When the Bush and Blair administrations justified the 2003 war on Iraq as an act of preemptive self-defense, this was greeted in many quarters with understandable skepticism. How can the right of self-defense be legitimately invoked when no prior aggressive attack has occurred and there is no evidence that one is imminent? This question, much debated in the months leading up to the war, invites us to reflect critically on the content of the right of self-defense. Yet there is a deeper question to be asked about the idea of a war of self-defense; namely, how is it that war can be considered an act of self-defense at all? How exactly is it that the concept of self-defense can provide a justification for war? It is this question that I ask in War and Self-Defense and the answer I arrive at is a surprising one.1
It may seem an odd question even to ask. Surely a nation, or at least a state, has a right to use military force in national defense in the same way that an individual who is attacked has the right to kill in self-defense. This seemingly straightforward idea, which Michael Walzer memorably called "the domestic analogy," has been a prominent feature of much political, legal, and philosophical thought on the ethics of war in the Western tradition. Indeed, so fundamental is the idea of an analogy between self-defense and national defense that it is difficult to imagine how any theory of just war could proceed without it.
It is fundamental but it is also, I contend, wrong. There is no valid analogy between self-defense and national defense, and the attempt to model international law and international ethics on domestic law and interpersonal ethics is philosophically misconceived and unhelpful in practice.
A Moral Theory of Self-Defense
In order to support such a claim it is necessary to begin with a proper understanding of the right of self-defense. We need a moral theory of self-defense whose function is first to explain the structure, operation, and limitations of the right, and second to provide a moral grounding to the right; to explain why it is that defenders are morally entitled to kill aggressors in situations of self-defense.
In order to generate such a theory, I first analyze the moral structure of self-defense, using as a starting point Hohfeld’s famous analysis of legal rights. Self-defense is constituted by a set of normative relations between four elements:
- the subject of the right (the defender)
the object against whom the right is held (the aggressor)
the act that is the content of the right (in typical cases of self-defense this will be homicide)
the end of self-defense (the good that the defensive act is intended to protect or preserve—this may be the defender’s own life, the life of a third party, or some other valuable, such as property or liberty)
By paying attention to the relationships between these elements we can begin to explain many features of the operation of rights of defense in their most general form. For example, rights of defense arise out of a normative relationship between the subject of the right and the end that he or she is seeking to protect: either the subject has a "right to" the end of defense (as when one defends one’s own right to life from attack) or the subject has a duty of care over the end (as when a parent defends his or her child) or the subject is acting out of a general duty of rescue (as when a defender comes to the aid of a third party who is being attacked). The rights and obligations of defense differ subtly in each of these cases.
Self-defense is partly grounded in the normative relationship between the defender and the end of her action, but it is also importantly grounded in the fact that the end she is seeking to protect is a good or a value sufficient to merit harmful acts in its defense. This gives rise to the three most significant limitations on the right of self-defense: proportionality, necessity, and imminence. These say roughly that an act of self-defense is only justified if the harm inflicted in the course of defense is not greater than the harm it seeks to avert (proportionality); that there was no less-harmful way to achieve the same result (necessity); and that the harm one seeks to avert is truly imminent.
Analysis of this kind can reveal much about the operation of the right of self-defense and the limitations implicit within it. But it leaves unanswered the fundamental question concerning the justification of that right: Why is it that defenders are morally entitled to kill aggressors in situations of self-defense, given that we think that the right to life is universal, and therefore presumably held also by the aggressor?
To answer this question broader theories of ethics need to come into play. For example, according to utilitarian theory killing in self-defense is justified as a preference for the lesser evil. But the utilitarian theory fails because it gives an inadequate account of our intuitive understanding of the right. Similarly, Cheney Ryan and others have developed a highly suggestive account according to which self-defense should be considered as a "forced choice" between lives for which the defender has limited responsibility. But I argue that this approach fails because it misconceives the nature of responsibility.
In my view the best way of justifying self-defense is through an account of the interaction of rights in situations of violent conflict. In brief, when a defender justifiably kills a homicidal aggressor in self-defense, the aggressor no longer has the right not to be killed, and the reason for this is that the aggressor is morally at fault for the attack, while the defender is innocent.
Underlying this account is a particular view of the nature of moral rights. First, rights such as the right not to be killed are most fundamentally normative relations between two people. This is true in the strictly logical sense: A's right not to be killed by B is the logical correlate of B's duty not to kill A. Correspondingly, A’s right to kill B in self-defense is the logical correlate of B's failure to possess the right that A not kill him. But beyond its purely logical form, I believe that the right not to be killed is grounded in an interpersonal normative relationship. Thus rights against being killed are dependent on a relationship of reciprocity: one has the duty not to kill or harm others just so long as they adhere to the same duty toward you. It is the breakdown in this relationship of reciprocal respect constituted by the act of aggression that explains why aggressors fail to have the right not to be killed by their victims, and why defenders possess the right to kill in the course of self-defense.2
Second, what tips the balance of interpersonal rights in this way, depriving the aggressor of the right not to be killed, must be some feature that has a normative connection with the aggressor viewed as a moral subject—paradigmatically it will be an act for which he has moral responsibility. As I put it in the book, to say that a man has the right to life "is to say that he has an interest in his living which cannot be overridden except on the basis of . . . his own choosing, willing, or acting. He can waive his right to life, or forfeit it (for instance, by violating certain laws or obligations), but we cannot override his right simply on the basis that it would be congenial or useful or even imperative for us to do so." "The reason for this is that the most basic function of a morality of rights is to locate certain extremely important normative considerations wholly within the sphere of the subject itself so as to make them unassailable by external contingencies such as are appealed to by consequentialism" (pp. 89, 88). A corollary of this principle is to rule out defense against certain "innocent aggressors" and "innocent threats" that many commentators have thought appropriate objects of self-defense.3
From Self-Defense to National Defense
Understanding self-defense in this way has significant consequences when it comes to extending the notion of self-defense to one of national defense capable of justifying war. First, if defensive rights are grounded in a set of normative relations among different elements of the right, then it is important to understand exactly what kind of relation war is, and what the different elements of national defense—subject, object, content, and end—actually are. Yet here arises a problem, for there is a basic dichotomy between what I call "the two levels of war": war may be described as both a relation among persons and as a relation among super-personal political entities such as nations or states. It is therefore an open question as to whether national defense is properly conceived as a right held by and against individuals, or by and against states.
This in turn generates two potential strategies for vindicating a right of national defense. The first is a reductive strategy that seeks to explain national defense as a special application of the right of personal self-defense. The second is an analogical strategy that takes seriously the idea of state-held rights and tries to account for them through an analogy with the personal right of self-defense.
Understanding self-defense in this way has significant consequences when it comes to extending the notion of self-defense to one of national defense capable of justifying war. First, if defensive rights are grounded in a set of normative relations among different elements of the right, then it is important to understand exactly what kind of relation war is, and what the different elements of national defense—subject, object, content, and end—actually are. Yet here arises a problem, for there is a basic dichotomy between what I call "the two levels of war": war may be described as both a relation among persons and as a relation among super-personal political entities such as nations or states. It is therefore an open question as to whether national defense is properly conceived as a right held by and against individuals, or by and against states.
This in turn generates two potential strategies for vindicating a right of national defense. The first is a reductive strategy that seeks to explain national defense as a special application of the right of personal self-defense. The second is an analogical strategy that takes seriously the idea of state-held rights and tries to account for them through an analogy with the personal right of self-defense.
The reductive strategy in its purest form (which treats national defense as nothing but a complex and coordinated exercise of the right of personal self-defense) is a clear failure. Soldiers engaged in a defensive war (and according to classic just war theory, the aggressive soldiers as well) enjoy a set of permissions that extend far beyond the rights of personal self-defense. They may, for instance, kill uniformed soldiers not engaged in combat. Furthermore, an army fighting a defensive war may engage in a sustained counteroffensive. It would be difficult to view such acts of military attack as cases of personal self-defense on the part of the soldiers.
But the reductive strategy could be interpreted more broadly. One important element of any justification of national defense must be the identification of a defensive end of sufficient value to make the harms inflicted in the course of a defensive war proportionate. Perhaps individuals are not the subjects of the right of national defense, but defending their lives may nonetheless be its end (as when one talks of the state "defending its people"). This runs contrary, however, to the way that the right of national defense is usually conceived in just war theory and international law. For a state can have a right of national defense even when the lives of none of its citizens are threatened, as for instance in the event of a bloodless invasion. Moreover, having the end of defending the lives of citizens is not sufficient to make military action that is proportionate and necessary an instance of national defense. Consider humanitarian intervention, which has the end of saving the lives of citizens of another state and therefore may be compared to third-party self-defense in criminal law: not only is humanitarian intervention not an instance of national defense according to traditional just war theory, but on the contrary it is a normative consideration in deep tension with a state’s right of national defense. Fundamentally, the end of national defense cannot be construed in purely personal terms, for national defense may be considered justified even when the lives of no citizens are threatened by an act of aggression and even when more citizens would be threatened by fighting the war than by not fighting it.
Perhaps the analogical strategy fares better in providing an account of the end of national defense. After all, it is natural to regard war as a means to protect the value inherent in what we might call the "common life" implicit in a nation or state. The challenge here is to provide an interpretation of the value of the common life in such a way that it could function as grounds for a right of national defense. The task is not an easy one, for the account must be minimal enough to ground a right for all states irrespective of their internal constitution, yet it must still identify an objective good of sufficient value to ground a universal right of national defense.
For instance, we might explain the value of the common life through an account of state legitimacy such as Hobbes's social contract theory, which explains the moral value of the state through its role in removing us from the horrors of the state of nature. Hobbes is useful because he provides a very minimal account of legitimacy and thus offers the hope of grounding a right of national defense shared by all states. Yet if the good represented by the common life is shared by any effective sovereign power (as it is for Hobbes), then it is unclear why a political community should have the right to defend itself from invaders who seek to conquer and rule. Minimal accounts of state legitimacy such as Hobbesian contract theory provide no reason to defend our particular form of common life against others.
On the other hand, if the value of the common life is thought to reside in its unique character or identity, then the converse problem arises: it is not clear that such a value is sufficiently objective to ground a universal right of national defense. This is because the character of a nation’s common life may be oppressive, unjust, or cruel. How could such a common life (not uncommon in today’s world) constitute a value sufficient to make proportionate the harms of a defensive war? John Stuart Mill and Michael Walzer both indicate an intriguing solution to this problem. Both claim that it is not the character of a nation’s common life per se that merits defense and protection, but rather a people’s collective autonomy that underlies the national character. Yet when one attempts to spell out what this collective national autonomy could amount to in practice, it is all but impossible to construe it as the expression of an authentic moral value.
Moreover, autonomy and communal identity are values of communities. How, then, could such a value ground a right of national defense for states, given that most states are not identical with a unitary national community and, indeed, many states are in relationships of antagonism with national communities? Conversely, if rights of national defense are grounded in the values of communities, how are we to stop the right of recourse to armed defense from proliferating to the multitude of non-state communities?
These difficulties arise from attempting to reconstruct just one aspect of the supposed right of national defense: a putative end sufficient to satisfy the requirements of proportionality. Yet equally grave challenges face the supporter of national defense when it comes to explaining the remaining constitutive elements of the right. To give but one example, if the object of the right of national defense is conceived as the aggressor state, then we need an explanation of why possessing a right of defense against the aggressor state entails possessing a right to kill the soldiers of this state, given that the soldiers are themselves moral subjects with rights that, as I suggested above, can only be alienated through some action for which they are normatively responsible. I argue that, at the very least, this will require us to abandon the traditional just war claim that soldiers are not morally responsible for participation in an unjust war so long as they fight in accordance with the war convention.4
Toward a New International Ethics
Most of the difficulties with vindicating the domestic analogy between self-defense and national defense stem ultimately from the complexities inherent in the two levels of war—the duality between war as relation among persons and war as relation among states. The problems raised by this duality have important practical as well as theoretical consequences. A basic task of international ethics is to explain how the rights and values of states are related to the rights and values of individual persons, and how we should resolve value conflicts between them, yet this question is obscured by an approach that treats states as though they were individuals writ large. In the last decade, this difficulty has emerged most concretely in the case of humanitarian intervention, where the defensive rights of states come into direct conflict with the rights of individual citizens. Appeal to the domestic analogy between individual and state has been decidedly unhelpful in resolving the difficult normative issues at stake here.
Rejecting the methodological primacy of the domestic analogy invites us to reconsider the foundational elements of our moral theory of international relations in ways that will better enable us to handle problem cases like humanitarian intervention. We must develop a rejuvenated international ethics that pays greater attention to the complex and nuanced normative relations that exist among persons and, crucially, between individual and state, rather than simply focusing on state-state relations conceived on the model of super-persons.
* The editors would especially like to thank David R. Mapel for helping to guest edit the articles in this symposium. [BACK]
1 David Rodin, War and Self-Defense (New York: Oxford University Press, 2003). All in-text citation references are to this book. [BACK]
2 The operation of reciprocity is apparent also in other elements of the right; for example, in the notion of proportionality. Though it appears at first sight to be a form of consequentialist requirement, proportionality does not in fact require us to balance the overall harms and benefits of an act of defense, for an innocent defender is entitled to kill literally any number of culpable attackers. On the contrary, proportionality requires us to balance the harm inflicted upon a particular person, against the harm with which he threatens us. Proportionality in self-defense is therefore best regarded as deriving from a kind of interpersonal moral reciprocity. [BACK]
3 Understanding which innocent threats and innocent aggressors one may properly harm in self-defense requires a detailed examination of the theory of excuse. [BACK]
4 This may provide grounds for rehabilitating the idea of military force as a form of law enforcement or even punishment, though careful attention must be paid to the idea of authority: in most cases states at war will lack the authority required for proper punishment or law enforcement. [BACK]
