¹ I assume throughout that liability is a central element of permissible defensive killing, but this means that it is neither a necessary nor a sufficient condition for such permissibility.

² I assume here the standard Hohfeldian distinction between, liberties, claims, powers and immunities. See Hohfeld, Wesley, Fundamental Legal Conceptions (New Haven, 1919)Google Scholar.

³ Thomson, J. J., ‘Self-Defence’, Philosophy and Public Affairs 20.4 (1991), pp. 283–310 Google Scholar, at 303. For a similar position, see Uniacke, S., Permissible Killing: The Self-Defence Justification of Homicide (Cambridge, 1994)CrossRefGoogle Scholar, chs. 5 and 6.

⁴ Thomson, ‘Self-Defence’, p. 285.

⁵ McMahan, J., Killing in War (Oxford, 2009), p. 157 CrossRefGoogle Scholar.

⁶ McMahan, Killing in War, p. 35.

⁷ McMahan, Killing in War, p. 35.

⁸ Quong, J., ‘Killing in Self-Defense’, Ethics 119.2 (2009), pp. 507–37CrossRefGoogle Scholar, at 516–17. For similar positions, see Fabre, C., Cosmopolitan War (Oxford, 2012), p. 59 CrossRefGoogle Scholar, and Benbaji, Y., ‘A Defense of the Traditional War Convention’, Ethics 118.3 (2008), pp. 464–95CrossRefGoogle Scholar, at 474.

⁹ Incidentally, this also entails that it would be impermissible for third parties to intervene on behalf of any of the individuals involved, in line with the moral difference between doing and allowing. Namely, it is often accepted that killing an individual is morally worse than letting her die, all other things being equal. See Quong, J., ‘Liability to Defensive Harm’, Philosophy & Public Affairs 40.1 (2012), pp. 45–77 CrossRefGoogle Scholar, at 75.

¹⁰ Quong, ‘Liability to Defensive Harm’, p. 75.

¹¹ McMahan, ‘The Basis of Moral Liability to Defensive Killing’, p. 386. There are different understandings of liability in the literature. I use this one for ease of exposition. Furthermore, I use liability in this particular sense, and not as a correlative of a normative power, in Hohfeld's sense.

¹² As indicated in section I, this loss or forfeiture could be in personam or in rem.

¹³ See e.g. Fabre, Cosmopolitan War, p. 61.

¹⁴ For a similar criticism, see Benbaji, ‘A Defense of the Traditional War Convention’.

¹⁵ On rights’ infringement generally, see, e.g. Botterell, A., ‘In Defence of Infringement’, Law and Philosophy 27.3 (2008), pp. 269–92CrossRefGoogle Scholar. Against this notion, see Oberdiek, J., ‘Lost in Moral Space: On the Infringing/Violating Distinction and Its Place in the Theory of Rights’, Law and Philosophy 23 (2004), pp. 325–46Google Scholar.

¹⁶ Usually the notion of rights-infringement is associated with some form of compensation being in order. I take here a more cautious approach, in that it leads to a form of moral residue that sometimes can be compensated, but which can also be acknowledged in other ways.

¹⁷ Steinhoff proposes a similar analytical framework. However, he claims that in this type of situation both innocent parties have ‘lost their claim-rights to life and to self-defence but retained their liberty-rights to life and self-defence’ ( Steinhoff, U., ‘Rights, Liability and the Moral Equality of Combatants’, Journal of Ethics 16.4 (2012), pp. 339–66CrossRefGoogle Scholar, at 347, emphasis in the original). Accordingly, he does not see the relationship between the liberty and the loss, or better, the overriding of the relevant claim-right by each of these two individuals.

¹⁸ See e.g. McMahan, J., ‘Self-Defence and the Problem of the Innocent Attacker’, Ethics 104.2 (1994), pp. 252–90CrossRefGoogle Scholar, at 276, and Otsuka, M., ‘Killing the Innocent in Self-Defence’, Philosophy & Public Affairs 23.1 (1994), pp. 79–94 CrossRefGoogle Scholar.

¹⁹ We may reach a point in which for other considerations, most notably lesser evil ones, the number of culpable or responsible attackers is so big that, all things considered, it would be impermissible for V to kill them. This, however, is irrelevant for present purposes.

²⁰ For a similar case see Ferzan, K. Kessler, ‘Culpable Aggression: The Basis for Moral Liability to Defensive Killing’, Ohio State Journal of Criminal Law 9 (2012), pp. 669–97Google Scholar, at 675.

²¹ McMahan, J., ‘The Basis of Moral Liability to Defensive Killing’, Philosophical Issues 15 (2005), pp. 386–405 CrossRefGoogle Scholar, at 387–8.

²² McMahan, ‘The Basis of Moral Liability’, p. 394.

²³ McMahan, ‘The Basis of Moral Liability’, p. 394.

²⁴ Tadros, V., ‘Orwell's Battle with Brittain: Vicarious Liability for Unjust Aggression’, Philosophy and Public Affairs 42.1 (2014), pp. 42–77 CrossRefGoogle Scholar, at 72; Ferzan, K. Kessler, ‘Can't Sue; Can Kill’, Criminal Law Conversations, ed. Robinson, Paul H et al. (New York, 2009), pp. 308–99Google Scholar, at 399; Rodin, D., ‘Justifying Harm’, Ethics 122.1 (2014), pp. 74–110 CrossRefGoogle Scholar, at 85; and Bazargan, S., ‘Killing Minimally Responsible Threats’, Ethics 125.1 (2014), pp. 114–36CrossRefGoogle Scholar, at 123.

²⁵ On the differences between intuitions as an empirical matter, see the research reported in Greene, J., Moral Tribes (London, 2013)Google Scholar, ch. 3.

²⁶ This case is meant to address McMahan's Conscientious driver case as replicated in IT above. In order to make it more easily comparable to Tram, we could stipulate that the weight of the 50 passengers was necessary for the bus to be able to kill V.

²⁷ See e.g. McMahan, Killing in War, ch. 4.

²⁸ Nozick, R., Anarchy State and Utopia (Princeton, 1993), pp. 34–5Google Scholar, and Thomson, ‘Self-Defense’, p. 287.

²⁹ J. McMahan, ‘Liability, Proportionality, and the Number of Aggressors’, The Ethics of War, ed. S. Bazargan and S. Rickless (forthcoming).

³⁰ I am not sure he can make that move. If proportionality is generally internal to liability, proportionality in the wide sense must also be.

³¹ I assume here that McMahan would want to resist the claim that both CA and the 50 ITs are all liable to being killed, and that it is permissible for V to kill them all.

³² McMahan, ‘Liability, Proportionality, and the Number of Aggressors’.

³³ But see also, Steinhoff, ‘Rights, Liability and the Moral Equality of Combatants’.

³⁴ My argument here roughly follows Quong's argument in his ‘Killing in Self-Defence’, pp. 520–1. For other agent-relative justifications for the permission to act in self-defence, see Davies, N., ‘Abortion and Self-Defense’, Philosophy & Public Affairs 11.3 (1982), pp. 232–45Google Scholar, and Levine, S., ‘The Moral Permissibility of Killing a Material Aggressor in Self-Defense’, Philosophical Studies 45 (1984), pp. 69–78 CrossRefGoogle Scholar. Cécile Fabre has also defended a more differentiated account in her Cosmopolitan War, which is also based on the argument of partiality. More recently, Saba Bazargan has built a more symmetrical account based on lesser evil considerations (see his ‘Killing Minimally Responsible Threats’), and Burri, S. has argued for flipping a coin, in her ‘The Toss-Up Between a Profiting, Innocent Threat and His Victim’, The Journal of Political Philosophy 23.2 (2015), pp. 146–65CrossRefGoogle Scholar. I will not be able to address their arguments here.

³⁵ McMahan, J., ‘Self-Defense and the Problem of the Innocent Attacker’, Ethics 104 (1994), pp. 252–90CrossRefGoogle Scholar, at 268–71. See also, Otsuka, M., ‘Killing the Innocent in Self-Defence’, Philosophy and Public Affairs 23 (1994), pp. 74–94 CrossRefGoogle Scholar.

³⁶ Quinn, W., ‘Actions, Intentions and Consequences’, Philosophy and Public Affairs 18 (1989), pp. 334–51Google Scholar, at 344.

³⁷ On this move, see Quong, ‘Killing in Self-Defense’, pp. 525–32.

³⁸ McMahan, J., ‘Self-Defense against Justified Threateners’, How We Fight: Ethics in War, ed. Frowe, Helen and Lang, Gerald (Oxford, 2014), pp. 128–9Google Scholar.

³⁹ Quong has responded to this claim by suggesting that such a use of the bridge may be opportunistic, rather than eliminative (see his ‘Killing in Self-Defense’, pp. 525–32). I will not consider this further argument here.

⁴⁰ A caveat is in order here. As McMahan himself acknowledges (‘Self-Defense against Justified Threateners’, pp. 120–1), his objection does not necessarily undermine the claim that it is permissible for NRT to kill V. All the objection supports is that grounding a permission to (eliminatively) kill a non-liable threat on the existence of a personal prerogative would commit defenders of this prerogative to accepting a permission to kill at least some innocent bystanders. This proposition seems to me correct. Yet, it neither provides an argument for subjecting NRTs and Vs to his preferred ‘all-or-nothing’ framework, nor does it mean that agent-relative considerations must play no role in self-defence situations, as I shall argue below.

⁴¹ For two plausible defences of this approach, see Thomson, J. J., The Realm of Rights (Cambridge, MA, 1990)Google Scholar and Feinberg, J., Rights, Justice, and the Bounds of Liberty (Princeton, 1980)CrossRefGoogle Scholar. I have defended this understanding of conflicts of rights at greater length in Chehtman, A., The Philosophical Foundations of Extraterritorial Punishment (Oxford, 2010)CrossRefGoogle Scholar, ch. 2. For an alternative take on this issue, i.e. the ‘specificationist’ position on rights, see e.g. Oberdiek, J., ‘Specifying Rights Out of Necessity’, Oxford Journal of Legal Studies 28 (2008), pp. 127–46CrossRefGoogle Scholar.

⁴² See e.g. Feinberg, J., ‘Voluntary Euthanasia and the Inalienable Right to Life’, Philosophy and Public Affairs 7 (1978), pp. 93–123 Google ScholarPubMed, at 102.

⁴³ It differs from the standard version of NRT, among other things, in that B is not using any type of weapon. As argued by Joshua Greene, this type of consideration may introduce a bias in our intuitive assessment of the case. See Greene, Moral Tribes, ch. 3.

⁴⁴ See, e.g. Fabre, Cosmopolitan War, p. 57. Admittedly, some people may argue that A and B would be under a duty to flip a coin. See e.g. Burri, ‘The Toss-Up’. I don't think that morality is so demanding, but this makes no difference for the point at stake, which is precisely about the ultimately symmetrical position A and B are in.

⁴⁵ Raz, J., Ethics in the Public Domain: Essays in the Morality of Law and Politics (Oxford, 1994), p. 48 Google Scholar.

⁴⁶ The moral difference between killing someone opportunistically and eliminatively would have to be assessed in this same ‘dimension’ of the conflicts of rights framework.

⁴⁷ And yet this conclusion is in turn compatible with the claim that under certain circumstances it may be permissible for five individuals to cross the bridge, or for a 25-year-old youngster to do so if B were about to die in the next few hours or weeks. That is, their position may ultimately be symmetrical all-things-considered or even ultimately asymmetrical in the other direction. Nonetheless, no one needs to be considered liable to being killed in order to reach any of these conclusions.

⁴⁸ The same fundamental claim applies to NRT and NRA. If anything, their case is more symmetrical to Vs than to ITs.

⁴⁹ Bazargan, ‘Killing Minimally Responsible Threats’, p. 136. He, by contrast, is concerned with how the fact that just combatants would be infringing the rights of unjust combatants may alter the threshold for permissibly waging a war.

⁵⁰ Bazargan, ‘Killing Minimally Responsible Threats’, p. 164.

⁵¹ In fact, he suggests that Resident bears less responsibility than Conscientious driver, because he acts ‘not merely subjectively permissibly but with subjective moral justification’ (McMahan, Killing in War, p. 166).

⁵² Bazargan, ‘Killing Minimally Responsible Threats’, pp. 114–15.

⁵³ Quong presents us with a situation in which, much like Bazargan, Resident has opened the door and he lunges against Twin brother with a knife. Unlike Bazargan, however, he shares McMahan's conclusion that Resident is liable to being killed because he intentionally and unprovokedly attacks his twin brother (Quong, ‘Liability to Defensive Harm’, p. 53). Again, I believe his conclusion falls on the right side, but this is because Resident would still be acting somewhat negligently, something which Conscientious driver is not. Otsuka also supports liability here (see his ‘Killing the Innocent in Self-Defence’, p. 91).

⁵⁴ This does not mean that Resident should be convicted for murder. But I do not think a conviction for manslaughter should be discarded.

⁵⁵ McMahan, Killing in War, p. 166.

⁵⁶ Interestingly, this case further supports the conceptual distinction drawn in section III. Even if it is permissible for V to kill Twin Brother in Resident 2, TB is not liable to being killed (but neither is Resident).

⁵⁷ This conclusion, in turn, is compatible with certain unjust combatants being actually liable to being killed.

⁵⁸ I am grateful to Daniele Bruno, Alasdair Cochrane, Marcelo Ferrante, Massimo Renzo, Eduardo Rivera López (twice), and Uwe Steinhoff for written comments and conversations on earlier drafts of this article. I am also grateful for useful discussions to the participants in the Oxford Criminal Law Discussion Group, the War Seminar at Oxford University and the Faculty Seminar at Universidad Torcuato Di Tella, especially Janina Dill, Cécile Fabre, José Luis Martí, Jeff McMahan and Horacio Spector. The reviewers for Utilitas provided me with two extremely useful rounds of comments. The usual disclaimer applies.