I argue that riskier killings of innocent people are, other things equal, objectively worse than less risky killings. I ground these views in considerations of disrespect and security. Killing someone more riskily shows greater disrespect for him by more grievously undervaluing his standing and interests, and more seriously undermines his security by exposing a disposition to harm him across all counterfactual scenarios in which the probability of killing an innocent person is that high or less. I argue that the salient probabilities are the agent’s sincere, sane, subjective probabilities, and that this thesis is relevant whether your risk-taking pertains to the probability of killing a person or to the probability that the person you kill is not liable to be killed. I then defend the view’s relevance to intentional killing; show how it differs from an account of blameworthiness; and explain its significance for all-things-considered justification and justification under uncertainty.
See, for example, Jeff McMahan, Killing in War (Oxford: Oxford University Press, 2009). Although ‘innocent’ is a term of art, it is used advisedly. On my understanding, with one possible exception, the fact that one is innocent of responsibility for a wrongful threat is sufficient grounds for one not being liable to be killed in order to avert that threat. The exception: I suspect that some kinds of duress can render one’s contribution to a wrongful threat morally innocent, i.e. blameless, but that one can nonetheless be liable to be killed in such cases. It is important to be clear, though, that innocence here refers exclusively to one’s degree of responsibility for the wrongful threat that is to be averted. It does not invoke a more general property of one’s character.
Philip Pettit, The Robust Demands of the Good (Oxford: Oxford University Press, 2015); Philip Pettit, On the People’s Terms: A Republican Theory and Model of Democracy (Cambridge: Cambridge University Press, 2012). See also Nicholas Southwood, ‘Democracy as a Modally Demanding Value’, Nous, 49 (2015), pp. 504–521.
As I argue in Seth Lazar, ‘Deontological Decision Theory and Agent-Centred Options’, Ethics,(Forthcoming); Seth Lazar, ‘In Dubious Battle: Uncertainty and the Ethics of Killing’, this is at most as a sufficient condition for epistemic permissibility, not a necessary one.