¹ See e.g. Broome, J., ‘Goodness is Reducible to Betterness: The Evil of Death is the Value of Life’, The Good and the Economical, ed. Koslowski, P. and Shionoya, Y. (Berlin, 1993), p. 77Google Scholar; Bykvist, K., ‘The Benefits of Coming into Existence’, Philosophical Studies 135 (2007), pp. 335–62CrossRefGoogle Scholar; Parfit, D., Reasons and Persons (Oxford, 1984), p. 489Google Scholar.

² Holtug, N., ‘On the Value of Coming into Existence’, The Journal of Ethics 5 (2001), pp. 361–84CrossRefGoogle Scholar; Roberts, M., ‘Can it Ever Be Better Never to Have Existed At All? Person-Based Consequentialism and a New Repugnant Conclusion’, Journal of Applied Philosophy 20 (2003), pp. 159–85CrossRefGoogle Scholar. Roberts concentrates on the view that existence can be worse for a person than non-existence.

³ Note that I take ‘S's existence’ to denote S's mere existence, rather than, e.g., the state of affairs that S exists and leads a life with a particular content. See further Section IV.

⁴ Palle Yourgrau is an exception (by which I mean that he denies actualism, not that he succeeds in exemplifying properties or relations without being actual). See his ‘The Dead’, The Journal of Philosophy 86 (1987), pp. 84–101.

⁵ Parsons, J., ‘Axiological Actualism’, Australasian Journal of Philosophy 80 (2002), pp. 137–47, at p. 139Google Scholar. Cf. Bykvist, ‘Violations of Normative Invariance: Some Thoughts on Shifty Oughts’, Theoria 73 (2007), pp. 98–120, at p. 101n.: ‘One could . . . identify merely possible people with certain complexes of actually existing properties.’ If such complexes exist, this suggestion leads to the view that merely possible people exist.

⁶ ‘Not-A’ refers, of course, to the state of affairs that A does not obtain.

⁷ Roberts, ‘Can it Ever Be Better Never to Have Existed At All?’, p. 168.

⁸ Roberts, ‘Can it Ever Be Better Never to Have Existed At All?’, p. 177.

⁹ This is probably because it does not have any capacity for mental states (nor does, of course, Nora in w2). Cf. Section VI below.

¹⁰ To avoid confusion, I shall use the term ‘Jeremy's life’ instead of using ‘Jeremy's existence’ in Holtug's sense. I will come back to how Holtug's view is related to the question of whether Jeremy's mere existence is better for him than his non-existence.

¹¹ Holtug says that he got this way of respecting actualism from Wlodek Rabinowicz (in personal communication): Holtug, ‘On the Value of Coming into Existence’, p. 374n. In fact, when I discussed my ideas with Rabinowicz in May 2007, it emerged that what he had had in mind was closer to my own view (introduced in Section V below) than to Holtug's.

¹² Holtug, ‘On the Value of Coming into Existence’, p. 364.

¹³ Cf. Bykvist, ‘The Benefits of Coming into Existence’, p. 346.

¹⁴ Holtug, ‘On the Value of Coming into Existence’, p. 366.

¹⁵ Holtug, ‘On the Value of Coming into Existence’, p. 366.

¹⁶ Someone might object here that though Jeremy's life contains a surplus of positive intrinsic value, this does not imply that Jeremy's life itself has a positive intrinsic value (which it must have in order to be intrinsically better – have a higher intrinsic value – for Jeremy than his non-existence). But I won't press this point.

¹⁷ Someone might think that Holtug uses the phrase ‘no value’ to mean zero value. On the contrary, however, he emphasizes that he does not: ‘On the Value of Coming into Existence’, p. 381.

¹⁸ Holtug, ‘On the Value of Coming into Existence’, p. 381.

¹⁹ Bykvist, too, seems to overestimate the importance of whether S's non-existence has zero value for S. He devotes several pages to this question: see ‘The Benefits of Coming into Existence’, pp. 342–7 (as well as elsewhere in the article). True, in contrast to the relevant passages in Holtug, Bykvist does not appear to be concerned exclusively with zero intrinsic value for S on those pages. But that hardly justifies the preoccupation with whether S's non-existence has zero value for S, for anyone who takes S's existence to be extrinsically better for S than S's non-existence is very likely to hold that S's non-existence has negative – not zero – extrinsic value for S.

²⁰ Cf. Feldman, F., ‘Some Puzzles about the Evil of Death’, The Philosophical Review 100 (1991), pp. 205–27, at p. 216CrossRefGoogle Scholar.

²¹ Of course, EV2 itself does not violate actualism.

²² To say that something has basic intrinsic value is to say that it has intrinsic value in the most fundamental way. For instance, given a simple form of hedonism, states of affairs with positive basic intrinsic value have the form S experiencing pleasure to degree n at time t. For more on basic intrinsic value, see Harman, G., ‘Toward a Theory of Intrinsic Value’, The Journal of Philosophy 64 (1967), pp. 792–804CrossRefGoogle Scholar, and Feldman, F., ‘Basic Intrinsic Value’, Philosophical Studies 99 (2000), pp. 319–46CrossRefGoogle Scholar. They are concerned with intrinsic value simpliciter rather than intrinsic value for a person. But regardless of whether we want to determine a world's intrinsic value simpliciter or its intrinsic value for a person, we need to appeal to the basic, as opposed to non-basic, intrinsic values (either simpliciter or for a person) of the states of affairs that obtain in that world. Otherwise there will be double-counting and other problems.

²³ Bykvist, ‘The Benefits of Coming into Existence’, p. 348 (emphasis his). I have corrected one typo in this passage.

²⁴ Bykvist, ‘The Benefits of Coming into Existence’, p. 348. More exactly, he points out that the requirement is captured by Accessibility and the principle that A is better for S than B iff B is worse for S than A. (This principle seems obviously true.) But we can disregard this, for my approach violates Accessibility directly, since EV3 is a principle of both betterness and worseness. I have changed some of the notation in Accessibility to make it cohere with the one I use in this article.

²⁵ Bykvist has told me (in personal communication) that he intended Accessibility to cover only cases where ‘symmetry’ holds, so that, for example, ‘w1 is the closest possible not-A-world and w2 is the closest possible A-world’ holds in w1 iff it holds in w2. However, it is still worthwhile to see why the unrestricted version of the principle fails. Moreover, notice that it is far from clear that symmetry holds in a case of a person's existence and non-existence. Suppose S exists in the actual world w1 and that w2 is the closest world where S does not exist. It may well be that, relative to w2, w3 (where w1 ≠ w3) is the closest world where S exists. And it may well be that the intrinsic value for S of w3 is very different from the intrinsic value for S of w1.

²⁶ Cf. Holtug, ‘On the Value of Coming into Existence’, p. 375.

²⁷ You might object that if there is a world where W has mental states (which, apparently, there must be if W has the capacity in one world), this shows that W does after all have the capacity for mental states in all worlds where W exists. But this completely distorts the notion of a capacity. I do not have the capacity of flying, even if there is a possible world where I fly.

²⁸ Someone might object that, necessarily, all beings with psychological features have psychological identity conditions, so that nothing that has psychological features in w2 can be identical to W. But it would be a great cost for my opponents to have to rely on this principle, which is rejected by most philosophers.

²⁹ For a much more thorough discussion of this and related problems, see Bykvist, ‘The Benefits of Coming into Existence’, pp. 350–3, ‘Prudence for Changing Selves’, Utilitas 18 (2006), pp. 264–83, at pp. 273–4, and ‘Violations of Normative Invariance’.

³⁰ In ‘Person-Affecting Moralities’, The Repugnant Conclusion, ed. J. Ryberg and T. Tännsjö (Dordrecht, 2004), pp. 136–7, it appears that Holtug intends the following ‘Wide Person-affecting Principle’ to allow that an outcome, C, where everyone is badly off is worse (simpliciter) than an outcome, D, where there are no individuals, regardless of whether C or D obtains: ‘An outcome, O₁, cannot be better (worse) than another outcome, O₂, if there is no one for whom, were O₁ to obtain, O₁ would be better (worse) than O₂, and no one for whom, were O₂ to obtain, O₂ would be worse (better) than O₁.’ However, if D obtains, there is no one – and hence no one for whom, were C to obtain, C would be worse than D, and no one for whom, were D to obtain, D would be better than C. So, the principle entails that, if D obtains, C cannot be worse than D.

³¹ It is particularly cases of this kind that are usually referred to as instances of the ‘non-identity problem’. See Parfit, Reasons and Persons, part IV.

³² Indeed, some philosophers are drawn to certain versions of ‘person-regarding morality’ precisely because they want to defend this asymmetry. Clearly, however, there are other reasons to find ‘person-regarding morality’ attractive: see e.g. Holtug, ‘Person-Affecting Moralities’, pp. 137–8.

³³ Krister Bykvist has objected (in personal communication) that in order for me to deliberate between different alternative actions, these alternative actions have to exist. Therefore, he suggests, it is plausible to regard them as states of affairs. But it seems enough for deliberation that there are states of affairs such as my performing action a1 and my performing action a2. Actions themselves – e.g. a1 or a2 – are still concrete (and exist only if performed).

³⁴ This does not violate the relevant version of the idea that e.g. an action's wrongness should not depend on whether it is performed. What should not depend on whether the action is performed is the truth value of the statement that the action would be wrong if it were performed. The normative view I discussed in the opening two paragraphs of the present section violates this principle. So does the view in the third paragraph.

³⁵ Many thanks to several anonymous referees, Matthew Liao, Rebecca Roache, Julian Savulescu, Nicholas Shackel, and, especially, to Krister Bykvist and Wlodek Rabinowicz.