The Nature of a Threat

Professor Stephen Kershnar, SUNY/Fredonia

 

      The notion of a threat is central to the theories of permissible wartime killing, because self-defense and defense of others warrant most, if not all, wartime killing and they focus on threats. Thus, we need a theory of threats. This theory attempts to provide one. It has three parts. In the first part of the paper, I set out a theory of self-defense and note that it focuses on threats. My assumption here is that defensive purposes justify most, if not all, just wartime killing. In the second part, I analyze the notion of a threat. I argue that a person is a threat if he brings about a significant amount of expected injustice. In the third part, I briefly note that given this analysis of a threat, a government may in some cases target civilian leaders (including legislators), military-production workers, and persons who provide for the needs of persons qua human beings (e.g., physicians and farmers).

 

PART ONE: SELF-DEFENSE FOCUSES ON THREATS

  1. The Right of Self-Defense is Held Against Threats

I shall assume that just wartime killing is justified by self-defense and defense of others. Following this assumption, I begin by briefly sketching out the right of self-defense and then apply this right to wartime killing.

One person has a right to use force against a second if and only if the first has a moral liberty (i.e., under no moral duty not) to use force against the second and a claim of non-interference against that person. Where the right is grounded in the fact that the second person endangers the first, or does so in a particular way, the right is one of self-defense. The right to self-defense is usually accompanied by a claim to non-interference against third persons, although this claim is not part of the right. For example, one person, Al, can have a right to act in self-defense against a second, Bob, even though he has waived his right to non-interference in the exercise of that right against another, Charles. Under such circumstances Al has the right to defend himself against Bob’s attempt to batter him by punching Bob in the face, even though he has no claim against Charles that the latter allow him to punch Bob.

A couple of intuitions confirm that we have such a right. First, a person who uses force against an unjust aggressor does not act wrongfully or unjustly. Second, the defender does not owe compensation for the harm he imposes on the unjust aggressor in the course of defending himself. That both of these intuitions apply in hypothetical cases in which there is neither a government nor a society suggests that this right is a natural one.

The right to self-defense is held against unjust threats. Roughly, a threat is a cause of a process that will likely harm a person. A threat is unjust if and only if the person performing it infringes on someone’s rights in bringing about the likely harm. This notion might have to be modified to take into account omissions that prevent or help to prevent a person from being harmed since omissions do not cause other events.

Threats can include persons who are morally innocent, i.e., they are not blameworthy for being a threat. Here are two examples of innocent threats.

 

Psychotic Aggressor: A person in an elevator goes berserk and attacks you with a knife. There is no escape: the only way to avoid serious bodily harm or even death is to kill him. The assailant acts purposely in the sense that his means further his aggressive end. His actions are frenzied and it is clear that his conduct is non-responsible. If he were brought to trial for his attack, he would have a valid defense of insanity.[1]

 

Innocent Projectile: Someone picks up a third party and throws him at you down at the bottom of a deep well. The third party is innocent and a threat; had he chosen to launch himself at you in that trajectory he would be an aggressor. Even though the falling person would survive his fall onto you, you may use your ray gun to disintegrate the falling body before it crushes and kills you.[2]

 

It is worth noting that these cases differ from other cases that target innocents. Consider the following. 

 

Killing Children: A terrorist has planted a bomb in an apartment complex in Israel, but won’t say which one. An Israeli officer who knows of the danger but can’t locate the bomb brings in the terrorist’s four children. He shoots one and threatens to shoot one for every subsequent ten-minute period that the terrorist holds out.

 

This defensive act, and ones like it, is clearly unjust because the children are not threats.

 

  1. The Right of Self-Defense is Explained by Rights Forfeiture

      The explanation of the right to self-defense is that the aggressor forfeits various moral rights that ordinarily protect him against harm. This forfeiture explains four features of the right to self-defense. First, it explains the content of the right to self-defense (in particular, the defender’s permission to touch or damage an aggressor’s body or property). These are rights that are forfeited by the unjust aggressor. Without such forfeiture, the defender’s actions would be unjust and would ground a claim to compensation on behalf of the aggressor. Neither intuitively seems to be correct, even in cases where the aggressor is innocent. Second, it explains why the aggressor may not have a justice-based claim against defensive action but third parties might. That is, if certain parties (e.g., the government) gain the right to any and all defensive action, then they might have a claim against a person acting in self-defense. Third, the forfeiture notion explains why a likely victim has a right to take defensive action but that this right by itself doesn’t entail that he ought to do so. The idea here is that rights protect act-options, but don’t guide the choice between the various options.[3] Fourth, the forfeiture account fits nicely with the notion of rights as forming a perimeter of non-interference again to create space for the making of choices. Persons thereby have a right to protect this perimeter regardless of whether someone is morally responsible for crossing into it. 

If forfeiture grounds the right to self-defense, then the right is grounded by another’s act. That is, the right should be viewed as having the following form.

(1)               If person B … to A, then A has a right to self-defense against B.

This is distinct from a conditional right (e.g., person A has the right to use-force-against-B-if-B …). Proposition (1) better captures the notion that the right is created through B’s forfeiture. In particular, it better aligns with the creation of a new claim and liberty in A.[4] If, however, contrary to my claim, the right to self-defense is conditional, nothing in my overall argument changes. 

      Forfeiture theory has come under a number of criticisms. First, it doesn’t explain why the aggressor forfeits his rights. Instead, the theory is circular in that it merely presupposes that self-defense is morally just rather than explaining it. This can be seen in the theory’s ad hoc explanation of the different constraints on just self-defense: necessity, proportionality, and imminence conditions on self-defense. This can also be seen in the theory’s inability to explain why persons can lose rights on the basis of something for which they are not morally responsible (e.g., consider Innocent Projectile). Second, the theory is so vague as to be useless as a guide in deciding when defensive force is just. In particular, the theory is unable to specify the rights that an attacker forfeits, the degree to which he forfeits them, and the duration of forfeiture. For example, does a rapist forfeit his right not to be raped, equivalent rights, or his right to life? [5] Both of these criticisms can be overcome if forfeiture is seen as a primitive and relatively fundamental feature of a person’s rights over his body and property.[6] If this is correct, then the right’s content can be discovered via our pattern of intuitions. It would then be argued that our intuitions are determinate enough to support a clear and coherent pattern that allows forfeiture to explain when self-defense is just, why forfeiture need not presuppose moral responsibility, and when and how it occurs. This is particularly true when rights are viewed as having a particular role, such as protecting or promoting persons’ autonomy or interests.

The likely victim alone has the right to self-defense. If the right to self-defense is intimately connected to a person’s rights over his body and property, then it appears that he exclusively has the right to self-defense. The antecedent rests on the notion that it is a person’s right over her body or property that makes the attack unjust and it is presumably this injustice (or the harm that will likely accompany it) that explains the right forfeiture.

In addition, it seems coherent for a likely victim to defend herself against attack while at the same time denying others the right to act on her behalf. This is analogous to a case in which a person might demand that a promise be fulfilled, yet deny third parties the right to help her enforce it.

This forfeiture account also explains the moral permission to attack various bystanders.

Culpable Bystander: A mineshaft has collapsed, leaving two miners trapped in a small open space. A radio communication has informed them that rescuers will reach them in five hours, but their instruments indicate that, while there is enough oxygen to allow one of them to survive for more than five hours, their oxygen supply will be depleted within three hours if both continue to breathe. … [O]ne of the minors (call him the “innocent miner”) then learns two facts: first, that the other miner (the culpable miner”) deliberately engineered the collapse of the shaft in an effort to kill him and, second, that the culpable miner has a small oxygen tank that will allow him to survive for two hours after the shaft runs out. The innocent miner therefore has two options: he can do nothing and die of asphyxiation while the culpable miner survives or he can take the culpable miner by force, thereby killing the culpable miner in self-preservation.[7]

Intuitively, the innocent miner may justly kill the culpable miner to take his small oxygen tank. This is true regardless of whether the culpable miner is viewed as a current threat since he caused the complex state that endangers the other miner’s life or whether the culpable miner is viewed as a past cause of the current threat. The forfeiture account explains this because it explains why the culpable wrongdoer is neither unjustly treated nor owed (nor his estate owed) compensation for the killing. If the miner with the small tank were not a cause, and perhaps not an unjust cause, of the situation, then it is not clear that the innocent miner would be justly permitted to attack him.

      If the forfeiture explanation is correct, then the right to self-defense is an objective rather than subjective one in that it focuses on whether a person actually faces a threat, not whether he believes that he faces such a threat. This objective interpretation allows us to distinguish, as we ordinarily do, between just and unjust self-defense and whether the agent is blameworthy for the defensive violence. It also avoids the oddness of one person gaining a right against another merely because the latter appears to be something. This is similar to the way in which in other areas of morality, merely appearing one way doesn’t ground a Hohfeldian moral claim or liberty. For example, merely appearing to be dead or to promise to do something doesn’t ground a right in another.

 

  1. Limitations on the Right of Self-Defense

      The right to self-defense is often thought to include various conditions, namely that the force used be (a) necessary to avoid (b) an imminent threat and (c) that the harm caused be proportional to that avoided. I will argue that the proportionality requirement is the only one and that it applies in an odd manner. There is no necessity requirement for self-defense. To see this, imagine that in Innocent Projectile, there is a 5% rather than 0% chance that the projectile will miss the person at the bottom of the well due to a small chance that the projectile will be sufficiently dismembered by the sharp edges on the side of the well. The fact that killing the innocent projectile is not necessary for the defender to save his life does not undermine his right to use lethal force. Nor would this requirement allow for clearly just cases of self-defense when the defender stands very little chance of winning.

      Imminence, understood as temporal distance between the attacker’s act and the harm, is also irrelevant. Once we control for the likelihood and severity of harm, temporal distance is no more relevant in legitimating defensive force than is physical distance. For example, consider a case in which terrorist is activating a series of well-hidden and undetectable bombs. So long as the likelihood and severity of the threatened harm is constant, the time until the bomb explodes and the harm occurs, e.g., 2 hours versus 5 years, is irrelevant to the right to use force to prevent the bombs from being activated. There might be consequentialist reasons for making the three conditions part of the law, e.g., to discourage people from defending themselves except in the clearest of cases, but these are not relevant to the right to self-defense.

Proportionality, if it is relevant at all, does not hold in a straightforward manner. Consider the following case.

 

Hell’s Angels: Six members of a Hell’s Angels’ chapter are about to rape a seventeen-year-old girl. The only way for her to protect herself is to shoot them with a machine gun she has taken off one of their choppers. Given the close range and the gun’s inaccuracy and power, her using it will likely kill most, if not all, of the six. 

 

Such defensive force seems permissible even though death is worse than rape and even though six persons will die. Even if proportionality is viewed in terms of goodness rather than net harm and even if the Angels’ well-being is discounted or even has negative value in virtue of their vicious nature, this still does not support the proportionality requirement.[8] To see this, consider the following case.

 

Blood Bank: A soldier has a rare type of blood that has been and will be continue to be used to save the lives of fellow soldiers who desperately need blood transfusions. In order to protect his life and ensure that he does not engage in any sexual activity that will contaminate his blood, the army orders him stationed in an administrative post in a hospital. There he tries to rape a nurse. The only way she has to protect herself is to inject him with a needle that contains so much morphine that it will kill him. 

 

Again, it intuitively seems that this use of defensive force is just even though doing so brings about overall worse results.[9] There still does seem to be something like a proportionality requirement. For example, it intuitively seems unjust to shoot someone in order to prevent him from stealing a candy bar. However, the limits of defensive violence intuitively are unclear. This can be seen, for example, in that the disproportionate results are far worse in Blood Bank than in the candy-bar-stealing case. Nevertheless, we can adopt the following sufficient condition for proportionality: an act of self-defense is proportionate if it brings about overall good results and better results than any other act or omission available to the agent.[10]

             

  1. Conclusion

      If this analysis is correct, then we should accept the following claims.

(2)               A person has a right to self-defense against, and only against, unjust threats to his well-being.

(3)               A threat is a cause of a process that will likely harm someone.

(4)               The explanation of the right to self-defense is that the threat (and perhaps agent of omission) forfeits some of his moral rights.

(5)               The right to self-defense is not subject to necessity and imminence constraints.

(6)               An act of self-defense is proportionate if it brings about overall good results and better results than any other act available to the agent.

 

  1. Just Wartime Killing Focuses on Whether a Person is a Threat

A right is a claim held by one individual against another. This is because they are designed to protect autonomy or interests against others, depending on the theory of rights, and only individuals have those features. Rights can be gained, lost, or modified. One way in which they can be modified is if they are combined. They can be combined when one person gains a right against another with regard to how the second exercises his right. An example of this occurs in the context of a partnership, where each partner has a right to use the firm’s assets only if he gets the permission of a majority of his partners. No matter how rights are gained, lost, or modified, they are still held by individuals.

If this is correct, then an individual can lose a right only if he does something (e.g., attacks another), changes in some relevant way (e.g., ceases to be autonomous because of Alzheimer’s disease), or there is a change in the things they own (e.g., his water holes dry up). In the context of potential military targets, soldiers and other relevant persons don’t change in some relevant way, nor are there relevant changes in the things they own. Hence, if they lose rights, they do so by doing something, specifically by waiving or forfeiting them.

Waiver occurs when a person validly consents or promises to give up a right. For example, a person waives his right to his car when he sells it. However, in most cases, soldiers and others do not waive their rights to be killed during wartime. Hence, if soldiers and others lose their right to be killed, they usually do so by forfeiting them. Other than by being a threat to another’s person or property, there doesn’t seem to be other ways for them to do this. This is true regardless of how rights have been combined into the collections that constitute nations. Hence, just wartime killing focuses on whether a person is a threat. As a result, we’ll need an account of threats that will allow us to identify who may be properly disabled or killed and it is to this we now turn.  

 

PART TWO: THE NATURE OF UNJUST THREATS

  1. Threats

      Threats are posed by events involving objects. For simplicity, I will talk as if objects alone pose threats. A threat is something that causes or enables a process that makes it more likely that a person’s life will go worse.[11] Not all injuries are produced via cause. A person’s act enables harm when it brings about a state that makes it more likely that a person will be harmed, but does not cause the harm.[12] For example, an attacker might remove a warning sign to a washed out bridge. The removal doesn’t cause the resulting crash since the absence of a sign isn’t an event. Similarly, if free human thoughts or actions are ones that are not caused by previous events, then an officer’s commands enable but do not cause his platoon to attack. If, for simplicity, we ignore the role of enablers, then a threat is a cause of a process that makes it more likely that a person will be harmed. For example, bears, assassins, viruses, and neighbors who covet one’s wife are all threats.  

      In discussing a threat’s effects, there are many categories that are relevant. Discussions of effects distinguish between intended and unintended effects, direct and indirect effects, gross and net effects, anticipated (e.g., expected) and actual effects, and particular effects and types of effects. In addition, a threatening act can also have multiple parts. For example, where a pharmacist promises to sell a person a blood thinner and instead gives him a sugar pill, the former might be considered to have endangered the latter even though the endangerment is comprised of a number of component acts.   

 

B.     Unjust Threats

A threat is unjust if and only if the person performing it infringes on someone’s rights. Not all threats are unjust. For example, a suitor who tries to win the hand of a man’s fiancée is a severe threat, but not an unjust one. Whether a mere attempt infringes on a second person’s moral right depends on whether persons have moral rights against having certain risks imposed. For example, if one person plays Russian roulette on a second when the second one is sleeping, the first might be seen to infringe on the second’s moral right.[13] I shall assume that persons have rights not be exposed to certain risks.

      Defensive violence is justified only if it is in response to an unjust threat. We are now in a position to set out the criteria for an unjust threat.

(7)               A person, X, poses an unjust threat to a second, Y, if and only if X’s act infringes on Y’s moral right and brings about a process that makes it more likely that Y is harmed.

Not all unjust threatens warrant killing or grievous bodily harm. Hence, we need to identify the magnitude of threats.

 

C.     The Magnitude of Unjust Threats

Any amount of an unjust threat grounds defensive action, although this action is constrained by the magnitude of the threat. One may not, for example, shoot a shoplifter or trespasser. What we need is a way to assess the magnitude of an unjust threat as a way of determining when a lethal response is just.

The magnitude is a function of some bad result of the attacker’s act, a result that would have occurred had the defender not responded to it. This bad result might be thought to be a function of the harm, injustice, or moral badness that the attacker would have brought about (or would have likely brought about) but for the defensive response. The first approach holds that the magnitude of the unjust threat is a function of the harm that the attack would have brought about. It runs into some troubles with aggression that has indirect and unintended benefits. Consider the following case. 

 

Gambler: Louie is a degenerate gambler who borrows money from the Gambino family and then fails to pay back all the interest (although he does pay back the principal plus 20%). As a result, a Gambino soldier attempts to force him to swallow a chemical stolen from the local pharmacy. Louie stabs the soldier in self-defense. The poisoning would have made Louie very ill but also would have killed his cancer, thereby extending his life by a couple of decades. 

 

The stabbing intuitively seems to be just even though it would not have brought about net harm. It might be that this harm-based approach is capable of being salvaged by focusing on the direct or intended harms. A direct effect is usually viewed in one that has an immediate causal nexus, a foreseeable connection, or one for which the agent is morally responsible. An intentional effect is one that a person sees as his goal or the means to his goal. It is not clear that this will work for aggressors who neither form the relevant intentions, nor are morally responsible for their aggression (e.g., Innocent Projectile). The assumption behind this last point is that the causal-nexus and foreseeable-effect versions of the directness theory don’t work. 

A second approach holds that the magnitude of the threat is a function of the stringency of the right that the soldier would have infringed on. In general, the first two approaches will likely align since the stringency of a right tends to correlate with the importance of the interest that it protects. This approach, however, is not burdened by concerns over indirect and unintended benefits.

A third approach holds that the magnitude of a threat as a function of how much worse it would have made the world. This approach should be rejected, since it suggests that the threat is a minor one if Louie is a vicious person and hence one whose death would have made the world a better place.

This second approach raises a few concerns. First, it puts enormous weight on moral rights. However, if justice (i.e., the perimeter of moral rights) sets the boundaries on spheres of non-interference and if is independent of other types of moral considerations, such as utility, then it will already have to bear much of this weight. Second, it requires that we be able to compare degrees of injustice. I would suggest that this shouldn’t pose that a problem since we already do this in setting punishment ceilings. For example, backward-looking justifications of punishment, assuming they address the issue of how much to punish, require that we assess past wrongdoings in terms of their seriousness and then assign punishment ceilings according to a shared metric. The seriousness of a wrongdoing might be filled out in terms of the harm that a particular injustice caused or typically caused, the evilness of the intention or motive that brought about the act or that typically did so, the importance of the right infringed or typically infringed by the act, or something else. An analogous approach can be used in this context so long as it is consistent with the possibility of the aggressor’s being blameless. A third concern consists of measuring contribution when there are different persons contributing to the unjust threat. Here the best approach in at least some contexts is to use the notion of marginal contribution to an injustice that is similar to one used to measure the economic value of a factor of production. 

The next problem that arises is the degree of unjust threat that warrants a killing. For my purposes we can rely on an intuitively satisfying sufficient condition for just defensive killing. If we rely on Hell’s Angels and Blood Bank, then we should conclude that when one person poses a significant unjust threat of a rape, he may justly be killed. A similar thing is true of those pose an unjust threat of death or severe bodily injury. By focusing on sufficient conditions for lethal defensive force, we are able to sidestep issues about the degree of injustice necessary for it.

If the above arguments succeed, then we should adopt the following conclusions.

(8)               The degree of an unjust threat is equal to its expected degree of injustice.

(9)               The expected degree of injustice is equal to the stringency of the right that the unjust aggressor would have infringed on but for the defensive action. 

(10)           If an unjust threat is at least as serious as a significant chance of rape, then, as a matter of justice, it warrants a lethal response.  

 

  1. Threats and Coordination

      Where there are multiple agents and they act in concert to kill someone, the right of self-defense would seem to generate forfeiture against any member of the group rather than merely against the persons carrying out the last part of the attack.[14] Since the group functions as an individual, they each forfeit the rights that would be forfeited by an individual who performed all the actions of the group. For example, since an individual who tried to kill an innocent would forfeit his right to life, the members of the group that plan and attempt to do the same also forfeit this right, at least with regard to defensive action. The explanation for this is that each member is a sustaining cause of a lethal threat and hence forfeits his right to life. This is true even where one conspirator’s contribution is neither necessary nor sufficient for the type of lethal unjust threat posed (i.e., there is causal overdetermination of the unjust threat).[15]

      This coordination requirement prevents forfeiture from applying to many independent and innocent contributors to an unjust threat. Consider the following case.

 

            The Diabetic’s Mother: A homicidal diabetic is chasing Mary through the woods

of an enclosed preserve, trying to kill her for sport with a pistol. However, because of his medical condition, he must return to a cabin in the middle of the preserve every hour in order that his aged mother can give him an insulin shot. Without it, he will take ill and die and will thus be forced to abandon his attempt to kill her. Blocking that insulin shot seems Mary’s only hope and the only reliable way to do this is for her to kill the mother.[16]

 

Here intuitions differ as to whether Mary may kill the mother.[17] This may depend on the degree of threat and how one determines it for different causal contributions to an aggressor. It intuitively seems that Mary may not kill the person who delivers food to them or the utility that provides electricity to the cabin thereby allowing for the mother to inject the insulin. However, it also seems that Mary may do so if all these persons conspired with a sane diabetic to kill her and are doing their part. It might seem that this is explained by the moral guilt of the parties, but this isn’t correct since they may be killed even if they are not responsible (e.g., they made a reasonable error about Mary’s identity or are mentally ill). It is unclear whether the relation of forfeiture to joint projects is a result of the participants causing the threat, intentionally causing the threat, being part of a threat-like project, or something else.

 

  1. Conclusion

We thus end up with the following conclusions.

(11)           The degree of an unjust threat is equal to its expected degree of injustice. 

(12)           The expected degree of injustice is equal to the stringency of the right that the unjust aggressor would have infringed on but for the defensive action.

(13)           If an unjust threat is at least as serious as a significant chance of rape, then, as a matter of justice, it warrants a lethal response.

(14)           If a number of persons jointly pose an unjust threat, then each one may, as a matter of justice, be treated as if he alone poses it. 

Using these notions, many different types of persons might end up being legitimate wartime targets.

 

PART THREE: WHO IS A THREAT DURING WARTIME

When applied to persons in an aggressor country, the application is straightforward. When a person’s action threatens as much injustice as a rape, the likely victim or his agent may use lethal force. Whether a person is a combatant or blameworthy is irrelevant. For an unjust aggressor, its military leaders are clearly legitimate targets, assuming they jointly bring about or increase a sufficient unjust threat. So are civilian leaders who do so. The fact that a leader is a legislator no more protects him against defensive violence than it protects drug dealers who democratically voted to pay for a hit on a new salesperson in their city. The fact that any one legislator’s vote is not necessary for his country’s unjust war is similar to other cases of overdetermination of an unjust threat. The same is true with regard to medics, farmers, and production workers and other persons whose marginal contribution generates an enough of an unjust threat, particularly if they are judged part of a joint project.

 

PART FOUR: CONCLUSION

In this paper, I assumed that just wartime killing is justified by self-defense and defense of others. Self-defense focuses on unjust threats. One person is an unjust threat to another if the first does something that infringes on the second’s moral rights and thereby makes it more likely that the second is harmed. I then noted that threats come in degrees. I argued that if a person poses as much unjust threat as one who imposes a significant chance of rape on someone, then that person may, as a matter of justice, be killed. I speculated in a country that is engaged in an unjust aggressive war, this might make many different people legitimate targets.

 

NOTES



[1] This example comes from George P. Fletcher, “Proportionality and the Psychotic Aggressor: A Vignette in Comparative Criminal Theory,” Israel Law Review 8 (1973): 367-390. A similar example occurs in Sanford H. Kadish, “Respect for Life and Regard for Rights in the Criminal Law,” California Law Review 64 (1976): 889.

[2] This example comes from Robert Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974), 34,

[3] Jeremy Waldron argues for this notion in Jeremy Waldron, “A Right to do Wrong.” Ethics
92 (1981): 21-39.

[4] In addition, the logic of a conditional right is not obvious. If we conjoin the statements (a) person A has the right to use-force-against-B-if-B … and (b) B …, it is not clear what follows. This is because the conditional is contained within the scope of the operator “has a right to” and this makes the logic less transparent than if rights are forfeited and corresponding ones created. The idea for this point comes from Jeremy Waldron, The Right to Private Property (Oxford: Clarendon Press, 1988), 118-1222.

[5] A nice summary of these objections occurs in Whitley Kaufman, “Is There a “Right” to Self-Defense?” Criminal Justice Ethics 23 (2004): 20-31, esp. 23-26.

[6] An example of his strategy can be seen in Stephen Kershnar, “The Structure of Rights Forfeiture in the Context of Culpable Wrongdoing,” Philosophia, 29 (2002): 57-88.

[7] This example comes from Jeff McMahan, “Self-Defense and the Problem of the Innocent Attacker,” Ethics 104 (1004): 258. It should be noted that McMahan views the culpable miner as a bystander rather than attacker, which I leave open, and rejects the right-based analysis of permissible self-defense. Ibid., 258 and 275-281. On his account, it is moral responsibility for initiating or sustaining a threat (or in some cases, for failing to eliminate it) that makes a person liable to defensive violence. McMahan, Jeff McMahan, “The Ethics of Killing in War,” Ethics 114 (2004): 718-722, esp. 721.

[8] The notion that the well-being of persons with negative desert has negative intrinsic value is defended in Fred Feldman, “Adjusting Utility for Justice: A Consequentialist Reply to the Objection from Justice,” Philosophy and Phenomenological Research 55 (1995): 567-585; Thomas Hurka, “The Common Structure of Virtue and Desert,” Ethics 112 (2001): 6-31.

[9] Thomas Hurka notes that one some accounts, the costs of defensive violence, at least in the context of war, exclude other agents’ wrongful choices, even when they result from the agent’s defensive actions. Thomas Hurka, “Proportionality in the Morality of War,” Philosophy & Public Affairs 33 (2005): 34-66. Even if this is correct, and this is not obvious, it is irrelevant to cases like Blood Bank.

[10] The intent condition is likely superfluous. I include it since on some accounts some benefits of wartime killing (e.g., economic growth) do not count toward the proportionality condition because they are not contained in the just aims of the just side. Hurka, “Proportionality in the Morality of War.”

[11] More specifically, since causes are events, a threat is an event that that involves a thing. However, for simplicity, I will talks as if things, such as persons, are threats.

[12] The notion of enabling comes from Lawrence Lombard, “Delaying, Preventing, and Disenabling,” Philosophia 24 (1995): 435. A harm is a setback to an interest. Joel Feinberg, Harm to Others (New York: Oxford, 1984), ch. 1.

[13] This example comes from Judith Jarvis Thomson, The Realm of Rights (Cambridge: Harvard University Press, 1990), 244.  

[14] The key feature here appears to be shared intent. Michael Bratman sets out the shared intent in his “Shared Intention,” Ethics 104 (1993): 97-113, esp. 106. On Bratman’s account, a shared intention occurs where two persons share the same goal, where their sharing the same goal rests in part on the mutual recognition that they share the same goal and plans by which to achieve the goal.

[15] Here the member’s contribution might be cause of the particular threat even if it isn’t the cause of the type of threat. The idea here is that a threatening event is a particular event and that the causes of an event are all necessary conditions of it. The notion that a cause is an essential feature of a particular event is seen in Peter van Inwagen, “Ability and Responsibility,” in John Martin Fischer, ed., Moral Responsibility (Ithaca: Cornell University Press, 1986), 158-160.

[16] This example, slightly modified, comes from Jeffrie Murphy, “The Killing of the Innocent,” in Retribution, Justice, and Therapy (Boston: D. Reidel Publishing Co., 1979), p. 22 n. 15.

[17] For those who think that Mary can’t kill the mother, it’s worth considering whether it matters whether the mother repeatedly fixes the diabetic’s gun after it jams and reloads it for him (she thinks he’s shooting rats at the local farm) or shoots him with insulin. It’s hard to see why this should matter and yet the former is analogous to persons loading howitzers in the military, i.e., paradigm legitimate targets.