¹ The arguments below deal with uncertainty about the situations in which we find ourselves, not uncertainty about what is the correct moral theory.

² Throughout, I will assume that the force used is proportionate. Thus, the liability we discuss is only liability to proportionate defensive force. I set aside other issues, such as whether the force is necessary, or what it exactly entails.

³ Throughout, when discussing moral permissibility, “ought,” and other similar terms, I will insert the phrase “other things equal” to indicate that I am talking about these terms to the extent that they track the presence or absence of rights.

⁴ See e.g., Hurd, Heidi, Moral Combat (Cambridge: Cambridge University Press, 1999), 7.CrossRefGoogle Scholar

⁵ See Zimmerman, Michael, Living with Uncertainty (Cambridge: Cambridge University Press, 2008)CrossRefGoogle Scholar and Ignorance and Moral Obligation (New York: Oxford University Press, 2014). Zimmerman discusses self-defense in Living with Uncertainty, 108ff. There are two differences between Zimmerman’s discussion and my own. First, Zimmerman offers no substantive theory of liability, but uses the case of self-defense to test his broader view. The final, rights-based part of the argument below thus serves to complement his discussion. Second, I disagree with Zimmerman on two counts. I find his arguments against objectivist theories of liability unsatisfactory because he claims that such theories cannot explain why aggressors become liable before the actual attack has occurred. This seems false. Moreover, Zimmerman says that if A threatens V, and A knows this, but V does not know this, then A nevertheless becomes liable to V’s defensive force. (See Zimmerman, Living with Uncertainty, 108–10.) My account does not support this conclusion, as, it seems to me, is the view’s natural implication. Zimmerman claims (p. 109) that V would still be wrong to attack A, but for some other reason than A’s rights. This strikes me as ad hoc. What is wrong about V’s attack if not how it treats A?

⁶ The example is a variation on a case presented by McMahan, Jeff, “The Basis of Moral Liability To Defensive Killing,” Philosophical Issues 15 (2005): 386–405, at 387.CrossRefGoogle Scholar Quong, Jonathan, “Liability to Defensive Harm,” Philosophy and Public Affairs 40 (2012): 45–77 CrossRefGoogle Scholar also discusses a variation of this example (p. 53).

⁷ Compare Rodin’s discussion about justifications and excuses in Rodin, David, War and Self-Defense (New York: Oxford University Press, 2002), 30–31.CrossRefGoogle Scholar

⁸ I draw the notion of a “best bet” from Zimmerman. For more detailed discussion, see Living with Uncertainty and Ignorance and Moral Obligation, throughout.

⁹ Throughout I assume that it is not Annie’s fault that the evidence is faulty.

¹⁰ Similarly, X’s best bet about Y’s liability cannot be simply understood as X being justified in believing that Y is liable. While the account proposed below refers to some of X’s justified beliefs, it does not follow that X’s best bet can be entirely understood in this way. See note 29 below.

¹¹ For related points, see Frowe, Helen, “A Practical Account of Self-Defence,” Law and Philosophy 29 (2010): 245–72;Google Scholar Kessler Ferzan, Kimberly, “Justifying Self-Defense,” Law and Philosophy 24 (2005): 711–49;Google Scholar and Kessler Ferzan, Kimberly, “Culpable Aggression: The Basis for Moral Liability and Defensive Killing,” Ohio State Journal of Criminal Law 9 (2012): 669–97.Google Scholar

¹² See Quong, “Liability to Defensive Harm,” 67–8. Quong introduces the evidence-relative qualification to deal with certain cases of risk-imposition. Since I do not see an important difference between the proposal below and that part of his account, I leave this issue aside. In what follows, I consider the theory’s suitability for dealing with problems of uncertainty, which may not have been Quong’s purpose.

¹³ Quong, “Liability for Defensive Harm,” 69ff.

¹⁴ Frowe, “A Practical Account of Self-Defence,” 250. She calls this the “myth of ‘full and accurate’ knowledge.”

¹⁵ Frowe, “A Practical Account of Self-Defence,” 260.

¹⁶ More precisely: [I]f defensive force is ever permissible, its use must be justified on the grounds of Victim’s reasonable belief that (a) if he does not kill this person, then they will kill him, and (b) that he is innocent” (Frowe, “A Practical Account of Self-Defence,” 252).

¹⁷ “[W]hether or not Victim is justified in inflicting harm is not the same question as whether his target is liable to bear that harm. What renders a person liable to bear harm are objective facts about culpability” (Frowe, “A Practical Account of Self-Defence,” 252–53).

¹⁸ Frowe’s account aims to improve upon Kimberly Ferzan’s earlier account in Ferzan, “Justifying Self-Defense.” Ferzan was one of the first to draw attention to the problem of uncertainty in the context of liability to defensive harm. Ferzan correctly points out that a theory of liability that presupposes that defenders know things that they could never know (as most “objective” theories do) is useless or worse. However, Ferzan’s own account insists that epistemic considerations only come into play once an objective condition of culpability on the part of the liable person is met. But of course the very same objection can be raised against this objective condition as Ferzan raises to objective theories more generally. For elaborations of this objection, see Frowe, “A Practical Account of Self-Defence”; and McMahan, Jeff, “Self-Defense and Culpability,” Law and Philosophy 24 (2005): 751–74.CrossRefGoogle Scholar

¹⁹ In fact, Frowe does not even think that we should base permissibility on evidence in general. The context of self-defense is special, she writes, because self-defense “is by its nature urgent: it does not allow for the deliberation or investigation that we ought to require in other parts of morality” (Frowe, “A Practical Account of Self-Defence,”257). If someone points a gun at you, you have to make a snap decision about whether or not it is permissible to kill her in self-defense. This, too, is unconvincing. Decisions about defense are hardly unique in being time-sensitive. And decisions about defensive harm can allow for extensive deliberation. Real-world examples include preemptive and, especially, preventive wars. Given that there is no clear distinction between decisions about defensive force and other parts of life, the cost of postulating different sets of moral rules for these is considerable. For one, it would make it hard to see when rival (inconsistent) sets of moral rules apply, and thus what we ought to do. It would also make it hard to understand why one set of rules applies rather than another, or to what reasons our actions ought to be sensitive, and thus to solve uncertainty or disagreement about right action.

²⁰ Perhaps I should say that I find these rationales in Locke’s discussion. I leave aside whether Locke would have endorsed these. My aim here is not Locke scholarship; it is an attempt at drawing insight from his text.

²¹ Locke, John, Two Treatises of Government, ed. Laslett, P. (Cambridge: Cambridge University Press, 1988 [1689]), Second Treatise, sect. 7.CrossRefGoogle Scholar

²² Acting in a rights-violating manner entails acting without moral justification. Violating a right is different from infringing a right. The latter are normally in conflict with the demands of other people’s rights, but are nevertheless done with moral justification.

²³ As Locke put it: “In transgressing the Law of Nature, the Offender declares himself to live by another Rule than that of reason and common Equity, which is that measure God has set to the actions of Men, for their mutual security; and so he becomes dangerous to Mankind, the tye, which is to secure them from injury and violence, being slighted and broken by him. Which being a trespass against the whole of Nature, every man upon this score, by the Right he hath to preserve Mankind in general, may restrain, or where it is necessary, destroy things noxious to them, and so may bring such evil on any one, who hath transgressed that Law, as may make him repent the doing of it, and thereby deter him, and by his Example others, from doing the like mischief. And in this case, and upon this ground, every Man hath a Right to punish the Offender, and be the Executioner of the Law of Nature” (Second Treatise, sect. 8).

²⁴ Thus, should you survive and be injured I ought to pay you compensation for the rights-violation. But it would not make sense for you to blame me for injuring you, as it was an accident. (It would make sense to blame me for my negligence.)

²⁵ He may, of course, still violate some other right of yours, such as the right not to expose you to the danger of being shot.

²⁶ As required, for example, by Quong’s theory.

²⁷ One might object that after the hit man repairs his gun, he is again a threat to the target and thereby again becomes liable. But this is mistaken. The hit man repaired his gun after his target became liable, and thus does not pose a threat to a non-liable person.

²⁸ The idea of being justified in believing something is complex. For present purposes, I will consider what X is justified in believing and X’s available evidence as more or less interchangeable. That is, I will consider X’s being justified in believing something in terms of X having sufficient evidence concerning that thing. However, this is ambiguous between the evidence that is in principle available to X and the evidence of which X has in fact availed himself. I think the correct account is organized around the latter for reasons outlined by Zimmerman in chapter 3 of Ignorance and Moral Obligation. But the account of liability does not turn on this, as it is a general concern about how to treat the evidence-relative dimension of morality. For this reason, I sideline these complicated issues.

²⁹ I do not claim, for instance, that X’s best bet is also made up of X being justified in believing that force is necessary, proportionate, and so on.

³⁰ Quong discusses a slightly different case because he considers a slightly different evidence-based view from the one I have proposed here. Nevertheless, his discussion contains the basic building blocks of this (for my account more threatening) objection. See Quong, “Liability to Defensive Harm,” 62.

³¹ The same is true, of course, for relevantly informed third parties. With respect to them, too, the duped soldiers are liable to defensive force. Note that, in these respects, the case of the duped soldiers is equivalent to Mistaken Attacker.

³² For a variation of this objection, see Ferzan, “Justifying Self-Defense,” 713 and 729.