A war of national defense is a war waged by a victim state (V) to defend itself against an attack whose ultimate aim is not to kill any of the members of V, but rather to acquire control over certain “political goods” belonging to V. These political goods include one or more of the following: (i) the control of V’s political institutions; (ii) the imposition of certain political and cultural values on V’s population; (iii) the acquisition of V’s territory; iv) the seizure of V’s natural resources. To be sure, the aggressing state (A) is ready to kill the members of V, but only if they attempt to resist A’s aggression. If they surrender, no blood will be spilled.1 (Hence the label “bloodless aggression,” which is sometimes used to refer to this case.)
Both international law and ordinary morality treat wars of national defense as the paradigm of justified wars, but on the face of it this view seems incompatible with another widely shared assumption, namely that defensive lethal force is subject to a proportionality constraint. I may permissibly use lethal defensive force in order to protect vital interests, whose fulfillment is necessary to have a minimally decent life, but I may not do so in order to protect lesser interests, whose fulfillment is not necessary to this end. For example, I may kill someone to defend my life or to avoid enslavement, but not to prevent them from stealing my tie clip. Although I do have a right over my tie clip, the interest over my clip is not sufficiently important to justify protecting it by resorting to lethal violence.
Wars of national defense seem unable to meet the proportionality requirement because they aim at protecting interests that, on the face of it, do not seem vital, such as the interest in retaining political independence or control over a certain territory. An example that is regularly discussed to support this conclusion is one in which you block my way while I am going to vote in a general election, and the only way in which I can reach the polling station is by killing you (Hurka 2011: 251–4; Frowe 2014: 133–8; Lazar 2014: 20; Fabre 2014: 105). It seems clear that in this case I am not permitted to kill you in order to exercise my right to political participation. Indeed, arguably I would not be permitted to kill you even if your obstruction was not a one-off, but something that happens every time I try to vote. This is supposed to show that political interests such as the interest in being enfranchised or in being part of a self-determining community are not sufficiently weighty to justify resorting to lethal violence. And if these interests are not sufficiently weighty to justify killing the wrongdoer in the voting case, a fortiori they cannot be sufficiently weighty to justify V in waging a defensive war against A, since wars always involve inflicting a significant amount of collateral harm on civilians and other innocent parties in addition to the harm inflicted on those responsible for the unjust attack.
So, if it’s impermissible to use lethal defensive force to protect non-vital interests, and if wars of national defense involve the use of lethal defensive force to protect non-vital interests, we should conclude that wars of national defense are impermissible.
Two strategies can be employed to resist this conclusion. First, we might argue that the moral principles that regulate the use of lethal force in war are different from those that regulate the use of lethal force in the domain of interpersonal violence. If that’s true, we don’t have to worry about the fact that ordinary interpersonal morality prohibits the use of lethal force to defend non-vital interests, as this moral prohibition need not also apply in war. Alternatively, we might agree that the moral principles that regulate the use of lethal force in war are the same principles that regulate interpersonal violence, but argue that the interest a political community has in defending itself is not reducible to the interests of its members. Even if none of V’s members has a vital interest in the political goods threatened by A, and thus none of them is permitted to resort to lethal violence to protect those goods, the political community in its entirety may be permitted to do so because the value of protecting V is not reducible to the value of protecting the interests that V’s members have.2
Neither of these strategies, however, is available to many contemporary just war theorists, since the two main tenets of the predominant approach to just war are: (i) that war is regulated by the same moral principles that apply to interpersonal morality (“reductivism”); and (ii) that the value of political communities is reducible to the contribution they make to the interests of their members (“individualism”). According to this approach, sometimes referred to as “revisionism,” we should reject the idea that states or political communities enjoy a privileged moral status when it comes to war. The defensive privileges that can be attributed to them are simply the sum of the defensive privileges that can be attributed to their members in order to defend the interests they have qua individuals. And thus, justified wars are ultimately reducible to the sum of a number of acts of killing, each of which is permissible according to the principles of interpersonal morality (Frowe 2014: 123; Hurka 2011: 252; Rodin 2002: 127).
Both reductivism and individualism have recently received powerful defenses (McMahan 2009; Fabre 2012; Rodin 2002; Frowe 2014). Until a convincing case is made as to why they should be rejected, giving them up would be too costly. What to do then? Some revisionists have responded by simply accepting the conclusion that wars of national defense are not permissible. David Rodin and Richard Norman, who first raised the bloodless aggression case, take this route (Rodin 2002, chap. 6; Norman 1995, chap. 5). But intuitively, the idea that V is not permitted to wage a defensive war against A is hard to accept. Other revisionists, including Jeff McMahan, Cécile Fabre and Helen Frowe, have offered a number of arguments that aim to vindicate the permissibility of fighting these wars without departing from reductivism or individualism. In the first part of this paper, I criticize some of these arguments, focusing in particular on those offered in Helen Frowe’s excellent new book Defensive Killing. In the second part, I outline what seems to me a more promising account.
2 Revisionist and Institutionalist Arguments
A strategy adopted by some revisionists, including Thomas Hurka and Helen Frowe, to vindicate the permissibility of wars of national defense consists in pointing out that the distinction between vital and lesser interests is not exhaustive. Whereas some lesser interests, such as my interest in retaining my tie clip, may never be defended by lethal means, others may be so defended when they are violated on a sufficiently large scale, despite the fact that this is not permissible when they are violated in isolation (Hurka 2011: 253–4; Frowe 2014: 140). Political interests, such as territory or the capacity to exercise political self-determination, are said to belong to this second class, and this explains why, according to Frowe, V is permitted to wage a defensive war against A in the bloodless aggression case, despite the fact that I’m not permitted to kill you in order to reach the ballots.3 In the second case, only my interest in exercising political self-determination is being set back, whereas in the first case millions of interests are. Although wars that aim at achieving merely political goals threaten interests that are non-vital when considered in isolation, they do so on such a wide scale that when aggregated, the interests in question do count as vital.
Two main objections can be raised against this strategy (Lazar 2014: 32–3; Fabre 2014). One is that it misleadingly aggregates the interests to political self-determination of V’s members without also aggregating the interests not to be killed which would be set back if V were to wage a defensive war. Even if we agree that a sufficiently large number of aggregated lesser interests of V’s citizens may in principle justify resorting to lethal defensive violence, more would need to be said to show that wars of national defense are justified. For while a significant difference between the bloodless aggression scenario and the voting case is that the number of political interests threatened in the former scenario is much larger than the one threatened in the latter, the defensive harm inflicted in the former scenario (the casualties of a war) is also much larger than the one inflicted in the latter (a single death).
A second objection is this: According to the aggregative strategy, in order to establish whether a given war of national defense is permissible, we need to aggregate the non-vital interests of V and weigh them against the interests of A’s members who would have to be killed to protect these interests. This means that whether a political community has a right to wage a war of self-defense will ultimately depend on its population size. The larger the population of V, the weightier the sum of their non-vital interest; the weightier the sum of these interests, the easier it is to justify their right to fight. But this seems implausible. Both ordinary morality and international law attribute the right to wage wars of national defense to political communities, regardless of their population size. And this seems intuitively right. It seems unfair to grant the right to wage wars of national defense to large political communities, but not to small ones.
A more promising formulation of the aggregative argument is offered by Frowe when she notices that certain harms, when aggregated, “produce a step-change in what threatens each individual” (2014: 144). To use her example, preventing a non-Londoner from voting won’t make a difference to the sort of country the uk is, but if all non-Londoners were prevented from voting, this “would change the uk from a legitimate democratic state to an illegitimate dictatorship. States coercively impose legal frameworks upon their citizens. A government that is elected only by Londoners, but then attempts to rule over the whole of the uk, will be imposing this framework illegally, unjustly coercing individuals to adhere to it. This is a much more serious wronging of each individual citizen than the wrong of preventing any one individual from voting, and it is this sort of wronging that is threatened by a political invasion” (p. 144).
This seems to me entirely correct. Frowe is certainly right that the wrong suffered by each British citizen in her example is different from the wrong each of them would suffer if they were prevented from voting. This is why discussing wars of national defense by appealing to the voting case is a non-starter. But merely pointing this out is not enough to justify the permissibility of waging wars of national defense. We need to explain why this sort of wrong may permissibly trigger lethal defensive harm. Failing that, we haven’t explained yet why it is permissible for V to resist A’s attack.
A separate argument is invoked by Frowe at this juncture, namely that it’s unrealistic to expect that the perpetrator of a successful political aggression be benign toward its victims. In discussing her example, she continues: “[t]he new government is not benign—rather, it threatens to quash legitimate and reasonable dissent with violence. And things are therefore very different in the uk, because now the securities that guarantee the citizens’ current way of life are gone. Instead, they live in a country where laws are illegally enforced upon them, backed by the threat of unjust violence if they don’t comply. This is true even if the laws are by and large the same laws that were in place prior to the invasion” (p. 144). Frowe then concludes that “not living in a state of perpetual oppression and fear” is a substantial good, and that “individuals living under such conditions could use lethal force to avoid it, and risk harm to innocent people whilst defending themselves” (p. 145).
This is all true, of course, but also beside the point. With the exception of pacifists, everybody will agree that (within the constraints of necessity and proportionality) it is permissible to use defensive lethal force to overthrow an oppressive regime that forces its subjects to live “in a state of perpetual oppression and fear” and that is “prepared to put down dissenters” and “quash legitimate and reasonable dissent with violence.” The question we are interested in is whether it is permissible to resort to lethal defensive force to prevent political aggressions, i.e. aggressions in which the only interest set back is the interest in maintaining control over political goods such as political independence or territory, even assuming that no further interests are set back.4
Here one might reply that there is no point in considering such an abstract question, since in real life there simply can be no purely political aggressions. Political aggression inevitably leads to situations in which the more familiar sort of harm that justifies lethal defensive force will also be present (Fabre 2014: 108–9). But this reply is unpersuasive. To begin with, even if we are skeptical that bloodless aggression cases could ever occur, they are useful heuristics to establish whether and to what extent violations of purely political interests can justify the use of defensive lethal force (Fabre 2014: 105). As such, they tell us something important about the scope and the resources of our favored account of just war, and thus it matters whether our theory manages to accommodate them or not.
Second, it’s not clear what the argument in support of the empirical claim that there haven’t been (let alone that there cannot be) any “bloodless aggressions” is. Pointing out that there has been hardly any war in which blood was not spilled won’t do because bloodless aggressions are not wars in which no blood is spilled, but rather wars in which no blood is spilled unless the victims of the attack fail to surrender. The fact that all or almost all the wars we know did involve spilling copious amounts of blood, is compatible with the possibility that many, and even most of them, were cases of bloodless aggression, if the violence was a consequence of the victim’s failure to surrender. And indeed, as Seth Lazar notices, many of the wars we have witnessed in recent times are likely to be instances of bloodless aggression: the aims of these wars were to pursue economic interests, to replace unfriendly regimes with friendlier ones or to otherwise secure some influence over the policies of particular areas of the globe. It’s not clear that violence would have followed if these goals had been achieved without fighting (Lazar 2014: 22–5).
Finally, as we have seen, both international law and ordinary morality acknowledge V’s permission to defend itself against A, independently of whether the regime instituted by A would be oppressive or not. Thus, all things being equal, a view that can justify this permission should be preferred to those that cannot.
In the next section I attempt to articulate such a view. Before I do that however, I briefly consider an attempt to justify V’s right to defend itself without appealing exclusively to the interests of V’s members in retaining control over the relevant political goods. This attempt is interesting because it raises doubts about the very methodology adopted by the authors I have discussed and that I will also be employing.
Allen Buchanan and Margaret Moore have each recently argued that one problem with the way in which the question of the permissibility of wars of national defense is normally discussed is that it focuses exclusively on the morality of war, without paying attention to the question of how such morality should be institutionalized (Moore 2015: 221, 224; Moore 2014: 186–9; Buchanan 2016). The reason why V is permitted to wage a defensive war against A, according to them, is that institutionalizing the prohibition to wage wars of national defense would lead to bad results, primarily because it would likely incentivize further perpetrations of purely political aggressions. Thus, even if the interests of V’s citizens by themselves are not sufficiently weighty to warrant a defensive war, V should be allowed to defend itself, as this will deter other states from engaging in similar aggressions. This institutional argument is what we should focus on, according to Buchanan and Moore, rather than investigating how the interests of V’s citizens are threatened by A’s attack.
This is, I think, an important point. Considerations of institutional design clearly play a crucial role in determining what is and what isn’t permissible within a given legal framework. But these considerations are not dispositive. The mere fact that international law permits states to act in a certain way does not mean that so acting is always morally permissible. Once we have addressed the question of which defensive prerogatives should be granted to states as a matter of institutional design, we still need to ask the question of whether states are morally permitted to help themselves to these prerogatives in specific cases of purely political aggressions.5
One of the strongest arguments in favor of the view that they are permitted to do so appeals to considerations of deterrence. But while the benefits of deterrence should be factored into the proportionality calculation when we decide whether a particular defensive war is permissible, they do not necessarily tip the balance in favor of permitting it. Compare: it is true that if I’m not allowed to use defensive force to protect my tie clip, the risk of other attacks on people’s tie clips will increase. But clearly this would not be enough to justify the use of lethal force to protect my tie clip (particularly when doing so will collaterally harm innocent parties), even once we factor in the value of deterring future attempts to steal other tie clips. Of course, things might be different in the case of national defense. But this is what we need to show. It’s not enough to point at the deterrent effect of permitting wars of national defense. We need to show that protecting political goods is sufficiently important to justify inflicting lethal defensive harm on innocent parties, at least when failure to do so would increase the risk of future attacks on these goods. To this task I turn in the remainder of this paper.
3 Political Self-Determination
One of the problems with the current way of approaching the question of the justification of wars of national defense is that different political goods such as territory, culture, or political self-determination are often discussed interchangeably. (See, for example, Frowe 2014; Lazar 2014.) This is a mistake, as the interests we have in protecting each of these goods are very different, and there is no reason to expect an argument for the permissibility of waging war in defense of one of them to be also able to justify war in defense of the others.
Here I offer an argument that focuses on the value of political self-determination. I argue that wars of national defense may be permissibly waged because individuals can have a vital interest in political self-determination. Thus, my argument will vindicate V’s right to wage war in defense of territory or culture only to the extent that retaining control over its territory or its culture is instrumental to, or constitutive of, V’s capacity to exercise political self-determination (a question about which I remain agnostic).6
Political self-determination normally refers to the process through which political communities govern themselves by autonomously choosing how to act. This characterization is correct as a first approximation, but leaves out a crucially important aspect of the notion, namely that in autonomously deciding how to act, communities in question shape their own identity as a result of their choices. This is after all, quite literally, the point of self-determination: that an agent “determines itself” through the exercise of its own agency.
There are at least three ways in which a political community can engage in a process of self-determination:
- i.the Collective Intentions Model: members of the political community develop “collective intentions,” whereby they adjust their conduct in response to each other’s mutual expectation that these intentions are being collectively pursued by all the members of the community;
- ii.the Delegation Model: the members of the community delegate someone to make decisions (within certain constraints) on behalf of the whole group;
- iii.the Deliberative Model: there is a deliberative process that enables the agency of the individual members of the political community to be combined in such a way that the outcome of their collective deliberation can be said to be the expression of the agency of the group.7
These three models are not mutually exclusive. They all seem to be plausible ways of articulating the central idea that political self-determination consists in the capacity of political communities to self-govern by choosing how to shape themselves in light of the choices they autonomously make.8 And political communities do seem to rely on all of them, at least to some extent; so the best way to conceive of them is as overlapping. They constitute different ways of giving expression to the central insight that the agency of a political community is not merely the product of a number of uncoordinated instances of individual agency. Political self-determination requires more than that. It requires that the members of the political community must be “unified in certain ways” and “together constitute a unit, a collective self” (McMahan 1996: 8). (See also Buchanan 2015). If all we have is a collection of individuals acting independently, what we witness is merely the sum of individuals acting, rather than genuinely collective self-determination.
An investigation of the differences between the three models above is beyond the scope of this paper, and of only secondary importance for my purposes here. What I am interested in is rather what the three models have in common, and in particular the two following elements: First, these three models ultimately ground political self-determination on a particular type of relationship between individual agency and collective agency. They do not appeal to metaphysically suspicious collective entities, whose agency is independent of the agency of its members. Second, they presuppose that the way in which collective agency is exercised reflects the way in which the members of the group exercise their individual agency. Indeed, the link instituted between individual agency and collective agency is quite strong: although the latter is not merely the sum of a number of instances of the former, there is a sense in which in each of the three models, the members of the group act with an eye to determining the way in which the group will act. They do so either by forming the intention that the group will act in certain ways (and by responding to the mutual expectations that this intention generates among the members of the group), or by granting someone the power to make decisions on behalf of the group, or by giving their input in the deliberative process that will eventually lead the group to act.
When they vote, pay their taxes, or engage in public discussion of important issues concerning their political community, at least when they do so on a regular basis, the members of the political community are engaging in one of these forms of agency. And it is in this relationship between the exercise of collective agency (how the collective acts) and individual agency (how the members of the collective act) that ultimately political self-determination is grounded. Call this account of political self-determination “interactional,” since it rests on the value of certain forms of interaction among the members of the political community.
Notice that the interactional account does leave room for recognizing the importance of shared values and traditions for political self-determination. For the way in which the members of a particular group will engage in these three modes of collective agency inevitably will be shaped by certain foundational values and shared traditions. For example, egalitarian societies will tend to rely primarily on some form of democratic government, whereas non-egalitarian societies might rely primarily on other variants of the collective intentions or of the deliberation model. And of course there are several different ways in which the values of equality can be declined, leading to different models of democracy being adopted by different egalitarian societies (Held 2007). So, there are two ways in which particular values inform the interactional framework:
- (a)the interaction takes place by appealing to certain values and traditions;
- (b)the interaction itself, and the way in which it is structured, are the expression and manifestation of certain values and traditions.9
An example should help make the picture more vivid, so consider the Italian political community. Many Italians10 seem to see themselves as making decisions together. For example, they vote in national elections, with an understanding that the elected government will have the right, as well as the responsibility, to choose policies and laws on their behalf. It is important to these individuals that the government in charge of making such decisions is the one they have delegated to do so, because only in that case is there a sense in which those choices can be attributed to them. That attribution clearly would be out of place if it turned out that the government had been secretly chosen by a foreign power that had hacked the voting ballots. In that case, Italians would understandably feel that someone else had taken control of their own laws and policies.
Delegation is just one aspect, though a prominent one, of the way in which the members of the Italian political community act together. Obeying Italian law and paying taxes is another important one. This is how members support the political institutions through which their community will make and implement its decisions. Referenda are another example. By voting on a specific issue (or set of issues), Italians engage in a process of collective deliberation that will lead the community to act in certain ways rather than others. Although the result of elections and referenda is never aligned with the preferences of all the members of the political community,11 there is a sense in which it bears the mark of their agency, insofar as they all contribute to determining it by exercising their right to political self-determination. The result is not an expression of the will of the Italian political community because each member wills it, but because it is an expression of how all members have intentionally interacted in order to produce it.12
And of course, the way in which Italians take part in this process of collective deliberation is informed by particular values and traditions. For example, Italy is an egalitarian society, where all adults have a right to vote and run for office. In this sense, equality, however imperfectly realized, is a value that structures the very way in which Italians interact in the political sphere. But equality is also one of the values that is routinely invoked when Italians interact within that sphere. For example, they campaign for certain policies or reforms in virtue of the fact that equality demands them, and reject others in virtue of the fact that they are discriminatory. By contrast, Catholic values, once regularly invoked to justify the adoption of particular policies, have played an increasingly limited role in public life in the last 50 years (though the Church still has a significant influence on many political parties).
Necessarily, I’m painting with a broad brush here. But suppose now that another country were to forcibly change some of these features – say, by depriving some Italians of the right to vote or run for office, or by requiring Italian laws and policies to be justified in light of particular religious values. Once again, that would significantly undermine the capacity of the Italian political community to autonomously choose how to act and shape its own future. The capacity of the Italian political community to exercise its self-determining agency would be seriously compromised.
To be clear, nothing I’ve said so far is meant to vindicate the view that Italy is permitted to wage wars of national defense to defend such capacity. My aim in this section was rather to give a sense of what political self-determination is, so that we are better placed to think about whether the conditions of its exercise can be protected by means of war. To answer this question, we need to show not only that it is wrong to deprive Italy of its capacity to exercise political self-determination, but also that the wrong is so serious as to warrant the use of lethal force to prevent it. How can we do that?
4 Political Self-Determination and Wars of National Defense
Start by noticing that it is commonly acknowledged that lethal force may be used to defend what we might call “personal self-determination,” i.e. the capacity of individuals to autonomously choose how to shape their own life and pursue the choices they make.13 This is because the vital interests that may be permissibly defended by using lethal force are those required to ensure a minimally decent life, but a minimally decent life is not simply one in which our biological needs are fulfilled. A sufficient level of personal self-determination also plausibly counts as a constituent of a minimally decent life (Raz 1986).
Norman and Rodin themselves recognize this when they grant that lethal defensive force may be employed against enslavement (Norman 1995: 128; Rodin 2002: 48), since enslavement is the paradigmatic violation of personal self-determination. To be sure, enslavement constitutes a particularly egregious violation of our right to exercise personal self-determination, insofar as it involves the complete subjection of the enslaved person to the master. But it seems clear that even more limited forms of interference with our capacity to autonomously choose and pursue our own path in life may be similarly protected. For example, if someone were to permanently hypnotize you so as to interfere with your practical deliberation and get you to choose a partner or a job you don’t want for the rest of your life, I believe you would be permitted to use lethal force against him.14 You would be permitted to do so even if the partner or the job you would be manipulated into choosing would make you happier, or be better in some other respect, than those you would choose otherwise. How happy you are with the job, the partner or the friends you have matters less to how well your life goes than the fact that they are the partners and the job you chose for yourself.
Consider now the relationship between personal and political self-determination as it has been outlined above. We have seen that political self-determination is ultimately an expression of personal self-determination.15 While no individual exercise of personal self-determination makes a difference as to how the political group acts, the way in which the group acts is a reflection of the personal agency of its members. When we say that a group is pursuing certain goals, what we really mean is that the individuals forming the group are pursuing those goals, but they do so indirectly: instead of directly deliberating and exercising their agency, they give their input in a process of collective deliberation that might take different forms. Still, respecting the agency of the group is ultimately valuable because it’s a way to respect the agency of its members.
Indeed, with respect to certain spheres of conduct, respecting the agency of the group is the only way to respect the agency of its members. This is because while there are issues in relation to which what matters to us is that we decide individually how to shape our future, there are also issues in relation to which what matters to us is that we decide how to shape our future as part of a group. For example, it matters to me that my partner and I decide together how to raise our children, i.e. that the final decision is not simply my decision but the decision of the couple I’m part of. If you were coercively to interfere with our decision, you would be violating the vital right we both possess to autonomously decide how to raise our children. This is a right that we possess as a couple. (Neither of us has a right to decide individually how to raise our children, as long as we are a couple.) But when you violate it, you are not wronging some mysterious metaphysical entity called “the couple” or “the family.” You are wronging myself and my partner. The right of the couple to decide for itself how to act is grounded in the value that membership of the couple has for its members.16
Interestingly, this is the case even if my partner and I disagree about a specific issue and your intervention would force us to act in the way she or I personally favor. Suppose, for example, that my partner wants to send our son to a private school, whereas I want to send him to a state school. You force us to send him to a state school. In this case, I am wronged by your action, even if the result is the one I was personally hoping for. I am wronged because you have interfered with my goal to decide with my partner how to educate our child; and the respect you owe me as a self-determining individual agent requires that you refrain from interfering with the way in which I autonomously decide to exercise my personal self-determination, whether I do so individually or together with someone else. I have a vital interest in deciding how to shape my life, and this encompasses the interest to shape my life as I please, both when I act as an individual and when I act together with other people I am in a valuable relationship with.
Now, if it’s permissible for me to use lethal force to prevent you from interfering with valuable exercises of my agency aimed at shaping my life in light of choices that I personally make, it seems clear that this right also extends to exercises of my autonomous agency that I carry out as a member of a collective. After all, exercises of self-determination as a member of a collective can be no less valuable than exercises of self-determination as an individual, as my last example indicates.
To be sure, not all exercises of self-determination, whether personal or collective, are equally valuable. I exercise my self-determination in deciding how formally I should dress, but this particular exercise of my personal self-determination is not as valuable as the one concerning whom I should spend the rest of my life with or how my partner and I should raise our children. This is why if the only way to prevent you from wrongfully forcing me to spend the rest of my life with someone I don’t want is to kill you, I may do so (subject to the traditional necessity and proportionality constraints on harming third parties that apply to self-defensive killing),17 but if the only way to prevent you from wrongfully forcing me to dress more formally is to kill you, I may not.
Thus, one way to establish whether a certain exercise of self-determination is so valuable as to warrant resorting to lethal force to defend it is to consider how central the sphere of conduct affected by it is to our capacity to live the life we have chosen for ourselves. Whether I dress more or less formally has no serious impact on my capacity to live such a life because this sphere of conduct is not one I value to a significant extent. Even if ideally I’d like to dress less formally, a life in which I have to dress more formally is still one I can properly regard as my own, given the relatively marginal role that matters of fashion play in shaping the ideal of life I’ve chosen to pursue.18
But looking at how central a given sphere of conduct is to our capacity to live the life we have chosen for ourselves is not enough to determine when we may permissibly resort to lethal violence in order to prevent interference with exercises of our self-determining agency that fall within that sphere of conduct. This is because not all exercises of our self-determining agency falling within the same sphere of conduct are equally valuable. For example, spending time with my family is certainly a crucially important element of the plan of life I’ve autonomously chosen to pursue. The life I would be living if I were prevented from spending enough time with them is not one I could properly regard as my own, given how significant my ability to do so is for the ideal of life I’ve chosen for myself. This is why I may permissibly kill you if this was necessary to prevent your attempt to significantly undermine my capacity to spend time with them. But suppose now that you only attempt to prevent me from spending this afternoon with them, and that the only way to stop you would be to kill you. Clearly this would be impermissible. Although spending time with my family is indeed a crucially important element of the plan of life I’ve autonomously chosen to pursue, being prevented from spending this afternoon with them does not undermine my capacity to live a life I can properly regard as my own. This is because of the marginal impact that this form of interference would have on my capacity to realize this central component of my self-determining agency.
Thus, there are two tests that need to be applied to establish whether we are permitted to resort to lethal force to protect valuable exercises of our self-determination. First, we need to assess how central the sphere of conduct affected by a given exercise of self-determination is to our capacity to live the life we have chosen for ourselves. Second, we need to assess how seriously a given attack will set back our capacity to exercise our self-determining agency within that sphere of conduct. With these two tests in hand, we are now in a position to see why it is misleading to appeal to the voting case in order to conclude that wars of national defense are impermissible.
Suppose that I value being part of a self-determining political community,19 and that I contribute to the self-determining agency of that community by exercising my personal agency in the ways required by the interactional account. This means that when you prevent me from voting, you are indeed interfering with a sphere of conduct that is important for my capacity to live the life I have chosen for myself. But this is not to say that I may resort to lethal violence to prevent you from doing that. For preventing me from voting is a form of interference that has only a marginal impact both on the capacity of my political community to exercise its collective self-determination and on my capacity to exercise my personal self-determination in the way required for me to be able to count as a member of the community.
Preventing me from voting (even repeatedly preventing me from doing so) has a negligible impact on the capacity of my political community to exercise its collective self-determination, because my vote will make no significant difference (indeed, it will make no difference) on the result of the election(s). Provided that enough of my fellow citizens are not similarly prevented from voting, the outcome of the election will still be an expression of the will of my political community. But preventing me from voting (even repeatedly preventing me from doing so) also has a negligible impact on my capacity to exercise my personal self-determination in the way required for me to be able to count as a member of the self-determining community. For voting is only one of the many ways in which I can interact with my fellow citizens, mixing my agency with theirs along the lines suggested by the interactional model. I can do so by engaging in public debates, assembling and associating with others, campaigning and being elected, etc. Thus, just like the fact that I may not kill you if you are trying to prevent me from spending this afternoon with my family doesn’t show that spending time with them is not a central component of my chosen plan of life, one that I may protect by lethal means when it is seriously threatened, the fact that I may not kill you if you are trying to prevent me from voting, doesn’t show that being a member of a self-determining political community is not a central component of my chosen plan of life, one that I may protect by lethal means when it is seriously threatened. In both cases, what explains the fact that I may not resort to lethal defensive violence is simply that the violation of self-determining agency in question is not sufficiently serious. The first of the two tests identified above is passed, but the second isn’t.20
By contrast, when V is attacked by A, both tests are passed. All of V’s members will be prevented from interacting in the way required to make the agency of V an expression of their own agency. And to the extent that V’s members significantly value their membership in V, this will seriously compromise their capacity to live the life they have chosen for themselves. In this case, V is permitted to use lethal force, even if no other interest, other than the one in maintaining its political independence is threatened.
We can imagine A to be a benevolent colonial power, which is not threatening to violate any of V’s members’ human rights, or to harm V’s members in any other way. Even so, V is permitted to wage a war to defend its right to political independence, provided that V’s members have an interest in making decisions together. This right, as we have seen, is ultimately grounded in the weighty interest that each member has to be able to autonomously shape her life as she wishes. To the extent that this interest is a component of a minimally decent life, it can be protected by lethal means, provided that the conditions spelled out in this section are fulfilled.21
5 Two Objections
I have argued that there are two related interests I have as a member of a political community. I have an interest that my community is genuinely self-determining, i.e. that it has the capacity to exercise its collective agency in a way that enables it to shape its own identity as it wishes in light of the choices it makes; and I have an interest in being able to take part in this process by mixing my personal agency with the agency of my fellow citizens in the way described by the interactional model. When one of these two interests is set back to a significant degree, resort to lethal defensive violence is in principle permissible. A number of objections may be raised against this view. In this section I consider two of them.
One objection targets the alleged relationship between personal and collective agency that is at the foundation of my account of political self-determination. I have suggested that to the extent that how V exercises its collective agency is an expression of how its members exercise their own personal agency, wrongfully interfering with the former is ultimately a way of wrongfully interfering with the latter. Some might object that this model works for small groups, where the way in which each member exercises her personal agency can be expected to have a realistic chance of affecting the behavior of the group (as is the case with a couple deliberating how to raise their child). But the same idea is implausible in the case of large groups, such as entire political communities, where the way in which any member exercises her personal agency typically cannot affect how the community will act (Altman and Wellman 2011: 18–20).
The thought is that even if in the end my partner prevails and we send our child to a private school, I nonetheless had the capacity to affect the decision in the sense that there was a realistic chance we could have sent our child to a state school. However, when I act as a member of a political community, realistically I don’t have any capacity to affect the decision of the group, which is why my model cannot be transferred from small groups such as couples or clubs to large groups such as political communities.
The problem with this objection is that it mistakenly identifies the capacity to affect the behavior of the group as the key element on which we should focus to establish whether the way in which the group acts can be said to be an expression of the way in which its members exercise their personal self-determination. That this is not the case should be clear if we consider the following case. Suppose that, after long discussions, my partner and I decide to send our child to a state school, but that in fact we could not have done otherwise, as the private school was much more expensive than we had thought, and we couldn’t have afforded it. In this case, my partner did not have the capacity to get the group to act as she would have wanted. Sending our child to private school never was an option for us, as it turns out. And yet, intuitively the decision to send our child to a state school is one that can be properly attributed to the couple, and thus indirectly, to the members of the couple, including my partner. It is the fact that we have deliberated in the right way, exercising our personal agency as required by our membership in a family that makes it apt to read our behavior as a genuine expression of what each of us wanted.
Allen Buchanan is skeptical about this move. Focusing on democratic forms of political self-determination, he writes that “it is simply false to say that an individual who participates in a democratic decision making process is self-governing: he or she is governed by the majority. … Far from constituting self-government for individuals, majority rule … excludes self-government for every individual” (Buchanan 1998: 17–8). (See also, Altman and Wellman: 19–20).22 However, this objection mistakenly presupposes that individuals can only be self-determining if they get to act in accordance with the preferences they have formed in deliberating independently. As we have seen, this is not the case. There are issues in relation to which what matters to us is that we contribute to shaping our future as part of a group. In these cases, the correct way to respect our self-determination is to respect the self-determining agency of the group to which we belong, even when the decisions of the group do not align with our personal preferences. (Remember how I am wronged when you force my partner and I to send our child to a state school, even if that was my preference.)
The second objection I intend to consider also targets the idea of applying the model of collective agency I have employed to discuss groups such as the family to the political community. Here the thought is that while there is significant value in being a member of a couple or of a family, it is doubtful that membership of a political community is significantly valuable. If so, it may not be permissible to kill to preserve our capacity to act as members of a self-determining political community, even if it is permissible to kill to preserve our capacity to act as members of a self-determining family.
My view is that membership in our political community is indeed significantly valuable, since interacting with the other members of a political community can be a crucially important element of a meaningful life. But while I hope the discussion in the previous sections will have done something to persuade any skeptical reader, this is not a question which I can hope to fully address here. What is worth stressing is that just like being a member of a couple is valuable only insofar as one values interacting with his or her partner in a number of ways, being a member of a political community is valuable only insofar as we value interacting with our fellow citizens in the way described by the interactional model. Those who don’t take part in this interaction because they don’t value it, cannot be considered members, and although they might have reasons to obey the law and to not interfere with the functioning of said community, they don’t have special reasons to care about its existence.
Thus, membership of a self-determining political community is to some extent voluntaristic: although we need not consent to being members of the political community, we need to interact with our fellow citizens in the way described by the interactional model in order to count as members. (Call those who do so interact “self-determining citizens” (sd-citizens).) Those who don’t, should not be considered members, and although they might have reasons to obey its laws and not to interfere with its functioning, they have no special reason to care about the existence of that particular community. (Call these “non self-determining citizens” (nsd-citizens).) This raises a problem for the account of wars of national defense I’ve offered.
I have argued that the members of V are permitted to wage a defensive war against A in order to protect their capacity to function as a self-determining political community. This is the equivalent of my right to kill you in order to prevent you from choosing for us how my partner and I should raise our children. Now, if all the citizens of V endorse their membership in the political community, then the parallel with the couple holds. But things get more complicated if some of the citizens of V do not endorse their membership, and thus fail to be part of V’s political community, as is the case with nsd-citizens. nsd-citizens might have reasons not to interfere with V’s exercise of political self-determination, but they have no reason to prefer being citizens of V to being citizens of A. In the case of a political aggression, nsd-citizens will clearly be in favor of surrendering, and on the face of it, V’s failing to do so seems impermissible, since it would impose significant costs on these citizens (exposing them to the risk of the harm inflicted by the war) in order to protect the political interests of sd-citizens.
The same problem is raised by children and other subjects who are unable, rather than unwilling, to exercise their personal agency in the way required for membership of a self-determining political community. If V has a right to wage a war of national defense against A grounded in the value for its members of being part of a self-determining political community, this right cannot be grounded in the value that being a member of V has for subjects who do not or cannot engage in the sort of interaction that grounds political self-determination.
I will start by addressing the general problem, and then will focus on the case of children, as the latter raises special considerations. The answer to the general problem is that if some members of V do not value membership of it, then that counts against the permissibility of V waging a war of national defense. This is not to say that V is not permitted to wage such a war, but that it can do so only if the interests of V’s sd-citizens in political self-determination are weightier than the interests of nsd-citizens not to be exposed to the danger of collateral killing. To establish whether V can wage a war of self-defense we need to trade off these competing interests.
I will not say much here about how weighty each of these interests is, save to mention that it is conceivable that the interest in preserving a self-determining political community can be strong enough to override the competing one. This is because it seems plausible that, under certain circumstances, we may defend valuable exercises of our self-determining agency, even if doing so comes at the cost of inflicting unintended foreseeable (non-manipulative) harm on others.23 After all, most people agree that it is permissible to inflict unintended foreseeable (non-manipulative) lethal harm on a limited number of innocent parties if this is necessary to prevent a sufficiently serious unjust harm from being inflicted on a much larger number of innocent parties. If so, this must be true not only when the harm prevented takes the form of injuries to life and limb, but also when it takes the form of significant interference with the capacity of individuals to exercise their right to self-determination. For our interest in personal self-determination is a constituent of a minimally decent life, and thus counts as a vital interest.24 If personal self-determination may be defended by lethal force, despite the fact that this might involve inflicting unintended foreseeable (non-manipulative) lethal harm on others, political self-determination may also be defended in this way since, as we have seen, it is ultimately an expression of personal self-determination.
As I have mentioned, the case of children raises special considerations. On the one hand, parents have a special duty of care toward them, and an important component of this duty is certainly the duty to minimize the risk of harm to which they are exposed. This counts as a reason in favor of parents choosing that V surrenders to A. On the other hand, both parents and children might have a significant interest in the latter having the opportunity to grow up as part of V, rather than A (or perhaps A+V), since this would allow the children to be exposed to what the parents regard (and children might grow up to regard) as an important source of value and a crucial component of their own identity. Sharing membership of V’s political community is something that parents and children might value non-instrumentally, as a constitutive element of their own plan of life, but also instrumentally. For it might be too hard, at least in some cases, for parents to fully adapt to the social and cultural practices of A; and if so, there’s a risk that they will never fully see themselves as meaningfully belonging to A’s political community.25 The problem here is that if their children do, parents might struggle to interact with them in meaningful ways, insofar as they will not be sharing the same values and traditions. If so, this shared interest will count as a reason in favor of V waging a defensive war against A’s attack.26
Here the following objection might be raised: Insofar as the permissibility of waging wars of national defense, according to the account offered above, depends on the trade-off between the interests of sd-citizens and nsd-citizens, the account is ultimately aggregative, and thus subject to the same objections I have raised against Frowe’s approach. But while my account does indeed have an aggregative dimension, it is not aggregative in the same way in which Frowe’s account is. On Frowe’s view, whether V may permissibly defend itself against A’s attacks depends on how large V’s population is. Only if the number is sufficiently large as to make the interests of V sufficiently important, may it be protected by the use of lethal force. This leads to the objection that, counterintuitively, the permission to wage wars of national defense will be granted only to large political communities. The same problem does not afflict my view, as according to it, no matter how small V is, insofar as a sufficiently large share of its population values its membership of the political community, V may permissibly resort to lethal force to defend itself (provided that traditional jus ad bellum principles are met). True, if enough members of V fail to value their membership of it, V may not wage a defensive war against A, but this is as it should be. No member of V has a right that V persists as a political community. They only have a right that V persists if this is what V’s members want.27
The aim of this paper has been to articulate a justification for the permission to wage wars of national defense. The central notion that I have invoked to this end is the notion of political self-determination, understood as the capacity of political communities to decide how to shape their own identity through the exercise of their autonomous agency. I have argued that the interest we have in taking part in the process of political self-determination of our political community (provided that we value such membership) is a vital one, in that it’s an expression of our more fundamental interest in exercising our personal self-determination. If we may kill to stop interferences with valuable exercises of our agency with respect to the choices we make individually, this right extends to valuable exercises of our autonomous agency with respect to the choices we make as members of a collective. This enables us to justify the permissibility of waging wars of national defense without having to give up reductivism or individualism.
Massimo Renzo is Professor of Politics, Philosophy & Law at King’s College London. He has held visiting appointments at the Australian National University, the universities of Virginia and Arizona, the National University of Singapore, the Centre for Ethics and Public Affairs at the Murphy Institute (Tulane University) and Osgoode Hall’s Nathanson Centre for Transnational Human Rights, Crime & Security. He is an affiliated researcher at the Stockholm Centre for the Ethics of War & Peace and the Honorary Secretary of the Society for Applied Philosophy. He is also one of the editors of the journal Criminal Law and Philosophy.
Earlier versions of this paper were presented at a workshop at the Carnegie Council for Ethics and International Affairs in New York, as well as at a number of seminars and conferences at the Universities of Chicago, Stirling, Loyola (New Orleans), Amsterdam, Bristol, Pavia and at the London School of Economics. I am grateful to all those audiences as well as to Helen Frowe, Jeff McMahan, Victor Tadros and Lorenzo Zucca for valuable discussions. Special thanks are owed to Saba Bazargan, Allen Buchanan, Cécile Fabre, Seth Lazar, Adam Slavny and Laura Valentini for helpful written comments.
Fabre Cécile . 2014. “Cosmopolitanism and Wars of Self-Defence.” In The Morality of Defensive War, edited by Fabre Cécile and Lazar Seth . Oxford: oup.
Hurka Thomas . 2011. “Proportionality in the Morality of War.” In Drawing Morals: Essays in Ethical Theory, 238–64. Oxford: oup.
Moore Margaret . 2014. “Collective Self-Determination, Institutions of Justice, and Wars of National Defence.” In The Morality of Defensive War.
Stilz Anna . 2016. “The Value of Self-Determination.” In Oxford Studies in Political Philosophy, Volume 2, edited by Sobel David , Vallentyne Peter , and Wall Steven , 98–127. NewYork: oup.
van der Vossen Bas . 2016. “Self-Determination and Moral Variation.” In The Theory of Self-Determination, edited by Teson Fernando R. , 13–31. Cambridge: cup.
Given that the political good in question need not be the protection of a particular nation, the label "wars of national defence" is misleading. But since it is commonly used in the philosophical literature (and for lack of a better label), I’ll stick with it.
Hurka only takes his argument to justify the weaker claim that V is permitted to use more force to protect its right to exercise its political self-determination than it is permitted to protect one person’s isolated exercise of her right to vote.
The same reply can be offered to Buchanan’s argument that it is permissible for V to defend itself if there is reason to believe that, once in control of V, A will violate vital interests of V’s citizens (Buchanan 2016). Effectively this argument is a justification for the use of lethal violence to pre-empt future attacks to life and limb rather than to avoid current attacks to political goods. Fabre acknowledges this (Fabre 2014: 105–6, 112) and bites the bullet, concluding that wars in defense of purely political goods are impermissible.
The same is true in domestic law. There might be good reasons why domestic law does not require that we give the food we don’t use to those who need it, but this is not to say that if I see a starving person, it would be morally permissible for me not to feed her.
Moore 2014 and Stilz 2014 also appeal to the value of political self-determination, though their arguments differ from the one I offer below. Moore’s argument appeals to associative duties and relies on the sort of institutional considerations I discuss on pp. 712–3. Stilz’s qualified defence focuses primarily on the territorial rights associated with political self-determination. (Like Fabre, Stilz doubts that her arguments can justify waging war to defend purely political goods [Stilz 2014: 228].)
The Collective Intention Model is built along the lines of the theories of collective agency developed by Michael Bratman, Margaret Gilbert, Chris Kutz, and John Searle. The Deliberative Model is based on the account recently developed by Christian List and Philip Pettit (List and Pettit 2011.)
Recent accounts of political self-determination have tended to focus on the Collective Intention Model (Moore 2015; Stilz 2016). For an attempt to develop the Deliberative Model, see van der Vossen (2016).
This is not to say that the values and the traditions in question need to be those of a particular culture (Kymlicka 1989), nation (Miller 1997) or “encompassing group” (Margalit and Raz 1990). The interactional account is compatible with the view that political communities are structured around values drawn from a plurality of cultures present in a certain territory (Waldron 2010).
I say “many” and not “all” for a reason which I will discuss below, in section 5.
I address this objection below, pp. 724–5.
While some Italians might see themselves as sharing a particular intention when their political community decides to act in a certain way, others might not. They might simply see the process of collective deliberation as a way of reaching decisions together, without assuming that they share any specific intention about how the group should act. This is where the difference between the Collective Intentions Model and the Deliberative Model lies. (In this respect, the former is more demanding than the latter.) List and Pettit famously argue that there can be situations where, even if we operate with a fairly simple procedure (such as the majoritarian rule), a group might act in a way that does not align with the intentions of any of its members, let alone all of them.
Thus, “personal self-determination” denotes what philosophers sometimes call “personal autonomy.” Joseph Raz (1986: 370, 204) for example, describes personal autonomy as “an ideal of self-creation.”
Assume for now that your defensive attack would not also produce collateral harm on innocent parties. I address this complication below, pp. 725–7.
As Raz puts it, “[r]ights, even collective rights, can only be there if they serve the interests of individuals. In that sense collective interests are a mere façon de parler. They are a way of referring to individual interests which arise out of the individuals' membership in communities” (Raz 1986: 208).
Henceforth, this qualification will be taken for granted. On the proportionality constraints on the use of lethal defensive force to protect self-determination, see footnote 23 and the surrounding text.
Things would be different if, for example, I had chosen to pursue an ascetic life and being forced to dress formally would undermine this important project of mine.
More on this condition below (section 5).
Things would be different if many of my fellow citizens were also prevented from voting, or if I were prevented not only from voting, but also from interacting with them in most of the ways required by the interactional model of political self-determination (say, if I was constantly prevented from engaging in public debates, campaigning, etc.). In these cases, resort to lethal force would be permissible (subject to necessity and proportionality constraints on harming third parties being respected).
A caveat is in order: my defense of the permission to wage wars of national defense is grounded in the value of protecting the capacity of a political community to exercise its self-determining agency. But there are obvious limits to what falls within the scope of our self-determining agency. Consider again personal self-determination: I have a legitimate objection to you interfering with my autonomous choice of a career or of a spouse, because I have a right to choose which career to pursue or whom I should marry. But I cannot legitimately object to you interfering with my autonomous choice to murder someone or to discriminate against women in the workplace. No legitimate objection can be raised in these cases, because I do not have a right to engage in the relevant conduct. For the same reason, V lacks a legitimate claim against (necessary and proportionate) military intervention aimed at preventing serious human rights violations. Engaging in serious human rights violations simply falls outside the scope of what V is morally permitted to autonomously choose to do. However, humanitarian intervention, even when permitted, is meant to be temporary. The intervening party may stop the human rights violations in V and, under certain conditions, oversee the process of the rebuilding of V’s political institutions, so that V can start exercising again its political self-determination. As such, this sort of military intervention is different from the one discussed in this paper, which is not temporary and whose aim is to undermine V’s capacity to exercise its political self-determination. In [Renzo 2018a], I discuss the conditions that respect for political self-determination imposes on military intervention. In [Renzo 2018b], I argue that the intervening party is not permitted to use military force to impose on V constitutional arrangements or political institutions that V has not autonomously chosen, even if they would be better likely to guarantee V’s future capacity to function as a self-determining community.
In this context “self-determining” and “self-governing” denote the same thing.
On the difference between “eliminative” and “manipulative” harm, see Quinn 1993, chap. 7–8; Tadros (2013: 242–6).
This view is compatible with the claim that our vital interest in personal self-determination is less important than other vital interests we have, such as the interest in bodily integrity. I myself do not accept this claim (no doubt, we would often prefer to suffer setbacks to our interest in bodily integrity rather than setbacks to our interest in self-determination, at least in its personal variant). But if the claim is true, what follows is simply that the number of violations required to justify inflicting unintended foreseeable (non-manipulative) lethal harm on innocent parties will be higher in the case of our interest in self-determination than in the case of our interest in bodily integrity.
As I mention above (pp. 715–7), the interactional account does not reject the importance of shared values and traditions for self-determination. Rather, it claims that these values and traditions constitute the background against which the relevant interactions of the members of the political community take place.
This is true in the personal case too. If you’re wrongfully trying to prevent me from spending the next 30 years with the person I love, I may permissibly kill you (if necessary), provided that she wants to spend the next 30 years with me. Only in that case you wrong me in separating me from her. I am obviously not permitted to kill you if that’s not what she wants (say, because she chose to share her life with you instead).