The Inconsistent Reduction: An Internal Methodological Critique of Revisionist Just War Theory

Abstract

This article argues that the reduction of the morality of killing in war to the morality of killing in self-defense by ‘reductive-individualist’ revisionist just war theories is inconsistent, because when those theories apply the moral notion of self-defense to the morality of killing in war, they do not preserve the two conceptions of the “individual” inherent in this notion. The article demonstrates this inconsistency in two steps: First, it disentangles the two conceptions of the individual inherent to the notion of self-defense, namely (1) that the individual is an “entity” potentially bearing a right to self-defense (unlike, e.g., groups) and (2) that the individual is a “particular,” where “particular” signifies that every human is different from every other human. The conception of the individual as a “particular” is tied to the idea that a justification grounded in a rule of self-defense is necessarily “concrete,” in the sense of referring to individually given and specific perceptions or cases, as opposed to “abstract,” in the sense of being detached from specific perceptions or cases. The article then demonstrates that reductive individualism reflects the first notion of the individual, but not the second. Due to the “loss” of the individual as a “particular”, the reductive-individualist reduction of the morality of killing in war to the morality of killing in self-defense is inconsistent, and hence its justification of killing in war grounded in self-defense is not concrete. Since such a justification must be concrete, reductive individualism cannot offer a justification for belligerent killing.

1 Introduction

Just war theories aim to explain whether—and if so why and when—war, and the systematic killing it involves, is morally justified.¹ The present article offers an internal conceptual critique of the particular methodological approach that is predominant in many contemporary representatives of these theories. The methodological approach is called “reductive individualism” and the contemporary just war theories that apply it are often referred to as “revisionist” just war theories (henceforth “revisionism”).

Revisionist just war theories reject either one or both pillars of the law of war: some deny that national defense can justify war (Rodin, 2002). Others deny that the same rules apply to combatants fighting on both sides of a war; rather, those fighting for a state whose war has a just cause, generally, retain their right not to be killed by their unjust opponents, but are entitled to kill them. Their unjust enemies, in turn, generally, lose their right not to be killed, and are themselves not allowed to kill the just combatants. The article focuses on the second group of revisionist arguments and refers to its representatives as revisionists. Revisionists offer different arguments for this “moral asymmetry” between just and unjust combatants. Hence, assigning a consensus view to them would be mistaken. Nevertheless, many famous revisionists insist on methodological reductive individualism (McMahan, 2004, 2006a, and, 2009; Rodin, 2008; Frowe, 2014; Fabre, 2012; Haque, 2017).²

Reductive individualism is a way of thinking about the morality of killing in war that “reduces” this morality of belligerent killing to the morality of harm and killing in self-defense in ordinary life. This article argues that the reductive methodology is inconsistent, because in applying this concept to killing in war, it fails to adequately reflect the notion of the “individual” that underlies the moral right to self-defense in ordinary life.

This conceptual critique unfolds in two steps. The first step involves an analysis of the notion of the “individual” inherent in the moral concept of a right to self-defense (Sect. 3). This article will argue that there are, in fact, two such notions: The first understands the individual as the proper or sole “entity” that is capable of bearing a right to self-defense, in contrast to other entities, such as collectives (e.g., states). The second understands the individual as a “particular,” in the sense that each human individual is different from all others, in contrast to the “general-abstract.” The latter conception of individuality also reflects the idea that a justification of harm by recourse to the concept of self-defense can always only ever be “concrete,” i.e., refer to a specific perception or case, as opposed to “abstract,” i.e., detached from specific and individually given perceptions or cases. Furthermore, a concrete justification cannot rely on mere speculative presumptions about a specific case of lethal harm. Consequently, self-defense—understood as a rule potentially governing many individuals—can only ever provide a justification for harm on the level of the individual as a particular. This section also describes the moral concept of other-defense and its relation to self-defense. Reductive individualist just war theorists often refer to both concepts, though with a focus on self-defense (e.g. Fabre, 2012, p. 55, 56; Frowe, 2011, p. 24, 84 ff.). Furthermore, this section discusses a limitation of the methodological critique put forth in the article. The critique addresses authors who endorse a so-called “internalist” account of liability, and misses the mark for “externalists.”

The second step in the critique involves demonstrating that the understanding of the “individual” that reductive individualists rely on when reducing the morality of killing in war to the morality of self-defense considers the first notion, but not the second one, because these theorists homogenize the moral status of individual combatants on the basis of a presumption that is divorced from any situational concreteness (Sect. 4). This is where the methodological inconsistency in the revisionist reduction is to be found: Neglecting the second notion is precisely what makes reductive individualism a defective methodology. Finally, and most importantly, since a concrete justification that appeals to the rule of self-defense cannot rely on speculative collectivizing assumptions, reductive individualists cannot justify killing in war.

2 Reductive Individualism and the Right to Lethal Self-Defense

“Reductive individualism”³ is a theoretical approach that reduces the morality of killing in war to the morality of harming or killing in individual self-defense and other-defense during peacetime—whence its name. ⁴ In other words, the moral rules of war are continuous with the moral rules of ordinary life. For reductive individualists, war is, morally speaking, an extension of ordinary life.

Reductions can be a very handy theoretical tool, allowing theorists to do more with less.⁵ When it comes to the morality of killing in war, they allow theorists to fall back on normative tools developed to address acts of harm and killing (i.e., individual self-defense) in peacetime.

The argument presented in this article stresses that reducing the morality of killing in war to the morality of individual lethal self-defense in ordinary life is also necessarily individualistic. The reason for this is that self-defense in ordinary life is morally⁶ conceptualized in inherently individualistic terms.

The following sub-section describes and explains the notion of the “individual” that underlies the moral concept of self-defense, arguing that the moral right to individual self-defense incorporates two distinct conceptions of the “individual.” On the one hand, the moral right to self-defense entails an understanding of the individual as an entity that can be a bearer of the right to self-defense in certain situations. On the other hand, it also involves an understanding of the individual as a particular or concrete individual. The latter understanding of the “individual” inherent in the moral concept of a right to self-defense goes hand in hand with the idea that a justification of harm grounded in the concept of a right to self-defense can only ever be concrete, never abstract. Once we have made some conceptual clarifications in the next section, we will proceed to an analysis and critique of the use of the concept of self-defense by reductive individualists.

3 Two Conceptions of the “Individual” Inherent in the Right to Self-Defense

3.1 The Individual as a Right-Bearing Entity

Violent or lethal self-defense is a violent or lethal act committed to prevent harm to oneself. ⁷ The normative basis for such an act is a right to life and to bodily integrity. The reason is that the right to life has as a core element the claim against every other person that they not take one’s life or interfere with one’s bodily integrity.⁸

If You attack Me, the act I take to ward off Your attack could count as an act of self-defense. Self-defense is regarded as a source of justification for acts that are usually impermissible, namely those that involve harming other people. This is why moral philosophers generally understand self-defense as a right held by an individual person: If I have a moral right to self-defense, then my act of self-defense is also morally justified. ⁹

The rationale behind understanding self-defense as a right can be put most simply as follows: Every person has a right not to be harmed. However, a person can forfeit this right against being harmed, in which case harm against her can become permissible.¹⁰

Consider the following scenario: You have not done anything other than peacefully being You, meaning that You have a right not to be harmed. I attack You, knowingly, willingly, and for no reason. ¹¹ That is, I attack you without any justification—meaning that my attack is unjustified.¹² In doing so, I impermissibly infringe (or violate) Your right not to be harmed. In infringing Your right, I forfeit My own right not to be harmed. In forfeiting My right not to be harmed, You acquire the right to ward off My attack in self-defense. I, in turn, am not allowed to defend Myself against Your self-defense against Me, because by attacking You, I have renounced this right. You and I stand in a bipolar, asymmetric, normative relationship with regards to the right to harm one another.¹³

This description reflects the first understanding of the “individual” inherent in the moral concept of a right to self-defense: self-defense is a right held by an individual person. It is not a right held by collectives, states, firms, villages, or families.

(1) The entity that bears the right to self-defense is a human individual.¹⁴

3.2 The Individual as a Particular

Inherent in the moral notion of a right to self-defense is another conception of the “individual.” Let us consider once again the above example involving You and Me: You have a right to self-defense in relation to Me, and not in relation to anyone else. The reason is that it is Me who is attacking You. This also implies a further point, namely that it is My attack that creates the situation in which You can potentially acquire a right to self-defense. The asymmetry in our normative relationship regarding the right to harm one another is created by my attack. Prior to this, You and I stood in a symmetric relationship in which neither of us had a right to harm. It follows that a right to self-defense is a right only in relation to a situational attacker, and it is a right that only arises in a situation where there is an attack. You have a right to self-defense only in relation to Me, because it is Me who is attacking You. Hence, an individual must be under attack in order to have a right to self-defense in relation to the attacker who has brought about this situation.

In order to grasp the second conception of the “individual” inherent to the notion of a right to self-defense, one more specification is necessary: In order for You to have a right to self-defense in relation to Me, there must be something particular about both My attack and Your defense. First, My attack must be imminent.¹⁵ You have no right to merely assume I could attack and therefore strike a first blow. Second, You only have a right to necessary self-defense. That is, there must not exist other, less harmful means of warding off My attack than the means You chose. Importantly, this implies a duty to retreat if possible.¹⁶ Furthermore, only that which is effective can also be necessary. Ineffective defensive means can never be necessary.¹⁷ Third, You only have a right to proportionate self-defense. The means You employ to ward off My attack are proportionate if there is no serious discrepancy in relation to the good achieved by warding off the attack. For example, You are not allowed to kill Me if all I want to do is pinch you, since there would be a serious disproportion between killing and pinching.

The crucial point here is that we can only ever know whether these conditions hold if we look at a particular attack and a particular defense. These conditions only affect the existence of a right to self-defense in a concrete case. In other words, those conditions refer to an immediately given, specifically perceptible, and individually experienced situation. Whether they hold—and hence whether You have a right to self-defense in relation to Me—depends on the concrete situation You and I find ourselves in. There exist innumerable possible concrete situations involving potential harm done in self-defense. Most importantly, the very fact that these conditions are situationally concrete makes each individual involved in each situation of attack and defense concrete as well. Moreover, a concrete individual is not simply a general entity who can be replaced by another individual understood as an entity. A concrete individual is a particular entity. In the event that My attack is imminent and Your defense is necessary and proportionate, it is not only You as an individual entity who bears a right to self-defense in relation to Me as an individual entity. Furthermore, it is not only You as an individual defender who bears a right to self-defense in relation to Me, as an individual attacker in a scenario where I attack You. It is You, precisely as You, who has a right to self-defense in relation to Me, precisely as Me, because it is precisely Me, in My concrete life situation, who launches an attack that, in this very concrete case, is imminent to precisely You, in Your concrete life situation. And it is You, precisely as You, who finds the necessary and proportionate means to ward off My attack in precisely this concrete case. You and I stand in an intimate, bipolar, asymmetrical, normative relationship that is grounded in the intimate, bipolar, and asymmetric structure of the concrete situation created by My imminent attack and Your necessary and proportionate defense. Your right to self-defense emerges only in virtue of this relationship, which is necessarily grounded in concreteness. This makes You and Me unique and not substitutable or replaceable by any other two individuals. With this, we reach our second conception of the “individual” inherent to the notion of a right to self-defense:

(2) The individual who bears a right to self-defense is a concrete and particular individual, or identity, in the sense of “precisely that person,” in contrast to the “general-abstract.”

3.2.1 Internalist and Externalist Accounts of Liability

This article is an internal methodological critique of reductive-individualist accounts of revisionist just war theories. In order to qualify the critique, it is, at this point, important to introduce a concept that is often used when reductive individualist just war theorists write about the concept of self-defense. The concept is “liability”. In the revisionist literature on the just war, “liability” is, generally, understood in two distinct ways. The article’s main critique is germane only to one way of understanding – the “internalist” account of liability, and misses the mark for the other – the “externalist” account of liability. Hence, this section presents a limitation of the article’ main critique.¹⁸ The remainder of this sub-section introduces the concept of “liability” and explains how it relates to the moral concept of self-defense outlined above, describes both its “internalist” and “externalist” version, and explains why the article’s main critique is germane to the “internalist” version.

“Liability” is a term borrowed from tort law. Revisionist just war theorists often use it to designate the very criterion that decides whether a person has lost her right against being defensively harmed. The idea is this: The harm to which You are “liable” does not wrong you. Applied to the picture of the concept of self-defense presented above, the concept of “liability” yields the following statement: You are “liable” to My necessary and proportionate defensive harm given your unjust imminent attack. My necessary and proportionate defensive harm applied to ward off your unjust imminent attack does not wrong you, because you have forfeited your right against being so harmed in precisely this situation, that is, because you are “liable” to be so harmed. However – and here we approach the core of the distinction between the “internalist” and the “externalist” account of liability – if I apply unnecessary defensive harm, i.e. harm that surpasses what would be required to ward off Your attack, I wrong you, because you are not liable to unnecessary defensive harm. This very idea describes the “internalist” account of liability, that is, a concept of liability of which the condition of necessity is constitutive. According to the internalist account of liability, one can only ever be liable to necessary harm, never to unnecessary one. This “internalist” account of liability is captured by the concept of self-defense that the present article’s argument presumes. Hence, the article’s critique applies to authors who endorse an internalist account of liability, such as McMahan (McMahan, 2008, p. 9).

An account of liability of which the necessity constraint is constitutive understands the harm done to a person who is liable as a means to avert the threat the unjust attacker is morally responsible for: the defensive harm must be necessary in order to ward it off (McMahan, 2009).¹⁹ It reflects the idea that harms are bad “(…) not only for those who suffer them but also from an impersonal point of view (…)” (McMahan, 2008, p. 8). Desert could be called the opposing view to an internalist account of liability, because deserving harm is itself a positive reason to inflict harm upon the person who deserves it. Getting what one deserves is, generally, regarded to be impartially good (McMahan, 8), which makes desert a concept that understands harms as ends in themselves Frowe, 2014, p. 91).

“Externalists”, in contrast to “internalists”, detach the permissibility of defensive harm from the concept of necessity. It does not matter whether the defensive harm is necessary to ward off the unjust attack. The condition of “being responsible for an unjust threat of attack” is sufficient to become liable to proportionate defensive harm, that is, it creates an “overall permissibility” of inflicting proportionate defensive harm (Frowe, 2014, p. 89). This means that, if You are responsible for an unjust threat of attack, You are not wronged if I apply unnecessary but proportionate defensive force.²⁰ Put differently, for externalists, a right to proportionate defensive harm against an imminent threat of attack arises against a person who is responsible for an unjustified threat of attack, period.

As opposed to desert, an externalist account of liability does not give the justified defender a positive reason to inflict harm upon the attacker. Harming her is not an end in itself that reflects the fact that she does not deserve it. However, for externalists, harm is also not to be understood as a means necessary to ward off a particular threat, as internalists would argue. Being liable to defensive harm in the externalist sense simply means that there exists no reason anymore not to proportionately harm the person who is liable. The important point is that this means that proportionately harming her does not need to serve a purpose. Hence, it does not even need to serve the purpose of averting a particular threat, which makes the externalist account of liability independent of a particular threat. This means that a person liable in the externalist sense is liable to be killed for any purpose. Harming her must simply be proportional to the threat for which she is morally responsible, although the harm done to her must not be necessary to avert this very threat (Frowe, 2014, p. 93, and 2016, 155).

Given that the concept of self-defense that the present article endorses holds that a right to defensive harm emerges only in virtue of an asymmetric normative relationship between a just defender and an unjust attacker, a relationship that is necessarily grounded in the concreteness and particularity of the threat of attack of the latter on the former, the article’s critique misses the mark for externalists, such as Frowe (2014, 2016). It would be wrong to claim that externalist reductive individualists of the just war follow an inconsistent methodology because they do not correctly apply the moral concept of self-defense to war, given that they lose the notion of the individual as a “particular” when applying said concept. Their concept of justified self-defense does simply not include the notion of the individual as a “particular”.

3.2.2 Other-Defense

The observation that the concept of self-defense also understands the individual as a “particular” is based on the claim that a right to self-defense arises only between two people whose asymmetrical normative relationship is rooted in a concrete situation. However, would it be permitted for a third concrete person to come to a potential defender’s aid? If so, on what grounds? What is the relationship between the right to self-defense and the right to other-defense, and does the fact that other-defense would extend the previous bipolar relationship between just defender and unjust attacker to, at least, a third person, somewhat undermine the claim that the right to self-defense reflects an understanding of the individual as a “particular”?

The right to self-defense is a full “liberty right.”²¹ This means that it consists in the simple freedom to either defend oneself or not. It is not a feature of liberty rights that they entail attendant claims that others have the liberty to do the same thing. If I have a liberty to draw funds from my bank account, this by no means entails that someone else has the liberty to draw those funds from my bank account in case I am for some reason prevented to do so (Rodin, 2002, p. 32). Also, a liberty to do something is usually understood to be discretionary, meaning that one may waive it if one wishes (Hart, 1982, p. 183). Other-defense, however, does not correspond to either of these observations. Other-defense is defense, or assistance to defense, on behalf of a victim who is, for some reason, prevented in part or in full to defend herself. Moreover, other-defense is often considered not as a mere right, but as a duty as well (Rodin, 2002, p. 32; Frowe, 2011, p. 24).²² Hence, the right to self-defense does not cover a right of a third to come to the victim’s aid.

The right to other-defense must rather be understood as derived from a duty of rescue. Duties of rescue arise from the mere facts that someone is in danger and that one is situated so as to be able to assist.²³ Duties of rescue usually consist in “half-liberties”, meaning liberties “(…) to perform an act which is compatible with (because entailed by) a duty to perform it” (Rodin, 2002, p. 38). The strength of a duty to rescue varies with one’s chance of succeeding in the rescue and the degree of risk the rescue imposes on the rescuer. When the risks are high and the chances of success low, the duty can be diminished and become almost null (Rodin, 2002, p. 39). If the risks are very high, other-defense may become supererogatory. If the chances of success are, in turn, very low despite only a small risk, the defense becomes almost ineffective and, therefore, difficult to fulfil the necessity constraint.²⁴

Two points are noteworthy for the present argument. First, although the third party potentially engaging in other-defense is not standing in the bipolar asymmetric normative relationship with the attacker same as the victim, her right or duty of other-defense is still grounded within the relationship of the latter two. It is a concrete attacker’s imminent unjust threat to a concrete victim as well as the third party’s possibility of assisting that impose a right or a duty on the latter to apply necessary and proportionate harm in defense of the victim. This makes the third party equally concrete. It follows that, same as the right to self-defense, the right to other-defense reflects an understanding of the individual as a “particular”.

And, second, given the proportionality condition, if a third person defends more than one victim against a concrete imminent threat, the amount of force that she may apply in said other-defense could be argued to be correspondingly higher: the good achieved by the other-defense is higher if, e.g., it saves many lives, which makes a higher amount of other-defensive force proportionate (Frowe, 2011, p. 24). Note, however, that said higher other-defensive force is still subject to the necessity constraint; what may be proportionate may not be necessary.²⁵

Same as self-defense, the moral right to other-defense has played a role in the reductive individualist methodology towards arguments for a just war. On the one hand, other-defense is often referred to in attempts to justify humanitarian interventions (Frowe, 2011, p. 24 and 84 ff.) On the other, combatants can be understood to defend their comrades, other citizens, their nations, as well as, in the case of unjust invasions, joint property rights (Fabre, 2012, p. 55 and 67). In reference to the proportionality condition, the latter approach can ground the argument that just combatants may apply more force than defenders have in usual cases of self-defense. However, as mentioned above, the justified amount of force must fulfil the necessity constraint, if an internalist conception of liability is endorsed. Moreover, if the risks to the defender are very high, she may not need to engage in other-defense. In addition, based on the present article’s claim that defense is justifiable only in case of a concrete individual attack, all victims that are to be potentially defended by a third would need to face said concrete attack, that is, they would each need to stand in the bipolar asymmetric relationship with the attacker.²⁶

3.3 Self-Defense as a Rule

Let us recapitulate: An answer to the question of whether an act of harm is morally justified by recourse to the concept of a right to self-defense can only ever be given for a concrete situation. More precisely, this means that a justification of harm by recourse to the concept of self-defense is always necessarily concrete as well.²⁷

Still, self-defense can also be understood as a general-abstract rule. How so? Consider that the moral concept of a right to self-defense is a set of normative relations between the concepts of “attacker” and “defender” that hold in the event of an “imminent attack” and “necessary and proportionate defense.”

One can imagine a great number of individuals and cases that could “fit with” or “correspond to” the concepts of “defender,” “attacker,” “imminent attack,” and “necessary and proportionate defense.” When understood as a rule, self-defense does not prima facie seem to reflect the individual concreteness that we claimed it does. One might say that anyone who faces an imminent attack and who is capable of warding it off using necessary and proportionate means is justified in doing so. However, the important point is that, as with any moral rule, the justification of violence by recourse to the rule of self-defense becomes effective only through the application of the rule to a concrete case. Hence, even if self-defense is understood as a general-abstract rule, it can only exercise normative force (i.e., it can only justify actions or come into effect) on the level of the particular individual in the evaluation of a concrete case.

This is also why justification with recourse to a rule of self-defense cannot be based on heuristic methods or (statistical) assumptions about a number of cases of harm, from which the presumptive right to self-defense of a concrete individual is deduced. The reason for this is that such a method would ground the bipolar, asymmetric, normative structure of self-defense in a presumptive or hypothetical situational asymmetry. A presumptive or hypothetical situational asymmetry, however, merely presumes the application of the rule, and hence the justification that is grounded in it. A presumptive justification, however, must necessarily remain speculative and can never become evaluative. But it is only in the evaluative application to a concrete individual case that a justification involving recourse to the rule of self-defense can take effect. Beyond the level of the particular individual, it loses its justificatory essence.²⁸ For example, to conclude from the hypothetical assumption that, in 90% of stabbings, the attacker was armed with a plastic knife that this is also the case in a concrete situation of stabbing would not suffice as an application of the rule of self-defense to a case of stabbing.

Note that this does not mean that a rule of self-defense cannot apply to situations involving uncertainty. A rule of self-defense accommodates uncertainty well and in a wide range of cases: Many concrete situations involving attacks or threats of attacks include uncertainty that is dealt with once the rule is applied: e.g., “Did the attacker really want to attack?” or “Could the defender have known that there was no bullet in the attacker’s gun before he applied lethal self-defense?” and so on. The crucial point is that a rule of self-defense accommodates and deals with uncertainty on the level of the particular (i.e., in its application), but not on the level of the general-abstract.

This means that, despite being understood as a rule, self-defense does not and must not be abstracted from the diverse range of appearances, phenomena, individuals, and possibilities of human action, but must grasp processes as concrete and dynamic and each individual as a particular.

4 The Reductive-Individualist Understanding of the “Individual”

We have reached an understanding of the idea of the “individual” that is inherent in the moral concept of a right to self-defense. At this stage, we have also established that reductive individualists reduce the morality of harm and killing in war to the morality of harm and killing in self-defense. Two notable representatives of reductive individualism articulate this reductive project in the following terms: Frowe states that principles of self-defense “directly apply in war,” while McMahan writes that “[…t]he difference between war and other forms of conflict is a difference only of degree and thus the moral principles that govern killing in lesser forms of conflict (i.e. self-defense) govern killing in war as well” (Frowe, 2014, p. 14; McMahan, 2009, p. 156).

If the moral concept of individual self-defense applies directly to killing in war, the two conceptions of the “individual” inherent in the right to self-defense explained above should govern reductive-individualist just war theories. We shall now examine whether they do so.

Although the reductive-individualist literature is extensive, diverse, and evolving,²⁹ its distinctive methodology runs through the bulk of the substantive arguments and conclusions in these works. The present article attempts to capture the understanding of the “individual” that some of the most seminal reductive-individualist texts share. It will contrast this understanding with the conceptions of the “individual” inherent in the moral concept of an individual right to self-defense presented above.

4.1 The Reductive-Individualist Individual is a Right-Bearing Entity

For reductive individualists, the individual, as opposed to the state, is the “proper” and “fundamental focus of moral concern” (Fabre, 2012, p. 2; Frowe, 2014, p. 13; McMahan, 1994, 2009).³⁰ As Fabre puts it, the theoretical principles of just war must “ascribe pre-eminence to individuals and not conceive of groups as having independent moral status” (Fabre, 2012, p. 7). Reductive-individualist theories also start with the concept of “rights of individual persons” (Rodin and Shue, 2008, 5). This is why those theories are sometimes referred to as reflecting a “rights-based approach” to the morality of killing in war (Lazar, 2010).³¹

Just like the attacker in the case of individual self-defense in ordinary life (in a peacetime context) described above, reductive individualists claim that an individual combatant can lose his right against being harmed and killed by an enemy combatant in war if the former is responsible for an unjust threat of attack to the latter. The latter thereby receives a right to defensive harm (Lazar, 2018). Hence, reductive individualists subscribe to the first conception of the “individual” inherent in the notion of a right to self-defense: It is the human individual and not a collective or a state who bears a right to defensive harm.

4.2 The Reductive-Individualist Individual is not a Particular Individual

Does the reductive-individualist approach also reflect the second conception of the “individual” as concrete or particular in contrast to the “general-abstract”? Following the moral understanding of a right to self-defense in ordinary life, McMahan states that, for a person to be permissibly harmed or killed in war, ‘[…t]he right against attack is […] forfeited only in relation to certain persons acting for certain reasons in a particular context’ (McMahan, 2009, p. 10). This seems largely analogous to the conditions in which a right to self-defense in ordinary life potentially appears: According to the moral understanding of a right to self-defense, the “particular person” in relation to whom the attacker forfeits his right not to be harmed or killed is the person he attacks, that is, the defender in this particular situation. The “particular reason” on account of which he forfeits his right is his unjust attack. The “particular context” in which he³² forfeits his right is the situation which he himself creates in virtue of his attack. A consistent application of the structure of the self-defense principle to war would imply that a combatant loses his right not to be killed by another combatant if he attacks him without justification. The combatant who is attacked would, in turn, have the right to defend himself against his attacker.

4.2.1 Detachment from Situational Concreteness

Reductive individualists depart from this account, however, in that they give the greatest weight to the reason for the attacking combatant’s attack, while downplaying the specificity of the person and the context or situation.

Reductive individualists hold that a combatant only loses his right against lethal harm for a particular reason—that is, if he is individually responsible³³ for contributing to an unjust threat of attack. It is important to note that “contributing to an unjust threat of attack” does not necessarily mean personally attacking or personally posing a threat of attack in a certain situation. “Contributing to an unjust threat” is a broad qualification that, in simple terms, could, depending on the author, potentially cover almost any military act that contributes to the overall unjust war aim of the combatant’s state.³⁴ Thus, the term “threat of attack” is intimately tied to and dependent on the overall attack by the state. Hence, simply fighting for a state whose cause for war is unjust results in a combatant forfeiting his right against being harmed or killed in war, because in fighting, he contributes to this overall attack or threat of attack by the state.³⁵ Reductive individualists call this combatant an “unjust combatant.”³⁶ He has the same moral status as an attacker in a scenario involving individual self-defense in ordinary life, because, like an ordinary attacker, the unjust combatant forfeits his right against being harmed or killed. The unjust combatant’s enemy, by contrast, is not responsible for contributing to an unjust war effort and, hence, does not lose his right against harm. He is a “just combatant” and the moral equivalent of a defender in an ordinary case of self-defense.³⁷

Whereas reductive individualists place great emphasis on the reason for a person’s attack or threat of attack, the concrete individual situation in which the attack takes place is not taken into account. Consider, for instance, a concrete battlefield situation in which an unjust and a just combatant are present and in which the just combatant attacks the unjust combatant. The just combatant is a situational “attacker” and the unjust combatant is a situational “defender.” However, in the reductive-individualist account of war, the just combatant’s imminent attack in this situation does not affect whether the unjust combatant has any right to engage in necessary and proportionate self-defense in reaction to the just combatant’s attack. Since the unjust combatant is “unjust” due to his responsibility for fighting for a war aim that is unjust overall, he has no right to engage in defensive killing in the first place. The fact that the just combatant is attacking the unjust combatant in this scenario does not change anything about that. Hence, the principle that a just combatant has a right to kill an unjust one but not vice versa is detached from any situational concreteness. It is not only irrelevant whether a situational attack is imminent, and the defense necessary and proportionate, but it is also irrelevant who is attacking and even whether there is a situational attack at all. Whether a combatant has a right to harm and kill in war is predetermined by the justness of the cause of war of that combatant’s state and his responsibility for contributing to the war, independently of concrete situational parameters. Unjust combatants who are responsible for contributing to an unjust war aim are moral equals to concrete situational attackers in ordinary self-defense, while just combatants are moral equals to concrete situational defenders in ordinary self-defense (Fabre, 2012, p. 6, 79; Frowe, 2014, p. 15). However, whereas the moral status of the attacker and the defender in ordinary self-defense depends on the concreteness of an individual situation of attack, the moral status of unjust and just combatants is detached from concrete individual attacks.

Reductive individualists may argue that the moral status of a combatant is still linked to a “situation of attack and defense,” namely the overall “situation” of attack and defense existing between two belligerent states. However, this approach broadens the concept of the concrete individual situation of attack and defense to include a collective or institutional situation of attack and defense.

Reductive individualists mainly refer to self-defense. Note, however, that the same observation holds if just combatants are assumed to also engage in other-defense. If they are regarded as also defending their comrades-in-arms, or fellow citizens (and, potentially, the latters’ property rights), the threatened attack they are supposed to ward off in other-defense is still the overall state threat, and not individually concrete threats of attack on each comrade or citizen they are to defend.

To sum up, for reductive individualists, just and unjust combatants do not share their moral status with individual defenders and attackers in ordinary self-defense for the same reasons that the latter have such a status to begin with. In any battlefield situation where an unjust and a just combatant are present, for reductive individualists, neither the fact that one of them is attacking nor their individual particularity matters when it comes to determining whether their acts of harm are morally justified. Hence, both the just and the unjust combatant could, in principle, be replaced by any of their comrades-in-arms. The general reductive-individualist conclusion reflects this replaceability: It holds that all combatants fighting for a state that is waging an unjust war are responsible for the occurrence of that war (McMahan, 2004a, 2009; Rodin, 2008; Fabre, 2012). Reductive individualism offers a moral dichotomy between a just and an unjust collective of individual combatants.³⁸ This means that any individual, understood as an entity who could bear a right to harm and who is part of either collective, can be freely substituted with any other member of that collective.

Reductive individualists understand the individual as a bearer of a right to harm and kill (corresponding to the first conception of individuality). However, this right does not appear to be connected to any concrete situation of attack: Every just combatant has a right to harm or kill his enemies. The sole concretization we are offered are the labels “just” and “unjust,” which potentially cover enormous numbers of individuals,³⁹ each of whom can theoretically be replaced by any other member from the same moral category of combatants. Besides the warring party they are fighting for, none of them matters as a particular or specific individual. Within reductive individualist reasoning, the second conception of the “individual” inherent in the moral notion of a right to self-defense—that is, the individual as something concrete and specific as opposed to the “general-abstract”—evaporates.

4.2.2 Moral Collectivization

Can any other aspect of the reductive-individualist method capture the idea of an “individual” as a particular (i.e., the second conception), and not just as a replaceable entity? There does seem to be one possibility: Whereas reductive individualists detach an individual combatant’s moral status from situational concreteness, they make it dependent on that individual’s “individual responsibility” for contributing to a state threat that is unjust overall. Does the notion of “individual responsibility” reflect an understanding of the individual as a “particular” rather than a mere abstract entity? In other words, does the term simply refer to the “responsibility of individuals,” or does it cover potentially innumerable manifestations of different concrete responsibilities?

Answering this question requires highlighting that reductive individualist thinkers have acknowledged the challenge of applying the rule of self-defense to combat situations in different ways. And, those different ways all address the notion of “individual responsibility” for contributing to an unjust threat. For instance, McMahan concedes that unjust combatants may, depending on their individual contribution to an unjust war effort, be more or less morally responsible. Furthermore, some may be partially or even fully excused, e.g., due to ignorance about the injustice of their war, or if they have been forced to enlist by a draft policy or by socio-economic circumstances (McMahan, 2009, p. 116; McMahan, 2006b, p. 49). The difficulty of applying the concept or rule of self-defense to war, then, arises in acknowledging those individual differences in responsibility. McMahan concludes from the observation of different levels of excuses, however, that unjust combatants who are partially excused still keep a residual amount of moral responsibility for contributing to the unjust state threat. And, this residual amount of individual responsibility is sufficient for those unjust combatants to forfeit their right not to be attacked by their just adversaries. Fabre argues similarly. She claims that unjust combatants’ individual contributions to the overall unjust state threat may be only marginal and, thereby, not significant enough on their own for those unjust combatants to forfeit their right against lethal harm. Based on this observation, Fabre then argues that only “sub-lethal” defense is justified in those cases. However, her conclusion is, given that, jointly, unjust combatants participate in an unjust war overall, that they are each individually liable to be killed for their participation (Fabre, 2009, p. 76). ⁴⁰⁴¹

Those two examples reflect the reductive individualist attempt of artificially re-constructing the conditions for “individual responsibility” for an unjust belligerent threat such that, in McMahan’s words, “there is a basis for liability in the case of virtually all unjust combatants” (McMahan, 2009, 159 ff., italics added). For McMahan, individual responsibility is immune from partial individual excuses. For Fabre, individual responsibility is detached from one’s individual contribution and re-attached to one’s participation in a joint unjust belligerent endeavor. For McMahan, however, the difficulty still arises with fully excused combatants: how can he account for the complete absence of their individual responsibility for contributing to an unjust war aim? He does not, but argues that all unjust combatants, also fully excused ones, are individually responsible for their just opponents’ reasonable belief that they are not excused. This responsibility for their enemy’s reasonable belief about their own responsibility level causes even fully excused unjust combatants to forfeit their right against being harmed or killed by that just enemy (McMahan, 2009, 2011).⁴² Hence, reductive individualists acknowledge the fact that not each unjust combatant may be responsible for contributing to an unjust threat to the same degree. However, this observation either invites them to change the condition for forfeiting unjust combatants’ right against harm, such as, e.g., Fabre, who concludes that participating in an unjust joint endeavor with others is enough. Or else, they detach the condition for forfeiting one’s right against lethal attack altogether from an actual contribution, however small, to an unjust state threat, like McMahan, who makes the forfeiture of an unjust combatant’s right against lethal harm dependent on his responsibility for the false belief of his own just opponent.

One may summarize this reductive-individualist debate concerning individual responsibility a debate about where to “set the threshold” of responsibility that triggers the loss of a right against being harmed or killed in war (Lazar, 2010; May, 2015; Arneson, 2017). In other words, the debate’s guiding question is: To what extent must a combatant be responsible for his contribution to an overall unjust state threat in order for him to become a legitimate target in war?⁴³ If the threshold is assumed to be low, e.g. as Fabre suggests with the simple requirement of participation in an unjust war aim, not only all unjust combatants, but also certain unjust civilians, would potentially lose this right and become legitimate targets in war (Lazar, 2010). The reason why certain civilians would be included in the “low-responsibility group” is that it is also possible, e.g., for a civilian arms manufacturer or politicians who took the decision to resort to war to attain this “low” level of responsibility (Fabre, 2009, 2012; Frowe, 2014). However, if the threshold is set “high,” not all combatant fighting for the unjust state lose their right against lethal harm (May, 2015). It would be permissible for just combatants to kill only some, but not all, of their unjust enemies. But how would they know which ones? No unjust combatant carries a tag indicating whether he has lost his right against lethal harm or not.⁴⁴ Accepting the insolubility of this problem would make just wars possible in theory, but impossible in reality (May, 2015). Perplexingly, some analysts seem to find this “unpalatable.”⁴⁵ Here, McMahan’s above-mentioned claim becomes effective: His point that unjust combatant who nonetheless do not attain the high threshold of responsibility necessary for them to forfeit their right against being harmed or killed are still presumed to be responsible for the fact that all just combatants can reasonably believe that they have lost this right seems to solve the problem. (McMahan, 2009, 2011).

The idea of establishing a threshold of responsibility is inspired by the following question: What can we reasonably believe a group of people to all be responsible for? Can we reasonably assume that all (or even some) unjust combatants are individually responsible for contributing to an unjust overall state threat? In other words, the threshold debate concerns a presumption of responsibility in relation to a collective of combatants and, if the threshold is set low, to a group of civilians as well. It presumes that a group of persons—which can be larger or smaller, depending on where one fixes the threshold—exhibits a characteristic necessary for its individual members to lose their right against lethal harm, namely “sufficient individual responsibility for contributing to an unjust overall state threat.”⁴⁶ Put differently, whether a combatant loses his right against being killed in war depends on a moral feature that he presumably possesses. What is more, it is not only him who is presumed to possess that feature to the extent necessary to become a legitimate target in war, but many (if the threshold is set high) or almost all (if the threshold is set low) of his comrades-in-arms are presumed to also have it to the necessary extent. Individual responsibility is presumed to be a characteristic of a great many individuals. On this basis, theorists mentally integrate individuals into a collective that is presumed to be morally homogeneous.

Presuming that a collective shares this essential characteristic, however, requires us to abstract from its different expressions: Not every single unjust combatant is responsible to the same degree, and not all unjust combatant contribute equally to the occurrence of an unjust war. The presumption of a threshold also requires abstracting from a wide range of other phenomena and processes, e.g. individual backgrounds, mental states, or intentions. What is more, given that this feature is merely presumed, the mental assimilation also abstracts from the fact that some combatants may not even possess it. In other words, some combatants may lack sufficient responsibility for their contribution to an unjust war aim, even if the threshold is set very low. This is a thought about certain concrete individuals, one which some reductive individualists have taken up (McMahan, 2009, 2011). They have not, however, been concerned about the fact that generalizing about individual human beings based on a presumed moral feature seems to go against the idea inherent in the concept of self-defense that individuals also matter as concrete and specific individuals. Reductive individualists have been concerned with the fact that if they considered different expressions of individual responsibility, wars could not be fought justly in practice: No just combatant would know whether he is facing an unjust enemy who has lost his right against being killed or not. In order to solve this practicability problem, yet another presumption about a collectively held feature is advanced: All unjust combatants are responsible for the false beliefs of just enemy combatants that they themselves all bear sufficient responsibility for contributing to an unjust war aim. ⁴⁷ In short, the theoretical practicability of just war trumps theoretical consistency.

The criterion of “individual responsibility” incorporates an understanding of the “individual” as an entity that has the characteristic of “being responsible to at least a certain degree for contributing to an unjust war aim.” This characteristic is a situationally detached presumption about the moral status of a great number (with a high responsibility threshold) or almost all unjust combatants (with a low responsibility threshold). Reductive individualists abstract from the different degrees of responsibility of concrete individuals. In doing so, any individual in the unjust collective can, in principle, be replaced by any other. An unjust individual is not understood as a particular in opposition to other unjust individuals. Likewise, a just individual is not regarded as a particular in opposition to other just individuals. The second understanding of the individual as a “particular” as opposed to the “general-abstract,” which is inherent in the normative concept of self-defense, is also not captured by the reductive individualist concept of “individual responsibility.”

McMahan himself would push back against the present claim that reductive individualists collectivize individual moral statuses of unjust combatants on the basis of presumptions. He writes: “I of course concede that reliable information about the degree of an individual’s moral responsibility is virtually never available in conditions of war. But this does not mean that a combatant’s liability is collectivized; it means that combatants must act on the basis of presumptions of liability” (McMahan, 2006b, p. 48). The problem with McMahan’s observation of the reductive individualist methodology is that the recommendation that all just combatants should act based on the presumption that all unjust combatants have forfeited their right against lethal attack means treating unjust combatants as collectives in reality.

Both McMahan and Kutz would reply by conceding that the moral rules of self-defense are not directly transposable to real-world battlefields. In their view, though, the upshot of this concession is that the morality of war and the law of war must diverge. The law should simply guide combatants such that their acts conform as good as possible to what the revisionist morality requires (McMahan 2013, 180–184; Kutz, 2008).⁴⁸

However, this claim is problematic for, at least, the following reason: Accepting a divergence between the morality and the law of war means denying systems of moral thought to develop and articulate unchangeable moral principles that are practicable in the real world, in other words, statements of truth or, at least, statements of justified knowledge. However, moral theories are meant to guide action in this world (Elster, 2011, p. 44; Prieto 2022).⁴⁹ This is why, as some argue, McMahan’s and Kutz’ position of a divergence between the morality and the law of war relegates morality to an abstract sphere, and accepts a largely inoperable moral philosophy as true.⁵⁰⁵¹

4.3 Self-Defense as a Speculative, Reductive-Individualist “Rule of Thumb”

Reductive individualists adopt a speculative presumption about a certain feature possessed by a considerable number of combatants and abstract from individual expressions of it. This feature—namely responsibility for contributing to a state threat that is unjust overall—is extremely relevant on the moral level, because its existence invalidates combatants’ rights against being killed in war. It makes these combatants, morally speaking, individual “attackers.” Simultaneously, its absence gives other combatants a right to kill those who possess this feature: i.e., its absence makes them individual “defenders.”

However, as we have seen above (3.3.), a rule of self-defense cannot justify acts of harm based on presumptions about a number of cases or about a number of individuals implicated in those cases. The reason for this is that the justification remains speculative here, rather than evaluative, and a speculative justification can never apply to a concrete individual case. However, the level of the particular and concrete is the only level where justification by recourse to the rule of self-defense is possible.

Reductive individualists derive the moral status of individuals, as well as rights to defensive harm, from generalized presumptions (4.2.2.) rather than from concrete situational analyses (4.2.1.). Hence, whether a concrete battlefield situation can be described in asymmetric, bipolar terms that correspond to the normative structure of self-defense is either presupposed (without being argued for) or regarded as irrelevant. Thus, the claim that any just combatant has a right to kill any unjust combatant (but not vice versa) is speculation, not evaluation. As a result, reductive individualists offer no justification for any individual just combatant killing any individual unjust combatant. As there is no justification grounded in the rule of or right to self-defense that is not concrete, reductive individualists offer no justification for harm in war.

5 Conclusion and Critical Assessment

As Frowe puts it, reductive individualists purport to utilize the moral concept of a right to individual self-defense “directly” to war in order to establish whether killing in war is justified. The present article assessed whether reductive individualism makes good on this claim, with reference to the two distinct conceptions of the “individual” inherent in the concept of a right to self-defense. Whereas reductive individualists, drawing an analogy with the concept of a right to self-defense, understand the individual human being as the sole entity capable of bearing a right to harm—as opposed to a collective (e.g. a state)—they fail to take into account the second understanding of the “individual”: i.e., a monadic understanding of the individual as a concrete and specific person, as an individual who matters for being precisely that person and not someone else. Hence, the reductive-individualist “reduction” of the morality of killing in war to the moral right to self-defense is inconsistent.

Furthermore, it is only on the level of the individual as a concrete and particular person that a rule of self-defense can ever justify harm. This is why there can be no justification by recourse to the rule or right to self-defense other than a concrete one. For this reason, a self-defense justification is necessarily evaluative and never speculative. Reductive individualists, however, derive a right to defensive killing in war from speculative presumptions bearing on collectives. Hence, whether this right to kill—and the harm it produces—is justified can also only ever be speculative. Remaining on the level of the speculative, however, means that the rule of self-defense has no justificatory effect. It follows that reductive individualists offer no justification for killing in war.

One might object that it is unsurprising that reductive individualists rely on a rather heuristic assumption about who is responsible for fighting for a state’s unjust cause and that they to some extent transform self-defense into a “rule of thumb.” The reason for this is that war—as a collective, integrated, somewhat unmanageable endeavor—is a situation involving massive uncertainty. Some speak of a “fog of war,” meaning that we may never know what really happens on the ground, who attacks whom, when, and why. Hence, the nature of war may simply be such that a rule of self-defense is not applicable to concrete cases of belligerent violence.⁵² If so, relying on a reductive-individualist presumption is the best we can do and is thus legitimate. There are two answers to this objection:

First, as described above, a rule of self-defense can properly accommodate uncertainty on the level of the particular. More importantly, however, war is not merely a situation of epistemic uncertainty. It is not “just” a situation in which it is impossible to know who is attacking whom, when, and where. War is a situation characterized by a constant collective background threat grounded in the fact that it is states that send their armies to fight. This fact is known. And it is on the level of the state that reductive individualists generalize moral statuses. Hence, their presumptive homogenization clearly goes beyond the uncertainty surrounding war. Their “rule of thumb” is based not on a presumption of uncertainty, but on a presumption of collectivity. This is why calling this rule of thumb “self-defense” does not seem legitimate.⁵³

Second, one could also argue that raising the applicability problem as a counter-argument is misplaced here, in a quite general sense, because the applicability problem depends on what war is understood to be as a matter of fact. However, the reductive-individualist approach remains inconsistent irrespective of any analysis of what war is and irrespective of any analysis of the reason why the concept of self-defense may not be applicable to it. Justification by a rule of self-defense requires the rule to be applied. Claiming that war is, however unfortunately, such that the rule cannot be applied, is no excuse for relying on an inconsistent reduction—that is, an incomplete understanding of the rule—to still “in some sense” justify killing in war.