asymmetric wars unfairly biases the rules in favour of rebels and against their regular opponents. My response is to argue that while (2) Civilian Disguise and (3) Human Shields, as distinguished in Section 7.4.1, would typically aim at an unfair (dis)advantage within the terms of the Standard JIB, Civilian Camouflage (tactic 1) would not.

In practical terms, the decision to use Civilian Camouflage might reduce the ability of regulars to do three things:

1. (i) to achieve legitimate military aims through combat;

2. (ii) to defend themselves individually from harm;

3. (iii) to assassinate or arrest enemy combatants who are not presently engaged in combat and prior to deployment.

Criticism of Protocol 1, Article 44(3) tends to focus on the first and second of these abilities.Footnote ⁴³ Irregular tactics reduce the ability of regular forces while enhancing that of the irregulars to achieve legitimate war aims. Moreover, it is much harder for regular soldiers to defend themselves individually when trying to fight discriminately against enemy combatants whose status is not clearly signalled. The problem with permitting irregular warfare in the LOAC, on this view, is that it introduces an unfair distribution of rights between the irregulars on one side and, on the other, both the opposing regular army as a whole and its individual members. To respond to these claims, it is necessary to specify precisely why irregular warfare might be justified in the context of wars against occupying forces specifically, when it is not permissible in a war between states even of very different strengths.

The reason why some irregular tactics are justifiable in these cases is not primarily due to asymmetries of power or resources as is often assumed but arises from the conditions of occupation or domestic political oppression.Footnote ⁴⁴ Wars fought against occupying armies and domestic oppressors have two important features absent from wars between states. The first is that the enemy's presence within and across the whole extent of the disputed territory makes it impossible to define a front line dividing the two opposing sides. Front lines are very important for the normal conduct of war: they provide a means by which one's civilians can be kept back from the fighting and they define the point at which combatants expose themselves directly to the hazards of war. Behind the line, combatants can withdraw, regroup, await orders for mobilization, or even retire and return to civilian life. As Benbaji writes, when ‘warfare is removed from the cities to the battlefield […] the soldiers’ family members are safer, released soldiers would have safer places to return to, and wounded soldiers would have protected healing places’.Footnote ⁴⁵ In a war with no front line, these benefits become at best problematic and at worst impossible as all combatants are continually exposed to the risk of attack or capture, at least as long as they are easily identified visually. A second important consequence of occupation is that decisions about where to engage enemy forces in fighting are largely taken out of the hands of the irregular forces because the regulars are already in the country, occupying its key strategic positions. Based on these decisions, the theatre of war is likely to overlap unavoidably with the spaces occupied by civilian life, whatever the insurgents might wish to do. All in all, as Michael Gross observes, ‘the weaker side, fighting within civilian population centres, does not have a safer place to retreat’.Footnote ⁴⁶ The forces of resistance are therefore left with little room for manoeuvre.

In circumstances like these, where enemy regulars have dominance across much or all of the territory in dispute, forgoing the use of insignia or uniforms on the rebel side and concealing weapons when not in use or deployment is likely to be necessary to avoid arrest or assassination while not engaged in combat. During any time when rebels engage in activities that are normally carried out away from the front in regular wars or if they simply retreat for relief from the fighting, wearing a uniform or insignia would surely invite elimination by the enemy. This is especially the case in the era of drones and other remotely controlled weapons. In the face of this kind of dominance, irregulars might use Civilian Camouflage simply to establish the same ability that regulars have to withdraw temporarily from fighting for purposes not inconsistent with the laws of war. And doing so for these purposes, I submit, is not only not at odds with the Standard JIB (and the LOAC which is built on it) but very much in its spirit.

The Standard JIB and the LOAC seek to regulate and limit conflicts (including those with non-state groups) by means of a principle of combatant equality that requires the setting of rules that are equitable between opposing sides. When it is directed solely at avoiding elimination by assassination or arrest, the use of Civilian Camouflage restores fairness in the distribution of rights and duties under the LOAC between soldiers on opposing sides of asymmetric wars rather than undermining it. The option of organizing or retreating behind a front is generally available to regular soldiers. This is especially so if the regulars fight for foreign occupiers or colonial states. But even in cases of domestic oppression and civil conflict, regulars are more likely than their opponents to have fortified spaces into which they can withdraw. Permitting non-state soldiers engaged in liberation wars to forgo the use of uniforms affords them to some degree the same ability. The protection that they thereby enjoy from assassination or arrest by regular forces is then reciprocated in the duty of irregulars to abstain from the ‘perfidious’ use of civilian disguise in ambush. This duty prevents irregulars from exploiting the discrimination rule as a means of diminishing the ability of regulars to defend themselves individually. So as long as the only consequence deliberately sought by legitimate irregulars using Civilian Camouflage is (iii) to avoid elimination prior to combat, then the use of irregular warfare is consistent with fairness to enemy regulars within the terms of the LOAC as well as to civilians on the irregular side.

By contrast, the aim of Civilian Disguise and Human Shields (tactics (2) and (3)) as defined cannot be the same as that of Civilian Camouflage (1). Where human shields are placed visibly near a military target, the purpose is not to facilitate retreat and restore balance in the ability to withdraw behind the front line. It is to exploit the principle of discrimination itself and the enemy's willingness to be bound by it as a means of eliminating their capacity to fight. The collateral damage that human shields will suffer is a deterrent aimed at limiting the regulars’ ability to pursue what the Standard JIB requires participants to regard as ‘legitimate military aims’. It may also be used to diminish the regulars’ ability to defend their lives. The tactic of Civilian Disguise aims specifically at this latter ability, exploiting the good faith of enemies who comply with the principle of NCI in order to render them defenceless.

Now it is true that, strictly speaking, insofar as they are the ones posing wrongful threats, the regulars in cases of just wars of liberation by hypothesis have no moral right to defend themselves from justified attacks by legitimate rebels. And I will explore the possibility in Chapter 8 that some conflicts or particular phases within some conflicts might be governed by rules that directly reflect this fact. For now, however, my focus is on conflicts in which the rebels determined that following the Standard JIB was the best way to fight based on the expectation that their enemies would do so too. In those cases, the rebels agree implicitly to a trade-off whereby they treat enemy engagement in war as permissible insofar as it is directed solely at standard military goals and defence against rebel attack in return for compliance with the principle of discrimination. So, the leaders of rebellion in a sense disregard their combatants’ moral immunity from attack in return for greater assurances of the immunity of their non-combatants. This trade-off is justified, as I argued in Chapters 4 and 5, primarily where it is part of the most proportionate strategy available for resistance. And rebels ought to abide by the commitments required by the Standard JIB if they are to realize this strategy successfully, a goal to which they are morally committed. This is the main reason why tactics (2) and (3) (Civilian Disguise and Human Shields) are illegitimate for a side that seeks to fight within the terms of the Standard JIB: both seek to make the principle of discrimination disadvantageous to one side while retaining its benefits on the other (all the while based on the assumption that the obligation to observe discrimination in the choice of targets still applies to both). Both, therefore, go back on the trade implicitly agreed within the Standard JIB and, to that extent, betoken bad faith; moreover, both will tend directly to undermine the enemy's willingness to fight within the constraints of the Standard JIB and so both will deteriorate the protection that civilians on the rebel side enjoy under the LOAC. They may therefore undermine the rebel project to realize legitimate goals of justice in the most proportionate way possible.