4.2.1 The rules of engagement: five ‘codes’

Let's imagine that there was a prima facie case for armed resistance against an oppressive Regime and that the leaders of Resistance therefore had to formulate the best possible strategy for challenging its rule, ideally with a view to convincing its leaders to make way for a popular revolution by stepping down. Note that if there is already a prima facie case for armed resistance, then by hypothesis it must already be clear that any attempt to confront Regime using peaceful protest or civil disobedience is likely to be met with armed force. This would have been discovered either (a) where Regime used violence in response to peaceful attempts at resistance or, possibly, (b) in the historical evidence available to Resistance indicating that the regime was highly likely to do so. The latter judgement could be based on a familiarity with the regime's normal routine of human rights violations if these predated the present attempts to resist, which might in any case (c) provide an independent prima facie justification for armed force.

Let's say for the sake of simplicity that, in Regime, the Resistance had been engaged in a preliminary form of peaceful protest, demonstrating against the regime in the hope that public attention internationally would be sufficient to inhibit the state from unleashing its full force. Resistance encourages its followers to assemble in Liberty Square but they meet with violence from State Security Police who attack them using batons, tear gas, and firearms with live ammunition. What would happen next? What alternatives ought the leaders of Resistance to weigh as they evaluate the situation? What sorts of tactics could they permissibly use?

The first alternative arises without the need for any decision by the Resistance leadership. The most basic form that armed resistance may take is ‘Purely Defensive Violence’. If agents of Regime attack protesters using armed force without moral justification, they are individually guilty of acts of aggression, morally speaking. Where they present an imminent threat to an innocent protester, the victim or someone assisting them may repel the attacker using such force as is necessary for effective defence and provided the harm it is expected to cause is not disproportionate to the injury likely to result from an unimpeded assault. In the absence of any decision to initiate further offensive action by the Resistance leaders, therefore, the first code arises naturally and it has two features: first, it reflects ordinary rights of individual self- and other-defence such as would arise either in peacetime or wartime. These arise without the intervention of political leadership and are, as it were, facts of ‘natural’ justice. Second, it is non-strategic in that the protesters’ use of violence and their right to do so merely responds to threats as they emerge rather than following any coordinated plan.

The ethical framework of Purely Defensive Violence is likely to have been applicable during the pro-democracy demonstrations in eastern Europe before 1989 (in Budapest, 1956, for instance, or Prague in 1968) and the events of the Arab Spring in 2011 (in Egypt, for instance, or Libya or Syria). In the city of Deraa in Syria, for example, in June 2011, Robert Fisk writes of civilians faced with repressive violence from Assad's forces that turned them away from the ‘peaceful’ models of protest in Tunisia and Egypt and saw them resorting to armed force to ‘defend their families’ and ‘“shooting back” for the sake of “dignity”’.Footnote ⁶ Some of the force used by protesters in response to police violence in the Maidan in Kiev in February 2014 is interpreted and justified by its defenders in this way.Footnote ⁷

In all such cases, where soldiers or the police threaten to harm innocent protesters without justification, they are liable to proportionate defensive force. There is also at least one case of a political organization that explicitly adopted these rules as a code of conduct to define the rules of engagement their soldiers were to follow. Menachem Begin's memoir of the Jewish resistance to British rule in Palestine describes how, during the Arab Revolt of 1936–39, the Haganah, the Jewish paramilitary organization whose name means ‘Defence’, ‘pursued a policy known as havlagah, a Hebrew word meaning self-restraint. It prescribed non-retaliation except under direct attack. As a result of this policy the Arabs were always sure of the initiative and never suffered any counter-attack. Their risks were reduced to what they were liable to suffer during attacks timed to suit themselves and on a battlefield chosen by themselves.’Footnote ⁸

A second possibility, this time one that the leadership of Resistance might choose is Strategic Nonviolence. It might seem at first glance that this ought, in a way, to be the first resort. The idea that all nonviolent avenues ought to be explored before a ‘last’ resort to force is permitted is a familiar one. However, there is an important difference between the Strategic Nonviolence that resistance leaders might enjoin their followers to adopt as means of resistance in the face of brutal repression and the peacefulness of initial protests prior to any (threatened) violence by Regime. In states where the government does not threaten violent repression, then peaceful means of protest are simply the default resort for resistance, that is, nonviolence is what ordinary, peacetime morality requires since armed resistance would simply be impermissible in the absence of prior, unjust violent threats. By contrast, once it is clear that the regime is violent, that is, either in the form of a persistent background threat of killing, torture or the equivalent or in an explicit threat of attack during or in anticipation of protests, peacefulness is not the default position. For those who are not absolute pacifists, committed on moral or religious grounds to nonviolence even in the face of lethal aggression, the default position in these circumstances is Purely Defensive Violence, that is, it is permissible for individuals to defend themselves using necessary and proportionate force and, arguably, a duty for some to use force on behalf of others where those others cannot defend themselves.Footnote ⁹ For the leaders of Resistance to call for nonviolence in these circumstances, therefore, is to demand that important individual moral rights (and corresponding duties) be waived in the circumstances. To uphold the practice of Strategic Nonviolence, individuals are asked not to defend themselves, even if it means allowing themselves to suffer wrongful arrest, injury, or death.

To justify making such a demand of resistance supporters requires that important values be identified that could weigh against their sacrifices on the other side of the scales. And those otherwise willing and able to defend others in appropriate circumstances would need good moral reasons to override their moral duty to do so. Moreover, the ability to persuade many to adhere to a principle of nonviolence in the face of imminent wrongful threats may depend on how much people are willing to trust the leadership of Resistance or to internalize a particular moral view on nonviolence or pacifism as a value (such as was seen historically in the movements led by Gandhi and Martin Luther King). Another factor that can contribute to the feasibility of nonviolent protest is the presence of a ‘neutral’ third force willing to ward off threats from the state. This was the case during the civil rights protests in the USA as well as more recently in the Egyptian revolution. In Egypt, the ability of many protesters to withstand the pressure of a notoriously brutal state security service and maintain a nonviolent stance against the government was possible only because the army adopted a neutral stance between them and Hosni Mubarak's regime. Martin Luther King's nonviolent movement also had its background support in the form of federal police officers whose presence was at times important in preventing violence against the protesters.Footnote ¹⁰ In effect, however, this does involve someone threatening violence even if it is not the protesters themselves, which may be justifiable with reference to rights of self- and other-defence. Alternatively, the more neutral the third force, the more likely it is that we could plausibly describe it as a circumstance facilitating nonviolence that either staves off the prima facie case for armed resistance or enables the adoption of Strategic Nonviolence with lowered costs. One other way in which the threat of force could contribute to the effectiveness of nonviolent methods is where the government fears that violence might ensue if it fails to respond to protesters’ demands, as Adam Roberts has suggested.Footnote ¹¹

A rationale for choosing Strategic Nonviolence over Purely Defensive Violence could arise from a number of factors. The first is the danger of any resort to violence by protestors in the face of unjust regimes, as many advocates of nonviolent resistance emphasize. Gene Sharp, for instance, whose From Dictatorship to Democracy may have influenced protesters during Egypt's revolution of 2011, has argued that,

Jeff Goodwin has suggested various reasons why Resistance might try to persuade protestors to adhere to the code of Strategic Nonviolence. In the absence of ‘overwhelming’ state repression, he writes, nonviolent methods may have greater success than ‘armed struggle’ in achieving ‘popular mobilization’. Armed resistance is typically most effective only in ‘peripheral or thinly populated regions where the state's infrastructural power is weak,’ but nonviolence may be effective over a wider range of spaces. Moreover, while using violence to suppress nonviolent demonstrations may provoke some of them into resorting to arms in response, Goodwin suggests that it can also ‘swell the ranks of nonviolent movements’ while sowing ambivalence in the security forces sent to attack them, both weakening state power and increasing the chances of ‘regime change’.Footnote ¹³

Some developments during the Arab Spring partly bear out Goodwin's conclusion. The violence of Mubarak, Gaddafi, and Assad in particular helped at times to increase support for the opposition domestically and to galvanize international support for opposition movements in their states and condemnation of their regimes. In Egypt, the persistence of generally peaceful revolt succeeded in ousting Mubarak in February 2011, while the support of NATO for an armed rebellion in Libya after what were believed to have been unprovoked attacks on peaceful protesters eventually led to a military defeat of the regime.Footnote ¹⁴ But this brings us from the second code (perhaps via purely defensive force while protesters initially defended themselves from Gaddafi's forces) to the outbreak of offensive force and war.

In our narrative, let's imagine that Resistance resolved to bypass Purely Defensive Violence (Code 1) on the basis that there seemed initially to be a better chance of success in the terms described by Goodwin through Strategic Nonviolence (Code 2). But the ability and willingness of the state to utilize the State Security Police and parts of the army in suppressing resistance was such that people were terrorized into clearing the streets. Resistance leaders would then have to choose between two alternatives: they might decide, first of all, to give up and wait for a better chance to resist in the future. Or, they might consider escalating on their side of the conflict by preparing for a more organized violent response, shifting from Purely Defensive Violence to some form of offensive force. This is how Begin viewed Jabotinsky's formation of the Irgun Zvai Leumi (the ‘National Military Organization’) which abandoned the Haganah's policy of self-restraint in Palestine during the 1930s in favour of a strategic use of force that permitted offensive operations (initiating ‘the first attack by Jewish arms’) followed soon after by the ‘Fighters for the Freedom of Israel’ (more often known as the Stern Gang).Footnote ¹⁵ During the Rivonia Trial, Nelson Mandela described a decision of this kind in his defence of Umkhonto We Sizwe's resort to force.Footnote ¹⁶ Offensive Violence has two distinguishing features as compared with Purely Defensive Violence: first, it is directed strategically by the leadership; second, it involves taking the tactical initiative away from Regime and its agents by engaging in offensive actions, that is, it permits various different forms of attack that need not be prompted by individual cases of interpersonal aggression from the enemy. I presume this is justifiable where the threat from Regime is organized collectively in such a way as to necessitate a collective coordination of forces for effective defence.

One reason why Resistance might not choose to stand down in the face of repression is that an unsuccessful revolt may lead to greater oppression in the aftermath than before. Sharp suggests that this is one of the risks associated with violent revolts, but it is also likely to be true of nonviolent protest.Footnote ¹⁷ As the response to mass protest in Iran in 2009 and in Libya and Syria in 2011 demonstrate, the sense of political insecurity felt by rulers faced with popular expressions of dissent may motivate harsh repressive measures. On the other hand, there is very good reason to be highly circumspect about escalating to one form or another of Organized Offensive Violence. What chance is there that organized force by non-state groups (‘an opposition made up of former doctors, farmers, pharmacists and so forth’, as Barack Obama described the Syrian rebels) could defeat a state equipped with modern military technology and supported by a well-trained army (augmented, perhaps, with foreign mercenaries as in Libya or by the support of Hezbollah in Syria, or by the force of a foreign power such as in Hungary in 1956 or Czechoslovakia in 1968) with the will to use them, however unjustly?Footnote ¹⁸

Although he cites a number of important examples of successful armed revolutions during the second half of the twentieth century, Goodwin (writing at the turn of the millennium) was pessimistic about the prospects for organized violence against regimes with highly developed infrastructural power. Nevertheless, he recognized that even peaceful protest could help make the case for organized, offensive violence. Where the state attempted to defeat unarmed opposition by means of military force, it would run the risk of dividing the army and thus evening out the asymmetry of hardware, expertise, and manpower available between the two sides as organized violence emerged. Leon Trotsky expressed a similar thought nearly a century before when reflecting on the unsuccessful Russian revolution of 1905.Footnote ¹⁹ Recently, whereas in Egypt the decision by the army to form a defensive barrier between the protesters and the armed security forces of the Mubarak regime was crucial to the initial success of a relatively nonviolent revolution, in Libya the ability of Gaddafi's regime to secure the support of a large part of the armed forces in using military weaponry against protesters meant that the regime was able to disregard nonviolent opposition. However, the defection of some parts of the army along with their weapons combined with the independent success of rebel militias and later the assistance of foreign states acting under the authority of the UN combined to present a viable case of organized armed resistance on a large scale.Footnote ²⁰ (The jury is still out, of course, on the question of whether human rights are likely to be better protected post-Gaddafi.) So while we should certainly be sceptical about the prospects for successful armed resistance in many – or perhaps, most – cases, we ought to keep in mind that there have been numerous historical examples in modern times where it has been deployed effectively as a means of reforming, replacing, or preventing the establishment of putatively unjust institutions: aside from Libya in 2011, think, to name a few examples, of Ireland in 1921, Indo-China in 1954, Cuba in 1959, and Algeria in 1962.Footnote ²¹

In a scenario where Purely Defensive Violence and Strategic Nonviolence are both unlikely to succeed and Organized Offensive Violence is therefore the only alternative, what code ought to regulate uses of force by the Resistance? One possibility is to try to fight under the regulatory framework of the Standard JIB. For Walzer, as I mentioned, the authoritative moral framework within which all cases of military violence are to be judged is what he calls ‘the War Convention’: ‘the set of articulated norms, customs, professional codes, legal precepts, religious and philosophical principles, and reciprocal arrangements that shape our judgements about military conduct’.Footnote ²² Within this convention, the jus in bello appears as a singular, unified set of rules for engagement between armed forces among whose core components are the principle of combatant/non-combatant discrimination, the equal moral status of opposing soldiers, and the conventions governing uniforms. All three elements are implicit in the principle of non-combatant immunity (NCI) as usually understood. Fighters on both sides are recognized as having the status of ‘combatants’ and corresponding privileges and duties: the key privilege is the one permitting combatants on either side and irrespective of the justice or injustice of that side's cause to attack and try to injure or kill combatants on the other side.Footnote ²³ The most important duty is to abstain from direct attacks on non-combatants and from attacks on combatants with an excessive – ‘disproportionate’ – expected rate of collateral harm to non-combatants.Footnote ²⁴

Where Resistance is clearly unable to bring down Regime through either Purely Defensive Violence or Strategic Nonviolence, it should review the prospects for deploying Offensive Violence under the Standard JIB. I will refer to conflicts guided by the Standard JIB as ‘Conventional Wars’. This is in reference to the conventional nature of the some of the rules participants follow in such wars rather than to tactics as such. Some influential theorists have questioned the ability of non-state political movements to claim the privileges indicated by the Standard JIB in any form, notably A. J. Coates.Footnote ²⁵ Others such as Tamar Meisels and Yvonne Chiu argue that while they might operate within the Standard JIB in some circumstances, they must do so under the same constraints as those that bear on soldiers from regular armies and cannot claim additional privileges such as the right to fight without visible insignia or uniform.Footnote ²⁶ I shall defend the ability of such groups both to invoke the Standard JIB and to claim special provisions concerning limited irregular tactics, respectively, in Chapters 6 and 7 below.

Generally, Walzer treats the War Convention as the single authoritative framework for evaluating military actions. But he also indicates some other ways in which non-state groups have regulated their partisans’ actions in ways that contravene elements of the War Convention. One example comes from the French resistance to German occupation during WWII. Walzer cites a case where partisans ambushed a column of German soldiers by disguising themselves as peasants at work in the fields. As soon as the Germans were close enough for a successful attack, the ‘peasants’ drew concealed weapons and killed a number of them. For Walzer, this example is highly problematic, showing how guerrilla warfare commonly falls short of the War Convention's requirements. While recognizing that the French resistance to Nazism can hardly be dismissed as wholly unjustified, he also argues that, insofar as the occupying authorities aimed ‘at the restoration of everyday peacefulness’, they would have been justified in treating insurgents as criminals rather than prisoners of war if captured.Footnote ²⁷

Political assassination is another technique of armed resistance that Walzer mentions. This was a prominent feature of some revolutionary movements in the nineteenth century as well as of groups in Europe during the 1970s. ETA, for instance, assassinated the Spanish Francoist Prime Minister, Luis Carrero Blanco, in 1973 and in May 1979, the army chief of staff, General Luis Gómez Hortiguela.Footnote ²⁸ In Germany, the Red Army Faction used targeted assassination against judges and others seen as ‘representing’ the capitalist state in West Germany. The technique of singling out representative figures in a (putatively) unjust regime, typically individuals who are perceived to have some important degree of responsibility for oppression, is distinguished in Walzer's discussion from ‘terrorism’ proper. Whereas terrorism in his sense is directed at individuals purely on the basis that they belong to a wider target group (Protestants, Catholics, Jews, Arabs), something Jeff Goodwin calls ‘categorical terrorism’, the technique of the revolutionary assassin follows a more discriminating ethic: not that of the legitimate combatant in war, but what Walzer calls the ‘political code’. Yet, while Walzer praises revolutionaries for discriminating in this way and choosing to risk their lives in some famous cases rather than cause harm to bystanders, he nevertheless stops short of endorsing the legitimacy of the code or the violence it permits.Footnote ²⁹

If the War Convention and the Standard JIB that it enshrines constituted the only appropriate framework for regulating armed struggle involving non-state parties, then both methods of resistance – the partisan ambush and the political assassination – should properly be regarded as morally questionable at best. One exploits the just war principle of discrimination (and its legal reflection in the principle of distinction) in a way that some believe violates a principle of fairness central to the War Convention, while the other violates it by attacking individuals who are often not in any literal sense ‘combatants’ and whose liability to attack depends, Walzer thinks, on subjective political judgements. Even if we might sometimes be inclined to sympathize with resisters in both types of case based on the justice of their cause, both, therefore, are morally problematic.Footnote ³⁰

The argument that I want to introduce here and defend more fully in subsequent sections (Section 4.3 and Chapter 8 as a whole), however, is that there is at least one other alternative that might be considered in certain circumstances. Sometimes there is a prima facie case for armed resistance but other conditions are lacking that would be necessary for full-scale war or guerrilla war as provided for in the Standard JIB. The Resistance might not have the military resources needed for such a conflict or may be unable, owing to its technological inferiority and the fortified spaces available to the enemy, to access enemy combatants and attack them effectively. At the same time, it might be that the permission granted under the rules of the Standard JIB to carry out attacks that foreseeably cause harm to non-combatants collaterally leads to casualty rates caused by repressive actions or counter-insurgency among the intended beneficiaries of resistance that would exceed those even of a concerted terrorist campaign against them.Footnote ³¹ Or it might be that mobilization of the wider population to a degree sufficient to support wider armed conflict is not yet feasible (though this raises further complicating moral and political issues that I discuss in Chapter 10). Crucially, if combatants fighting on behalf of Regime chose to fight indiscriminately, disregarding their responsibilities under the Standard JIB (a common feature of wars fought against insurgencies, whether domestic or in occupied or colonized territories), this too could make resort to a further framework necessary to the Resistance. Where these circumstances prevailed, it might be that the justifying aim of armed Resistance – to defend innocent parties from harms to which they are not morally liable while pursuing legitimate political goals – would be unattainable by any of the means considered so far, Purely Defensive Violence, Strategic Nonviolence, or war under the Standard JIB. And finally, it may be the case that armed resistance within the permissive range of the Standard JIB offers insufficient chance of success against an enemy whose victory would constitute a genuine moral catastrophe.Footnote ³² In such cases, an argument might be made for setting aside the Standard JIB in favour of fighting by an alternative jus in bello built on a different, non-conventional criterion of discrimination or, indeed, of engaging in terrorist war.

If the Standard JIB offered no real opportunity to resist successfully, one alternative that Resistance might contemplate once again is to capitulate, leaving people to manage life under unjust (and now, more violent) institutions as best they can until a better chance of escape presents itself. I argue in Chapter 8, however, that there is a further code that might sometimes be justifiable as an alternative framework for guiding the actions of Resistance fighters. I call resistance in this form ‘Partisan War’ and the code which governs it, ‘The Partisan Jus in Bello.’ It would set aside the legal criterion of combatant status as the basis for discrimination and involve selecting targets, instead, on the basis of their moral responsibility for (the threat of) unjust harm. The significance of this is as follows: although combatants in a just war (i.e. a war where one side fights for a just cause while the other fights against that cause) are treated in the conventional perspective of the Standard JIB as if they were morally all in the same position, they would not be regarded as uniformly liable to attack if viewed from outside that framework, that is, based on ordinary morality. If they fight discriminately on the just side in a war, then they do no wrong thereby that could forfeit them their usual moral immunity from deliberate harm. Yet the Standard JIB treats attempts to kill them impunible and licit. (On the enemy side, too, some combatants might be genuinely (and faultlessly) ignorant that their side is in the wrong and may sometimes be morally though not legally immune from attack.) In some cases many non-combatants, however, will bear a degree of moral responsibility for wrongful threats, some to an extent that might be sufficient in some circumstances for a moral justification for attacking them. But under the LOAC, which follows the pattern of Conventional War, such attacks are rendered legally impermissible. Resorting to Partisan War involves a decision to observe a principle of discrimination built more directly on moral innocence and non-innocence rather than legal immunity and liability, permitting attacks on those who bear greatest moral responsibility where they can feasibly be identified. But it differs from Purely Defensive Violence by permitting offensive attack. I argue in Chapter 8 that such a code would continue to permit attacks on most soldiers and other armed personnel who were active in securing unjust institutions and imposing their rule but could also extend the category of legitimate targets to include political leaders in the oppressive regime or some of its key functionaries if they were responsible for Life and Limb Rights violations. Political assassination could therefore sometimes be permissible. Moreover, the Partisan JIB permits certain guerrilla tactics to the Resistance that would not be permitted in Conventional War such as the use of civilian disguise in ambush.

The two practices that proved troubling in Walzer's account thus find their place in what I argue is a legitimate moral code for armed resistance in some circumstances. But whereas the Political Code has a highly uncertain status by comparison with the War Convention in Walzer's account, my claim is that fighting according to the Partisan JIB (its close relative) could be justifiable in cases where the Standard JIB for one reason or another doesn't apply. Where the additional tactics permissible on grounds purely of moral (as against conventional) liability could contribute in some way towards the chance of successful resistance, then it might offer Resistance a fourth way of conducting its campaign against the Regime (subject to quite stringent requirements concerning the gravity of injustice faced and proportionality, as Chapters 5 and 8 indicate). This is especially likely to be true if it turned out in some cases that a smaller vanguard or ‘entrepreneurial’ resistance movement could permissibly initiate armed force in this smaller-scale manner as a prelude to expanding the struggle to support a Conventional War (or successful Strategic Nonviolence).Footnote ³³ But above all, the degree to which the regime shows itself willing in any case to violate Life and Limb Rights_{1 and 2} and to disregard the constraints imposed by IHL will be key grounds for justification.

Finally, let's suggest that our narrative of Resistance versus Regime does not succeed in spite of exploring all avenues considered so far. Attempts to overthrow the regime have been thwarted or the chances of success were manifestly insufficient when compared with the probable costs of invoking and resisting through them. Could it ever be justifiable in such circumstances to resort to deliberate attacks on civilians simply qua civilians, that is, independently of whether they were morally responsible for injustice as individuals (and given that, by definition, they are not ‘combatants’)? I will refer to this fifth possibility as Terrorist War and to the code of rules by which resistance fighters would guide their actions as the Terrorist JIB. I will defer discussion of the terms in which it might conceivably be justifiable until Chapter 9. By contrast with other philosophical accounts, I treat it as a question of justifying the availability of terrorism as a tactical resort during the entire duration of a war or during a sustained campaign. I take this to be more usual in historical cases than one-off terrorist attacks and more likely to prove effective in realizing the goals of resistance. So it is not a question of justifying a particular act but a wholesale adjustment to the rules of engagement in certain circumstances.

Hence, there are five different ways in which actors faced with violent repression and, hence, with a prima facie justification for resorting to violence could resist, guiding their actions, deciding on targets, tactics, and the limits on permissible armed force. Each has its own ‘code’ of rules defining acceptable engagements with the enemy and their limits: (1) Purely Defensive Violence; (2) Strategic Nonviolence; (3) Offensive Violence I: the Standard JIB; (4) Offensive Violence II: the Partisan JIB; and (5) Offensive Violence III: the Terrorist JIB. Some of these are likely to be relatively uncontroversial partly because they are well established in ‘common sense’ as well as political and moral theory and also because they don't enjoin resisters to engage in actions generally prohibited under IHL: in particular, the code of Purely Defensive Violence permits to resisters only the rights that they would have either in peacetime or wartime. They include the rights of individual self- and other-defence that are activated by direct, imminent, unjustified, and unprovoked threats to innocent lives. There are circumstances, clearly, in which individuals may claim these rights and act on them without thereby invoking the Standard JIB or being subject to it. Strategic Nonviolence is different from Purely Defensive Violence in that it posits a rule that is in tension with ordinary morality and peacetime individual rights, namely, the prohibition on exercising (rights of) self-defensive violence. But the claim that this code might legitimately be followed in preference to the Standard JIB in some (phases of) violent conflicts with violent oppressors is unlikely to prove a controversial claim even among those who are convinced that the Standard JIB is the only framework through which to regulate armed resistance.

More controversial are the Partisan JIB and the Terrorist JIB since both permit tactics that are prohibited under the Standard JIB and that exceed what would usually be permissible in interpersonal conflict during peacetime. The Partisan JIB permits two kinds of action that diverge from Conventional War, the targeted assassination of individuals not part of the military who are nevertheless morally responsible for threats to Life and Limb Rights, and the use of some forms of ambush that IHL deems perfidious.Footnote ³⁴ The Terrorist JIB permits the targeting of civilians in a form of categorical terrorism. To claim that the codes defining either one of these forms of war might justifiably be followed by just rebels inevitably raises a range of possible objections, which I seek to address in later parts of this book.

4.2.2 Codes as ‘frames of war’: action guidance, interpretation, and judgement

So far I have concentrated on just one function of the ‘codes’ of resistance but each is distinguished from the rest in two respects: on the one hand and most fundamentally, as I have indicated, each offers a normative framework within which some types of action are permitted, some are required, and some are prohibited. Each successive code, particularly of Offensive Violence, is distinguished by certain tactical possibilities that it offers to the resistance and that are normatively excluded by numerically prior codes (respectively, justified offensive force against unjust warriors as well as discriminate force by unjust combatants under the Standard JIB; targeted assassination of culpable non-combatants and some ostensibly perfidious ambush tactics in the Partisan JIB; finally, categorical terrorism against civilians as such).

The second function, one that is intimately related to the first and yet nevertheless worth differentiating and thinking about, is to guide third parties observing a conflict in how to interpret and judge the performances of different actors in it. Different codes sometimes describe the same events in different ways. So, for instance, the killing of legitimate rebels by state forces defending a violently oppressive regime would be interpreted as discriminate and permissible in the framework of the Standard JIB (and the LOAC) but would be seen as illicit killing, perhaps ‘murder’, under Purely Defensive Violence, Strategic Nonviolence, or the Partisan JIB. Similarly, the assassination of a key civilian political leader who was responsible for some of the unjust policies against which resistance was directed might be seen as the indiscriminate murder of a non-combatant in the framework of the Standard JIB but as a permissible and discriminate act of war under the ‘Partisan JIB’. The Code that a third party chooses as a frame of reference thus determines, in some cases, the description of a particular act; and the description determines the appropriate moral or legal evaluation of that action. In order to make the right judgements in a given case, it is therefore necessary to know which set of rules is appropriate.

The ability of opposing parties in a conflict to invoke different codes and persuade observers to interpret their actions through them thus clearly has potential political importance. Public support either nationally or internationally is often a decisive factor in determining whether a resistance group succeeds in securing its aims. Persuading the public to accept a description that leads to a sympathetic evaluation of the group's actions is therefore likely to be a key strategic priority, one that can shape both the language rebels use in attempting to justify their actions to the public and, arguably, their choice of tactics. Thus, for instance, a group that might find the direct targeting of innocent civilians expedient purely in terms of its tactical efficacy by putting pressure on its enemies to make concessions might be motivated to sacrifice this advantage and limit its attacks to targets that can plausibly be described as ‘military’ in order to validate its self-description as a discriminate, non-terrorist resistance group. Frantz Fanon appreciated the significance of this issue when he indicated the importance of ‘educating the local leaders [of the FLN in Algeria] politically […] that it is of immediate advantage to the movement that its supporters should show by their actions that they respect certain international conventions’. He advises that ‘an army which takes prisoners is an army, and ceases to be considered as a group of wayside bandits’. By the same token, the opposing authorities have an interest in denouncing even attacks on military personnel, for instance, as ‘terrorist’ and to find ways to justify the description.Footnote ³⁵

One instructive example is that of the Provisional IRA which commonly targeted members of the Northern Irish police (the Royal Ulster Constabulary), the British Army, and the Ulster Defence Regiment, describing them as ‘legitimate targets’. As David McKittrick has said, ‘Sinn Fein, in its efforts to build a political machine in both parts of Ireland, has [always] been concerned to project IRA violence as the clinical and carefully directed use of force.’Footnote ³⁶ The British government, by contrast, insisted that such killings were acts of ‘terrorism’, especially when the victims died while off duty, as was often the case. In this light, the IRA Hunger Strikes of 1980–81 may be seen as part of a rhetorical contest in which prisoners sought to redescribe themselves as having the ‘special status’ of combatants in a political struggle rather than ordinary criminals to whom no code of armed conflict properly applied. Another interesting case is the attempt by the Red Army Faction to have its captured members recategorized as prisoners of war and brought under the Geneva Conventions in 1976–77. Their aim was to redescribe the struggle as an international conflict between, on the one side, the RAF and the anti-colonial movements with which it identified itself, and on the other, the German Federal Republic with its allies. As Andreas Baader explained it, the aim was that ‘people will be mobilized and radicalized around the critical issue: that the state is at war […] and, therefore, is in a dialectic that – because war frames the question of legitimacy along military lines – destroys the ideological justification of the constitutional state itself’.Footnote ³⁷

All else being equal, it will generally be easier to persuade third parties to approve the legitimacy of rebel tactics if appeal is made to the more familiar frames of reference, Purely Defensive Violence, Strategic Nonviolence, and the Standard JIB. It is also likely to be easier to substantiate the justification claimed for particular actions under these codes than under the other two. In particular, with the recent use of camera phones at the scene of confrontation in Iran, Egypt, and elsewhere, it is at least sometimes possible to show cases where personal self-defence (i.e. Purely Defensive Violence) was appropriate (though these kinds of source need to be viewed with due care and scepticism). Also with Strategic Nonviolence, as long as a movement generally persuades its followers to avoid responding to threats or provocation using force, it will be easier to persuade third parties to accept the ‘nonviolent’ (re)description of its tactics.Footnote ³⁸ At the same time, regimes faced with nonviolent or relatively peaceful demonstrations have a motive to try to tarnish the reputation of the opposition by alleging the involvement of ‘terrorists,’ as, for instance, in Yemen under the rule of Ali Abdullah Saleh, who consistently argued that opposition was infiltrated by al Qaeda, and Bashar al-Assad, who has persistently used an ‘al Qaeda narrative’ when describing the opposition in Syria (following which, the facts seem to have caught up with the story).Footnote ³⁹

Conventional War, however, marks that point in the sequence of escalation where movements have to start working harder to justify themselves. Rebels must persuade third parties to set aside ordinary, peacetime morality and legality and to view the violence as permissible warfare that must be interpreted and judged with reference to the Standard JIB (and, hence, the LOAC). Otherwise the same acts will be condemned as criminal violence and murder. As I will argue in Chapter 6, the ability to invoke the LOAC through engagement in a war governed by the norms of the Standard JIB in this way sometimes requires the fulfilment of a crucial prior criterion of ‘legitimate authority’. That is, the group needs to be able to differentiate itself from ordinary (i.e. non-political, unjustified) criminal organizations and from marginal political groups that also happen to use violence. It has to show that in deploying arms against its enemies it can rightfully claim the rights of belligerency as set out in the Standard JIB and bestow upon its partisans the privileges and immunities (as well as the responsibilities) of ‘combatants’.

But as difficult as it may sometimes be for rebels to convince others that they are engaged in legitimate acts of Conventional War, it will be even harder for those engaging in Partisan War. The Standard JIB is widely recognized publicly as a normative framework, and its central principle of combatant/non-combatant discrimination is generally accepted. By contrast, to justify the targeted killing of culpable non-combatants in a Partisan War would require much more careful attention from the public and a much more receptive attitude towards the group in the first place. This is why armed groups like the Red Army Faction made such efforts to explain the justification for individual assassinations in public declarations.Footnote ⁴⁰ While the Partisan JIB can quite readily be reconciled with ordinary moral intuitions about innocence, moral responsibility, and liability to harm, as I shall argue in Chapter 8, it is nevertheless less recognizable by the general public than the Standard JIB and its legal relative in the LOAC/IHL. If a resistance group is able to fight effectively within the Standard JIB, therefore, then the job of persuading observers to interpret its actions in a sympathetic light will be easier than if it assassinates political leaders, even if the latter are morally legitimate targets. Any offensive force that deviates from the Standard JIB will quite likely be depicted by the authorities as ‘terrorism’.

Finally, if Terrorist War were ever justifiable morally, then the task of spokespersons for the resistance would be hardest of all. The only way to woo third parties without distortion of facts is to try to demonstrate why the use of terrorism is the only viable means of resisting in the circumstances. What is no longer possible is to try to redescribe their actions in an honest way as ‘non-terrorist’. In reality, practitioners of terrorism are more often likely to try to convince readers of their pronouncements that civilian targets were sufficiently culpable to merit attack on moral grounds, thus redescribing what is actually a case of Terrorist War as one of morally discriminate killing.Footnote ⁴¹

4.3.1 The doctrine of moral equality

One way to argue that all wars ought to be regulated by the Standard JIB is to maintain, as many have done, that it reflects the requirements of ordinary morality in a more or less direct way. This is one of the ideas that defines a ‘traditional’ view in just war theory as some recent revisionists call it. If the Standard JIB is on all fours with ordinary morality, rather than being partly or wholly conventional in nature, then it is less likely that legitimate alternative sets of in bello rules exist.

The central component that a defence of the Standard JIB along these lines must focus on is the egalitarian principle according to which opposing soldiers are granted the same war privilege and the same constraints against attacking non-combatants. Many philosophers now regard this as at least partly conventional in nature, granting the war privilege and post bellum impunity to soldiers who abide by its in bello rules in return for the immunity of those designated non-combatants and regardless of the morality of war under the jus ad bellum. By contrast, the ‘Doctrine of Moral Equality’ (DME) – as Walzer calls it – holds that all soldiers are morally liable to attack by their enemies in war irrespective of whether they fight for a just cause or an unjust one, while non-combatants are morally immune.

A series of arguments have been offered at one time or another for the DME. One defends it on the basis of mutual threat: combatants generally pose a material threat of harm to one another; all therefore have a right to use force in self-defence.Footnote ⁴² But civilians do not. Therefore while it is morally right that opposing soldiers in war ought to be permitted the use of force against each other, they are not permitted to take aim at civilians. Civilians are ‘innocent’ in the etymological sense of the word, that is, presently harmless.Footnote ⁴³ Another argument is made from what Walzer calls ‘common servitude’: in an era of mass conscription such as the first half of the twentieth century, soldiers of many nations increasingly came to be subject to similar kinds of coercion, having been recruited to their nations’ armies by threat of force and being subject to a regime of threat during the war. All, therefore, were absolved of any crime in fighting since none were morally responsible for the threats they posed.Footnote ⁴⁴ Third, common ignorance about the causes for which they fight has also been cited in defence of soldiers who participate in war: lack of information or of a competence to judge the validity of their countries’ claims about justice could mitigate responsibility.Footnote ⁴⁵ Finally, the authority of the state could be cited as a basis for arguing that the duty of soldiers to obey outweighs their responsibility towards those whom they fight; so even if it turned out that they were fighting in an unjust war, it might be argued, their political obligations still override their moral duty not to attack and kill just warriors. A classic statement of this view is in Kant's essay, ‘An Answer to the Question: What is Enlightenment?’: ‘For many affairs that serve the interests of the commonwealth a certain mechanism is required’, he writes, ‘by means of which some members of the commonwealth must play only a passive role, so that they can be led by the government in the pursuit of public ends by means of an artificial unanimity, or at least be kept from undermining these ends. In these cases, of course, one may not argue, but rather must obey. […] It would thus be very harmful if an officer who receives orders from his superiors were to publicly question the expediency or usefulness of his orders; he must obey.’ While soldiers and other post-holders in the state ought to be permitted to make ‘public use’ of reason by ‘commenting as a scholar, on the mistakes in the military service and submitting these remarks to judgment by the public’, they are prohibited from exercising ‘private use’ of reason by putting their arguments into action in disobedience.Footnote ⁴⁶ One way to support the idea that soldiers are justified in obeying the orders of their governments to fight even in unjust wars is from epistemic concerns: if soldiers are generally in a much less privileged position epistemically than their political leaders, then it is reasonable to demand that they defer to the government for judgements about the justice of resorting to war.Footnote ⁴⁷

The problems with these arguments are now familiar from a wave of revisionist criticism. None of the moral arguments offered in defence of an egalitarian ethic can really show how it is generally – or, as the doctrine claims, universally – justified for those fighting for an unjust cause to kill at all in war. The key premise of this line of criticism is that a person cannot forfeit their moral immunity from intentional attack unless they are already responsible for contributing to a wrongful threat of harm against their attackers or those on whose behalf they act. In any truly just war, therefore, the claim of the just warrior to immunity from intentional harm seems from a purely moral point of view to remain intact in spite of the fact that he threatens to kill. The threat he presents to aggressors (domestically or internationally, depending on the type of war) is justified on the hypothesis that his is a just war and it therefore entails no moral liability to attack in return. So the first argument in favour of the moral grounding of the jus in bello falls down. Once it is accepted that the discriminate just warrior remains morally immune and that his claim against intentional attack retains its full force, the other arguments prove too weak too: in many cases, unjust warriors fight without being subject to a severe coercive regime, but even if they had been subject to one, unjust warriors would still be at best excused of their crime or enjoy some partial mitigation, but would not be justified in killing in the way their opponents were. And the excuse is unlikely to be comprehensive enough: a threat to the life of the unjust warrior would probably not wholly wash his hands of the crime of killing even one other innocent just warrior, and soldiers are usually expected to try to kill more than one. Epistemic arguments have some salience in providing excuses for unjust warriors – or ‘subjective’ or ‘evidence-relative’ justifications – but this factor varies according to the degree of indeterminacy of the justice of a particular war, the political conditions and availability of information in the unjust warrior's state, and so on. And the more egregious the crime of war by the unjust side – the more blatant its aggression or the more horrifying the human rights violations it perpetrates – the less reason there will be to accept that unjust warriors could be epistemically justified in believing their side to have just cause.Footnote ⁴⁸ Finally, it is doubtful that the soldier's duty to obey superior orders outweighs the duty of non-harm to the morally innocent just warrior whom he is ordered to kill, especially (but not only) in states that cannot claim to be committed to the public, progressive ends indicated by Kant. And even the most optimistic variants of the argument for deferring to government on questions of the justice of war admit that this cannot be true in many undemocratic regimes (to say nothing of the doubts that can be raised about the competence of democratic governments in this regard).Footnote ⁴⁹

Through its central principles of combatant/non-combatant discrimination and the DME, the Standard JIB therefore permits actions in many cases that are not justified from what McMahan calls a ‘deep’ moral perspective, that is, in terms of ordinary morality and basic intuitions about innocence, moral responsibility, immunity, and liability to harm. Centrally, the killing of those who fight in good faith for an objectively just cause by those who fight for a side aiming to uphold or perpetrate injustice is treated as licit in this framework and such actions are rendered legally impunible in the LOAC which follows it. But the Standard JIB also imposes a prohibition on action by just warriors that modifies the rules that might arise from ordinary morality in a further way: if those who held the greatest moral responsibility for the injustices leading to war serve in civilian roles and are designated ‘non-combatant’ by their state, then the Standard JIB strictly prohibits attacking them intentionally. Yet, on the same premise that revisionists cite to condemn the killing of just warriors by unjust, the targeting of at least culpable civilians would seem to be permissible to just warriors in certain circumstances (and to be more justifiable, even, than the killing of some unjust warriors, e.g. those with significant excuses for fighting.) To this extent, then, the Standard JIB is in tension with more basic moral principles in two important respects. If it has validity as a framework for guiding the actions of participants in some wars, it must be on some other, perhaps conventional basis as a number of philosophers have suggested. This raises two questions to which I now turn in Sections 4.3.2 and 4.3.3: if this analysis is correct, then what sort of convention is the Standard JIB? And, if it is a convention, then by way of specifying the circumstances in which it is binding, what can we say about the limits of its application? In what circumstances, in other words, might it be permissible to regard some of its demands as void?

4.3.2 The Standard JIB as a convention

According to Jeremy Waldron, there are three possible views on the principle of NCI that forms a core part of the Standard JIB: it must either be a ‘Lewis convention’ (following the precise, technical account offered by David Lewis), not be conventional at all, or else follow the pattern of some ‘looser’ notion of convention.Footnote ⁵⁰ He grants that the principle may be partly conventional in nature, but argues (rightly) that it cannot be interpreted as a Lewis convention, which is distinguished by two essential parts:

1. 1. The ‘convergence of interest feature’: ‘it is better for all concerned if all or most of the others follow some rule; and if all or most of the others follow rule R, it is better for oneself to follow R’. Hence, this feature is defined in such a way that the rule about driving on the left of the road in the UK has the right feature insofar as no one has a private interest in free-riding on other people's compliance with the rule.

2. 2. The ‘arbitrary alternative feature’: alternatives are both available and arbitrary ‘in the sense that any differences between them pale in comparison to the importance of following one of them or the other as opposed to no such regularity’.Footnote ⁵¹ The choice should be substantively a matter of indifference (again, illustrated by conventional decisions on which side of the road to use for driving).

Waldron argues that if NCI is a convention, it is one that lacks the first feature because it is sometimes in the interests of a party to free-ride on the compliance of others. Moreover, even if others don't abide by the rule, he maintains, it is still better to respect NCI oneself.Footnote ⁵² I'm inclined to think, however, that there is an equally striking divergence from Lewis on point of (2) since it is less clear to me that there are arbitrary (i.e. equally attractive and feasible) alternatives to NCI, which I will argue suggests it is something closer to a Humean convention.Footnote ⁵³

It is a common – and I think reasonable – belief that the principle of NCI (plus other elements such as prisoner-of-war status) is the most humane compared with feasible alternative conventions, so to choose that rule rather than another can't really be seen as purely an arbitrary matter.Footnote ⁵⁴ Clearly a rule that followed the distribution of moral responsibility (or culpability) for the injustices causing a war – on McMahan's approach or Fabre's, for instance – could not be offered as a candidate for an alternative ‘convention’; nor would it be said by proponents of such an analysis of war that choosing between NCI and one of these other rules is a matter of indifference, prior to conventional agreement. So this sort of view doesn't fill the gap. It might be objected (following George Mavrodes’ suggestion) that the Biblical story of combat by champions would be more humane still. But this alternative is unlikely to be viable for other reasons. Mavrodes, for instance, thinks that ‘there is almost no chance that such a convention would actually be followed’.Footnote ⁵⁵ Yitzhak Benbaji suggests this is because, unlike conventions such as wearing gloves during a boxing match, agreeing to resort to the use of champions is likely to have a distorting effect that will undermine compliance. He argues that the ‘Goliath convention’ alters the chances each side has of victory, regardless of the justice of its cause, making it ‘unfair’. By contrast, the principle of NCI is closer to the use of boxing gloves: each side ‘is able to do less damage to the other, but each also suffers less at the other's hands’. The constraint is equal and therefore fair since, although ‘both sides are subsequently denied important opportunities for military advantage, they are denied them to the same degree’.Footnote ⁵⁶ There may be some circumstances in which one side finds that the convention effectively eliminates the ability to fight with any chance of success, a possibility to which I will return.Footnote ⁵⁷ But NCI is generally thought to avoid the problems of the Goliath convention across a sufficiently wide range of possible wars to have a chance of effectiveness.

The most distinctive alternative that Waldron suggests is the following: ‘we could have a rule that no women and children are to be targeted, for example…’.Footnote ⁵⁸ I doubt, however, that we could really say that it was a matter of indifference whether we adopted this rule or NCI as it currently operates in the LOAC. Under Waldron's alternative, a much wider range of individuals would be liable to attack in war than is presently the case. And their liability would track no morally significant, non-arbitrary feature of the category to which they belong, that is, criteria such as guilt or culpability, moral responsibility, posing a direct threat to someone, causal contribution to injustices or threats posed by others, having means of defence, and so on; rather, it would take unchosen and inescapable group membership as such as a criterion for targeting.Footnote ⁵⁹

Some other variations are worth mentioning briefly in order to introduce a further nuance to the claim that the Standard JIB is in part a convention. Consider the following ways of distinguishing the category of legitimate targets: being a threat, being a contributor of a certain kind to threat (based, say, on making a non-welfare-based contribution to the ability of soldiers to fight),Footnote ⁶⁰ being a member of the armed forces, wearing a uniform, carrying arms, and so on. One could imagine a range of alternative rules by which to decide who is and who is not a legitimate target in war and these might seem like the arbitrary alternatives that would be needed to fulfil the second feature of a Lewis convention. I think, however, that these variations are not candidates for alternative conventions to NCI – or not, at least, in the relevant dimensions – but respond to the need to specify more concretely how NCI itself will be applied. The term ‘combatant’ is notoriously imprecise (as is the alternative ‘participant’ or others like ‘contributor’). In order to offer action-guidance to opponents in war, we need an account of who has the status of combatant and in which circumstances or by what means. Different alternatives could be proposed, and the question of which should be chosen is a contentious matter, philosophically. Some elements are less controversial, such as that combat soldiers are combatants, others more so, such as whether armaments workers contribute in such a way to combat as to make them liable to attack, and so on. To make a convention of NCI workable, an answer must be decided upon between different (potential, future) parties to war. But whichever answer is agreed, the basic elements of the convention remain the same: the duty to discriminate, the liability of combatants and immunity of non-combatants, and symmetry between opposing sides.

So, on the one hand, we can say that deciding on a particular way of filling out the categories of combatant and non-combatant is conventional insofar as it is a non-natural, agreed standard chosen from a set of alternatives between which choosers might otherwise be indifferent (more or less). But on the other, this addresses only a layer of detail within a larger, more general convention that endures between such variations. Revisionism in contemporary just war theory challenges not only the level of detail but also the convention as a whole, not just one or the other way of defining combatant status, but the very idea of regulating a war within the convention of NCI.

Arguably, then, NCI meets neither of the distinguishing features of Lewis’ conventions. So, if it is a convention, then what sort is it? I think a better account can be derived from Hume's analysis of the foundations of a conventional principle of justice as it determines the principles appropriate for the distribution of material goods. In An Enquiry Concerning the Principles of Morals, Hume speculates:

Instead, Hume argues, it is necessary to establish as fundamental principles of justice, conventional rules that can be operated by fallible human beings without descending into civil conflict. Let ‘Natural Moral Principle’ stand for the ideal of rewarding moral goodness in Hume's account while ‘Peace’ signifies the maintenance of society (avoiding its ‘total dissolution’). The relevant type of convention arises in the following way:

1. I. All have a fundamental interest in the best possible satisfaction of the requirements of Natural Moral Principle and Peace.

2. II. Rule R is best for optimizing fulfilment of Natural Moral Principle in particular cases. In fact, justice might be ideally realized by universal satisfaction of R, assuming R could ever be satisfied from an agent-neutral perspective.

3. III. The problem with R is that uncoordinated action based on Natural Moral Principle by the rule R is likely to cause widespread conflict, and no authoritative means of coordinating judgement under R is possible for epistemic reasons.

4. IV. Whereas all might be said to value both Natural Moral Principle and Peace equally, ceteris paribus, one fact forces them to prioritize Peace over Natural Moral Principle, namely, that a collapse of Peace makes any fulfilment of Natural Moral Principle impossible; whereas, some degree of compromise on honouring Natural Moral Principle is compatible with Peace.

5. V. Only an artificial rule – call it, R′ – is capable of achieving Peace (though it has a more problematic relationship with Natural Moral Principle, offering only a second-best realization). (R′, incidentally, is the only candidate for such a convention since it is the only practicable one, in Hume's account: ‘Tho’ the rules of justice be artificial,’ he writes, ‘they are not arbitrary’.Footnote ⁶²)

Roughly speaking, R in Hume's account seeks to distribute material goods according to desert, whereas R′ allocates goods according to more easily verifiable criteria such as established possession. Hume's worry is that, whereas natural moral sentiment inclines us to R, the lack of an authoritative epistemic vantage point from which to reach authoritative judgements under R will mean that attempts to follow this rule lead us into a kind of Hobbesian conflict. So we adopt a second-best principle, R′, that uses epistemically less opaque criteria to make decisions and agree to allow R′ to override R where the two point towards different allocations. Following Rawls, we might therefore say that R′ replaces desert with ‘legitimate expectations’.Footnote ⁶³ So, someone couldn't expect to gain a legal hearing in a society organized around these conventions for a claim that they didn't get the material goods they deserved (or that someone else did get something they didn't deserve), but they could make a complaint under the fundamental conventions of justice about expectations generated by the conventions themselves.

I think the status and force of NCI should be interpreted in similar terms from a revisionist perspective. In that case, NCI is represented by R′, whereas a rule such that targeting should track culpability or moral responsibility for the relevant set of injustices is R. It might seem like ideally the best way to honour the immunity of the innocent is to follow some version of R. But it is generally recognized that this would be untenable as a rule for wars in general, potentially leading to a deepening and widening of armed conflict to such an extent that it would vitiate any expected gains in fulfilment of Natural Moral Principle. That is, for instance, on the one hand, just warriors might target more discriminately under R than under R′ in moral terms in a way that widens their range to include some culpable non-combatants, but because unjust warriors are likely to believe mistakenly that Natural Moral Principle requires them to kill ‘culpable’ non-combatants on (what is from a fact-relative perspective) the just side, it may lead to more fact-relatively innocent people being killed than adhering to R′ instead. So, even if the just warriors’ actions were guided more directly by considerations of Natural Moral Principle under R than under R′, any actual gains in justice would be vitiated by the failure to contain war within regulated limits to the greatest extent possible (a goal which occupies the place of Peace in the analysis of Hume above). To put it another way, whereas the actions of just warriors would conform more directly with Natural Moral Principle insofar as they follow a rule (R) that reflects its requirements directly, they would realize the purposes of Natural Moral Principle less well insofar as potentially a greater number of innocent people would be killed as a result than if they reverted to R′.

4.3.3 Applicability of the Standard JIB convention

Based on this alternative Humean interpretation, we can outline the rationale for a second-best rule and, hence, indicate the sorts of case in which it applies by thinking about the outcomes likely to flow from different choices a just side might make in setting and following a rule to follow (where R′ is NCI and R is discrimination based on moral responsibility for injustice). Think of the following as possible just wars, each distinguished by the principles of conduct followed by different sides:

1. (1) ‘McMahanian’: R for just warriors; no permissible violence for unjust warriors in pursuance of their side's war aims.

2. (2) R followed by just warriors; R′ followed by unjust warriors.

3. (3) R for just warriors; unjust warriors fight without restraining rule.

4. (4) R for both sides, that is, both assume that they have just cause, that their opponents wrongfully oppose them, and that R permits them to discriminate amongst their enemies on grounds of moral responsibility for what they believe to be an unjust war.

5. (5) No rule restraining either just or unjust warriors.

6. (6) Conventional War: both sides restrained by R′ (i.e. both sides constrained by Standard JIB).

7. (7) R′ restrains just warriors; unjust warriors follow R (in the belief that they have just cause).

8. (8) R′ restrains just warriors; unjust warriors fight without restraining rule.

Number (1) follows the logic of the ‘deep’ morality of war on McMahan's analysis; (2) bears some resemblance to what Rawls proposed in A Theory of Justice where he wrote that, ‘[e]ven in a just war certain forms of violence are strictly inadmissible; and where a country's right to war is questionable and uncertain, the constraints on the means it can use are all the more severe’.Footnote ⁶⁴ In cases where I have indicated that unjust warriors will follow R, I mean that they will try to do so based on the false assumption that theirs is a just cause and that enemies who fight against it are morally liable to harm in defending it.

My question concerns the decisions that just rebels ought to make in choosing which principle to follow, R, R′ or no restraint at all. They should be guided in doing so by two aims: first, to fight in a manner that is as consistent as possible with moral principle; and second, to fight in a manner that will preserve the lives of the innocent as much as possible. Each aim may limit, to some degree, the pursuit of the other in different circumstances. One way, for instance, in which the aim of preserving the lives of the innocent may restrict the degree to which rebels follow moral principle is if doing so is likely to motivate more unrestrained fighting by their enemies. Rebels are therefore constrained in their choice of guiding principle by the response it is likely to face from their enemies and the sort of conflict that will emerge as a result. In some instances, as I will argue in later chapters, a non-conventional alternative such as R is the appropriate rule for just rebels to follow, particularly if they fight an enemy that refuses to abide by R′. If the rebels continued to adhere to R′ in such circumstances, it would put them in the positions indicated by (7) and (8). For now, let me indicate why in other types of case where offensive violence is necessary for the resistance, conventional rule R′ is the best rule to follow.

The alternative that most directly tracks the requirements of moral principle is number (1) but I presume this is a non-starter since, whereas just warriors might agree to fight under the restraint of R, it is unlikely that their unjust opponents will forswear the use of arms. We would have to suppose that unjust warriors, by hypothesis, are prepared to fight even in the face of a prohibition on doing so. So number (1) is likely to decline into number (3): if they are not granted any use of force at all, the unjust warriors might reason that they may as well be hung for a sheep as for a lamb and fight indiscriminately when it is most expedient to do so. Moreover, if the unjust side thereby lapses into terrorism, the likelihood that resorting to terrorism in return will be the only way for the just side to win might increase and, with it, the chance of being able to do so justifiably: hence, number (3) might deteriorate into number (5). Number (1), we might say, is therefore radically unstable. While it tracks the demands of moral principle closely, it is likely to fail to realize the moral purpose of preserving as many innocent people as possible.

The same is true to a degree of number (2): it is plausible to imagine that in many or most cases, the unjust warriors will either sincerely believe or act consistently with the belief that they have justice on their side (even if they have doubts); if they see their opponents setting aside NCI in favour of targeting people based on claims about moral non-innocence, then they are likely to do the same either by way of reprisals or as a means of claiming back the additional advantage their opponents have seized. We thus see a shift from number (2) to number (4). If it is likely that each side will regard attacks on its own putatively non-innocent non-combatants as terrorist outrages, there is also an increased risk at (4) that mutual provocation and retaliation will lead a descent towards (5). So, on this account, the two alternatives that most closely track moral requirements of justice – (1) and (2) – lack the ability to contain the tendency of conflicts to escalate, leading in a range of cases to a deterioration into (3), (4), or (5). These last three alternatives are even more radically at odds with Natural Moral Principle – and are likely to produce a greater number of injustices as an outcome – than number (6), that is, Conventional War governed by the Standard JIB.

The argument for choosing to follow R′ rather than R is that R′ is likely to offer the best chance of limiting the destructiveness of war in circumstances where the enemy is likely to follow it too. In other words, one should aim for a war modelled on (6), if possible. R′ – the Standard JIB and the principle of NCI – establishes neutral, symmetrical criteria for distinguishing legitimate targets, thus avoiding the risks of deterioration inherent in (1) and (2). Insofar as a stable regime of rules that binds both sides in this way can be expected to cause fewer innocent casualties than the alternatives considered so far – (1), (2) and the wars they may lead to: (3), (4), or (5) – it is likely to be the best option all things considered where it is available. Rebels should therefore have a preference for following R′ (NCI) in those circumstances.

However, by the same token, (6) is not possible where a just side is confronted by an enemy that refuses to be constrained by R′. In such a case, adhering to R′ would put the resistance into one of the scenarios represented by (7) or (8). The question then is whether resisters are obliged to endure the disadvantages of (7) or (8) as against rejecting R′ and adopting R (or even resorting to Terrorist War) if doing so offers the best chance of success in defending Life and Limb Rights. I return to these questions in Chapters 8 and 9.

4.3.4 Limits of the Standard JIB's application

To summarize, then, the standard jus in bello is partly conventional in nature. Some of its tenets (sometimes) are in line with deeper moral principles but other conventional tenets are in tension with them. If adherence to this framework is to be legitimate, then conventional elements must be justified by reference to basic principles of morality in a more indirect way. On the account offered above, these conventional elements seem to be justified on two assumptions: first, that it is likely that those on the unjust side will wrongly suppose that they are in the right or, where they realize that they are not, that they are likely to act as if they had a just cause; and second, given that this is generally the case, that fighting a war within the egalitarian framework of the Standard JIB will reduce the number of wrongful harms suffered overall as compared with fighting it through an asymmetric ethic that more directly tracks the moral innocence or non-innocence of individuals.

While the first assumption clearly has to be accepted in any conflict (since to do otherwise would be to assume that the unjust side did not have sufficient will to fight), it seems to me, however, that the second need not always be true. In conflicts where it isn't, the rationale for upholding the conventional dimensions of the Standard JIB is moot. My argument, on this basis, is not simply that exceptions might sometimes be permitted to those generally acting within the Standard JIB; but that in conditions where it is not likely to reduce the level of relevant harm in a particular conflict, or where it achieves reductions in Life and Limb₂ harms at the expense of a failure to resist a greater number of Life and Limb Rights₁ violations, an important barrier to considering alternative rules of engagement – rules that do not reflect the conventional egalitarianism of the Standard JIB – is removed.

The conditions that would need to be fulfilled before the Standard JIB could conceivably be set aside in favour of another code for regulating the organized use of force by the resistance need a good deal more attention, and I will fill them out in detail in Chapters 5, 6, 8, and 9.