Is God Our Benefactor? An Argument from Suffering

Journal of Philosophy of Life Vol.3, No.3 (September 2013):145-167 Is God Our Benefactor? An Argument from Suffering Dagfinn Sjaastad Karlsen* Abstract Modern theodicies normally involve the premise that God benefited us through creation. The assumptions on which such a premise relies have, however, rarely been discussed in this context. I argue that causing someone to exist cannot benefit the person whose existence is at stake, and hence that God is not our benefactor. If anti-frustrationism is a correct theory of welfare, the concept of a benevolent Creator is incoherent. 1. Introduction1 If God has – or can be thought to have – a morally sufficient reason to permit the existence of evil and suffering, the logical problem of evil in its traditional form is solvable.2 In his ‘free-will defense’, Alvin Plantinga suggests that one such morally sufficient reason could be God’s preference for a world containing both evil and free agents – rather than a world with no evil, if this world without evil would imply that no free agents existed.3 His solution has * Alumnus, Faculty of Theology, University of Oslo. Email: dagfink[a]yahoo.no. Special thanks to Lars Øystein Ursin for valuable input and encouragement throughout the process of writing this paper. Atle Ottesen Søvik, Ian Peter Grohse, and Erin Bachynski have also given insightful comments and advice. Thanks to Dag Flem Mæland for years of influence and co-brooding. Thanks to Ellen Grimsmo Foros for putting up with her partner’s pessimism. Last but not least, I should thank Solrun Sjaastad Karlsen and Reidar Karlsen for being such loving parents – despite the thoughtless primal act. 2 Antony Flew formulates the atheologian’s traditional challenge to theism with atypical brevity: ‘The issue is whether to assert at the same time first that there is an infinitely good God, second that He is an all-powerful Creator, and third that there are evils in his universe, is to contradict yourself’ (Flew (2005), p. 60). Most philosophers would say that there is no formal contradiction here. Over the last five decades, the locus has shifted from the logical to the ‘evidential’ problem of evil (evil could still count as evidence against theism.) 3 Plantinga (1974), pp. 166–167. Plantinga distinguishes between a theodicy and a defense. A theodicy purports to give God’s real reasons for allowing evil to exist, whereas a defense only tries to rebut the accusations of contradiction. Plantinga explicitly says that his defense only establishes the logical possibility of God’s existence concurrent with actual evil: ‘So there is a significant difference between a Free Will Theodicy and a Free Will Defense. The latter is sufficient (if sucessful) to show that set A [a set of propositions entailing an existent, omnipotent, and morally good God and the existence of evil] is consistent; in a way a Free Will Theodicy goes beyond what is required. On the other hand, a 1 145 been extremely influential, and the majority of analytic philosophers of religion agree that it is valid. Its soundness, however, is not undisputed.4 Interestingly, it is primarily Plantinga’s incompatibilistic view of human freedom and the difficulties of harmonizing divine foreknowledge with free choice that have that have been attacked. Couldn’t God create us in such a way that we always freely choose to do good? If God knows the future, wouldn’t this mean that our choices are already determined?5 Some writers have called the free-will defense valid, but irrelevant, since the possibility of non-contradiction between God’s existence and the proliferation of evil doesn’t in itself make theism plausible. The theist usually believes propositions about God that are much stronger than mere possibility.6 In my view, there is another route by which the free-will defense could be criticized. I think that the philosophy of religion is somewhat lacking in its development of the meta-ethical assumptions on which it typically relies. The concept of morally sufficient reasons deserves a closer look. It is my contention theodicy would be much more satisfying, if possible to achieve. No doubt the theist would rather know what God’s reason is for permitting evil than simply that it’s possible that He has a good one. But in the present context (that of investigating the consistency of set A), the latter is all that’s needed’ (Plantinga (1977), p. 28). In this paper, I will not make much use of any distinction between between theodicy and defense, as it inevitably makes the argument more convoluted. Nevertheless, I do hope my objections will be seen as relevant no matter how high the ambitions of the free-will defender/theodicist are. 4 ‘Det råder enighet om att den fria viljans försvar, som Plantinga framför för att påvisa konsistens hos en mängd tankar i vilka Gud och det onda ingår, är giltigt. Dess sundhet är dock ifrågasatt p.g.a. hans tankar om mänsklig frihet og gudomlig kunskap’ (Rehnmann (2004), p. 171). [My translation: ‘It is agreed that the free-will defense, which Plantinga presents in order to show consistency in a set of thoughts including both God and evil, is valid. Its soundness, however, has been questioned, because of his notions about human freedom and divine knowledge.’] 5 In God, Freedom, and Evil, Plantinga addresses (and tries to rebut) the objection that divine omniscience and free will are incompatible (Plantinga (1977), pp. 66–73). According to Richard Swinburne, Plantinga makes things too complicated for himself by insisting that God knows the future (Swinburne (1998), pp. 138–145). Although interesting, the problem falls outside the scope of this article, and will not be discussed. As I share Plantinga’s incompatibilistic view of freedom and find his argument convincing in this respect, I will not address this question, either. Others have made contributions that I presently have no ambition to challenge or improve. Much has been said already about the nature of free will, and the one parameter of theodicy I think deserves further exploration, is that of goodness. 6 To be relevant, the free-will defense has to go further that it does: since it doesn’t contribute to establish how God reconciles evil with His goodness and omnipotence, it fails to show that theism as it is believed can be justified. Theodicists like David Ray Griffin and Richard Swinburne have stressed, against Plantinga, that a theodicy must be plausible and not only possible (Feinberg (2004), p. 88; Søvik (2009), p.145). I find this point instructive and important, and, in my judgement, it should trouble the theist, especially if he seeks to convince others of the viability of his position. However, if we restrict the free-will defense to logic, which is what Plantinga does, it doesn’t really do any damage to his argument. 146 that an important obstacle for theodicy lies in a reliance on notions of beneficence that are incoherent and, usually, inadequately articulated. My discussion will challenge the view that creation can make sense within the framework of morality, and thus also the very coherence of the idea of a benevolent creator.7 2. An Argument from Suffering God could have forfeited creation altogether, or he could have created a world of lesser ‘richness’ – one where evil never occurs. It would, perhaps, be sad for God to spend eternity in his own company, but no-one else would be troubled by the situation. It seems plausible to say that those who never exist suffer no harm, while existers arguably do. The principle of benevolence tells us to avoid harm to others, unless this avoidance implies greater harm. This inference looks valid: (1) There exists an omnipotent, omniscient, and morally good Creator (a ‘true God’). (2) Beings who suffer have been caused to exist. (3) A true God meets any requirement entailed by benevolence. (4) Benevolence requires that harm be avoided unless its avoidance implies greater harm or deprivation of benefits that outweigh the avoided harm. (5) The never-existent cannot be deprived. (6) The falsity of (2) avoids harm. (7) The falsity of (2) entails no harm discouraged by benevolence, nor any deprivation of benefits (from (5)). (8) A true God could have caused (2) to be false (from (1)). (9) Benevolence requires that (2) be false (from (4), (6), and (7)). 7 My critique will be structured only loosely around the free-will defense, but Plantinga lays out the ground nicely for anyone who seeks to examine contemporary theodicies more generally, and his solution looms large behind almost everything that has been written on the subject since. One could even say that most theodicies include some version of the free-will defense as one of their most essential elements. Richard Swinburne has emphasized this point on several occasions: ‘The “free-will defense,” carefully spelled out, must be a central core of theodicy, as it has been for the last two or three thousand years’ (Swinburne (2001), p. 251). ‘It would, I suggest, be very difficult to construct a satisfactory theodicy which did not rely on the doctrine of human free will’ (Swinburne (1998), p. 252). 147 From (3), (8), and (9), we see that (1) and (2) contradict each other. Hence, there cannot be a true God. There are two interesting ways to counter this argument. One could either deny that propositions (3) and (4) give a correct description of divine goodness, or one could challenge proposition (5). It may be possible to circumscribe the concept of divine goodness in such a way that it becomes ‘inscrutable’ – to liberate divine goodness from human standards of morality, so to speak. This avoids the logical problem of evil, but at some cost: it may become difficult to maintain that God’s goodness is moral in content. Most contemporary theodicists opt for the latter strategy and claim that God is our benefactor. What they do, in effect, is to deny proposition (5), but usually only implicitly. In the following, I argue that the free-will defense, though formally valid, is defective because its concept of morally sufficient reasons lacks a sound meta-ethical foundation (Chapter 3.1). I proceed to defend the view that God must be judged as a moral agent in the ‘ordinary’ sense; otherwise we cannot meaningfully praise Him for His (moral) goodness (Chapter 3.2). The rest of this paper is devoted to the question whether God can be said to have benefited us through creation (Chapters 4.1–4.4). 3.1 Reasons, and Good Reasons The theodicy par excellence is founded upon human freedom. Its tradition can be traced back to St. Augustine, and in modern times, Plantinga famously formulates his version through the use of modal logic. Plantinga’s reasoning is similar to that of Nelson Pike, who in his essay ‘Hume on Evil’ (1963) pointed out that it is difficult to find a necessarily true proposition that entails a contradiction between God’s omnipotence and goodness on one side and the existence of evil on the other. For the contradiction to arise, one has to show that God cannot have any morally sufficient reasons for allowing evil to exist. Neither Plantinga nor Pike purport to present the real reasons why God chose to create as He did. Pike observes that these could be beyond our capacity to grasp. Even if the theist lacks knowledge of what God’s morally sufficient reasons could be, this doesn’t prove that there aren’t any. Pike thinks the atheist would have to dispute an enormous list of reasons in order to prove that the theist is guilty of contradiction, and for his attack to succeed, this list must be 148 exhaustive of any reason God could possibly have. Such an ‘ennumerative procedure’ would be highly difficult and nowhere near conclusive.8 Plantinga thinks it is misguided when the atheist demands from the theist a clear answer as to why God runs the world the way He does: ‘The fact that the theist doesn’t know why God permits evil is, perhaps, an interesting fact about the theist, but by itself it shows little or nothing relevant to the rationality of belief in God’.9 The free-will defense is clearly formally valid. The logical problem of evil is solvable; its resolution only requires a morally sufficient reason for God to allow evil. However, it remains to be shown that there can be such a reason. If morality prohibits that there be a morally sufficient reason to allow evil to exist, the resolution of the logical problem provided by the free-will defense amounts to very little. Throughout the history of theodicy, the goodness of human freedom has, by and large, been understood as self-evident. It has been said that a world containing free agents is better than one without them; their existence has intrinsic value. In City of God (XI, 18), St. Augustine puts it poetically: ‘For God would never have created any, I do not say angel, but even man, whose future wickedness He foreknew, unless He had equally known to what uses in behalf of the good He could turn him, thus embellishing the course of the ages, as it were an exquisite poem set off with antitheses’.10 St. Augustine cannot plausibly be interpreted as saying that God’s reason for creating was one of aesthetics. For if God adorned the universe with agonies out of some aesthetic preference, few would view this as a sufficient reason, and hardly a moral one. In a morally ordered world, evil’s good purposes must (greatly) outweigh the badness of the total state of affairs. Theism must entertain notions of divine goodness that are compatible with its own moral world view. Terence Penelhum touched upon the problem in a short article published in Religious Studies in 1966–1967. According to him, (Christian) theists have to say ‘that the universe is run on Christian principles, and when they encounter a state of affairs which, by those principles, is evil, they must in consistency hold that it is permitted by God for reasons which are applications of those principles’.11 Penelhum argued that the concept of morally sufficient reasons 8 Pike (1963), pp. 187–188. Plantinga (1977), p. 10. 10 St. Augustine (1996), p. 384. 11 Penelhum (1990), pp. 79–80. 9 149 may be restrictive to the degree that an omnipotent God cannot justifiably allow some of the actually existing evils. If this is the case, the problem of evil is still a logical obstacle for theism.12 I have yet to see a better formulation of the challenge of moral coherence. It would be interesting to see someone of similar philosophical clarity address the topic after Plantinga. Plantinga puts great effort into explaining how human freedom logically limits God’s headroom. Important clarifications in his books (The Nature of Necessity and God, Freedom, and Evil) address the problem of evil concurrent with omnipotence, but he does surprisingly little to explain the coherence of his notion of goodness. I suspect the explanation to be that he has no clear idea; in his argument, moral goodness is simply left as a blank variable. 3.2 God as a Moral Agent The discourse of theodicy is possible only on the condition that God can be viewed as a moral agent in some ‘ordinary’ sense.13 To be a moral agent is to be subject to some standard for choosing how to act. Goodness is compliance with the standard. The criteria for goodness must be the same for all moral agents. If an agent conforms to no consistent criteria for goodness, she conforms to no standard. Morality is a standard, and a standard is by definition consistent.14 God's all-goodness implies, by necessity and definition, that He never fails to meet any of the criteria laid down by morality. There may, of course, be disagreement about what these criteria should be. This does not show that there is disagreement about the meaning of ‘goodness’; rather, it shows that ethics must seek to investigate the standard with which any moral agent is expected to comply. Morality may require different things from different agents. Some moral agents are subject to restrictions of circumstance that do not apply to others. Access to information is an obvious example. Since God is omniscient, He is a fully informed agent and has access to all knowable information about the consequences of whatever He or anyone else chooses to do. ‘Ought’ implies 12 Penelhum (1990), p. 82. ‘There is only a problem of evil, a problem of why an all-good God permits or ordains evil, either natural evil or moral evil or both, if the goodness of God bears some similarity of character to human goodness’ (Helm (1999), p. 248). 14 Hare (1952), p. 132. 13 150 ‘can’; God knows and can more than anyone else. He also has special authority over other beings. All this is reflected in morality's demands on Him.15 Some object to treating God as a moral agent like ourselves. In his book The Problem of Evil and the Problem of God (2004), D. Z. Phillips criticizes contemporary theodicists and philosophers of religion for speaking about evil in ‘contextless’ language. According to him, their consequentialist, analytic approach misconstrues the nature of religion. Phillips seems to understand religious language as something outside ordinary moral discourse.16 Against this, I would argue that it is precicely by ascribing moral agency to God that we can meaningfully call Him good or bad. I willingly admit that for the believer, there may be more imminent concerns than philosophical coherence and consistency. However, no theodicy will get off the ground unless one views God as subject to moral requirements in a way similar as any other agent. If you say that there is another standard for God, this standard wouldn't be the standard of morality. As a strategy for tackling the problem of evil, this inevitably becomes fideistic. While fideism may avoid certain problems, it is of very little help in their resolution. If by God’s goodness we mean that He takes human interests into account, and cares and suffers with us, any talk of higher purposes will be rendered meaningless unless these purposes can withstand moral scrutiny. In Christian ethics, moral goodness is normally viewed as relational. A good person displays benevolence; she is disposed to promote the welfare of others. Acts done out of divine whim and desire can only be justified by reference to God’s obligations to Himself. The moral content of divine acts, however, cannot be decided without reference to how they affect other beings. Postulating an otherwise logically coherent greater good doesn’t contribute to any moral justification of God’s acts. That greater good must be compatible with the requirements of morality.17 15 Maring (2012). Phillips (2004), p. 107. 17 Antony Flew makes the same point: ‘The possibility of affirming, and praising, God’s goodness depends on the possibility of recognizing, and condemning, the evils of the world. If you want to have it that God is good, and to praise his goodness, then you are already judging God. Furthermore, since you are yourself a man, you are judging him necessarily by some human standards—regardless of any claim that these standards of yours happen in fact to be congruent with those of your God’ (Flew (2005), p. 61). 16 151 4.1 Creation as Beneficence The primary objective of morality is to avoid harm.18 In a world which contains suffering, the non-existent will necessarily be the least harmed.19 This provides a prima facie reason not to bring sentient (and subsequently suffering) life into existence. So it seems that God must have attained some greater good by creating us, a good that outweighs the harms suffered. Were we, perhaps, not just harmed, but also benefited by being brought into existence? Against atheistic arguments from evil, D. Z. Phillips suggests that the religious believer’s response could be formulated as such: ‘I know all the evils to which you draw my attention as well as you do, but religion still offers me a way of being grateful for existence despite them; a way of thanking God for life as a whole’.20 So, by my reading, the theist holds some notion of existence as a benefit, and God, being the Creator of all life, should be viewed as our benefactor. And as our benefactor, our gratitude to Him is only appropriate.21 The notion of creation as God’s beneficence is also present in more explicitly apologetic accounts. In a concluding chapter to The Coherence of Theism, Richard Swinburne invokes the parent–child relation as an analogy in support of the traditional tenets of theism – in particular the view that the Creator is worthy of worship, thankfulness, and the like: ‘From the general duty of respect to benefactors there follow particular and strong duties of respect to those on whom we depend much for our existence, growing and flourishing’.22 Swinburne’s analogy clearly implies the claim that being brought into existence actually benefits the person. The premise seems to be that children owe their parents gratitude for being created. In a similar but even more fundamental way, any being (whose life is worth living) is benefited by God through creation and 18 Some hold that harm avoidance is the sole objective of morality. Bernard Gert is close to this position, and he concludes his Common Morality by specifying ‘the goal of common morality, which is the lessening of the amount of harm suffered’ (Gert (2004), p. 149). 19 Although it never happens, it is, I admit, logically possible to exist as a sentient being with no experience of suffering. However, an adequately informed and morally sensitive person should still see herself as harmed by the suffering of others. 20 Phillips (2004), p. 95. 21 As mentioned, Phillips objects the discourse of theodicy and claims not to partake in it. The God of the believer isn’t necessarily the God of the philosopher, and Phillips argues that most philosophers of religion distort faith as a phenomenon and miss the point by treating it analytically. Nevertheless, Phillips does formulate propositions of some interest for the analytically inclined as well. The trivial point can be made that even he speaks of faith as transitive to some degree. 22 Swinburne (1993), p. 296. 152 should show Him respect and gratitude accordingly. The parent analogy also plays an important role in John Hick’s ‘soul-making theodicy’.23 An interesting twist on the argument is found in Atle Ottesen Søvik’s doctoral dissertation, which discusses some modern theodicies in light of their conception of divine power. Søvik leans on a line of reasoning that is strikingly similar to what Derek Parfit calls the ‘non-identity problem’: born at a different time, under different circumstances, we wouldn’t be us.24 ‘One cannot complain that one should have been born in another world, for that person born in another world would not have been you’.25 Had God created a different universe, one with a lower risk of suffering, or one where He interacted more often to stop the occurence of evil and suffering, the individuals who now exist would never have come into existence. Søvik concludes that non-existence may be the only real alternative to the way our lives play out: ‘So the alternative for those who suffer in our universe is not existence in a better universe, but no existence at all’.26 There are good objections to Søvik's reductio. Could God have avoided some of the suffering in the actually existing world? Certainly. He could have created a world with different beings, or without sentient beings at all – or He could have abstained from creation altogether. The price to pay would be the non-existence of all actually existing beings. But before the theodicist can put this observation to use, he has to assert that existence is a benefit. (This is true even if one says that God’s creative act was supererogatory.) Hardly anyone would think that, before creation, God was under an obligation to create. And had the world not been created, no suffering would have ensued. Since no beings would have existed, there would also have been no-one around to be deprived of the good that the actual world contains. Thus, God’s abstinence from creation wouldn’t have caused any harm. Having created the world, however, He has caused beings who suffer to exist, and this suffering is, arguably, a harm to the existers. Since perfection is one of God’s prominent attributes, creation may even seem superfluous. Had God not created, all that would have existed would have been perfect. One might surmise that the existence of evil and suffering detracts from the goodness of a world. Furthermore, if an originally perfect state of 23 24 25 26 Hick (1990), p. 171. See Parfit (1984), pp. 351–356; and Roberts (2009). Søvik (2009), p. 255. Søvik (2009), p. 293. 153 affairs could be improved, it wouldn’t be perfect in the first place. So it appears that the current state of affairs isn’t as good as a world in which God is alone. Considerations like these lead Richard R. La Croix to argue that God should have forfeited creation: ‘Perhaps God could not, for some perfectly plausible reason, create a world without evil, but then it would seem that He ought not to have created at all’.27 Why does a being who exemplifies perfect goodness in His own person choose to create new, less perfect beings? An answer could be that goodness tends to disperse itself.28 But how does this happen? I presently see no other alternative available to the theodicist other than to claim that existence is a benefit. This premise has received some attention in population ethics, but despite its obvious importance, it is rarely discussed explicitly in the context of theodicy.29 If we now return to the free-will defense, we see that it doesn’t deserve its reputation as a solution to the logical problem of evil until it’s extended with a demonstration of how God benefited us in creating. Unless this premise can be shown to be true (or minimally: possible), I believe that the free-will defense fails, along with the other attempts at theodicy. 4.2 The Value of Existence Bertrand Russell observed that it is quite commonplace to believe that existence is better than non-existence. Russell himself wasn't able to make much sense of the notion: ‘There is a general belief (which I have never understood) that it is better to exist than not to exist; on this ground children are exhorted to be grateful to their parents’.30 Derek Parfit holds a different view. In Reasons and Persons, he raises the question in terms of whether causing a person to exist can be said to benefit this person. Parfit admits that, morally speaking, procreation is an act out of the 27 La Croix (1974), p. 28. ‘A standard answer […] is that goodness has a tendency to diffuse itself and the creation is a diffusion of the divine goodness’ (Helm (1999), p. 245). 29 It could be added that this also challenges some cosmogonies that leave God out of the picture. Although it lacks the concept of the Creator Himself, a theory like John Leslie’s extreme axiarchism (see Mackie (1982), pp. 230–239; Holt (2012), pp. 196–214) is related to divine command ethics. Leslie suggests that the world is an emanation of an objective, metaphysical sort of goodness; the universe exists because it ought to. 30 History of Western Philosophy, quoted in Holt (2012), p. 255. 28 154 ordinary, but suggests that his answer to the question could be affirmative: ‘[W]e may believe that, in causing someone to exist who will have a life worth living, we are thereby benefiting this person’.31 He postpones further discussion to an appendix, but the question has become central to modern ethical theory.32 Part four of Reasons and Persons, where it comes up, is also considered to be the locus classicus for problems of population ethics. Nils Holtug expounds on Parfit’s suggestion. He argues that existence can both benefit and harm the person whose existence is in question. His claim is founded on the view that existence is better (or worse) than non-existence. Moreover, he holds that non-existence can harm (or benefit) the non-existent. Holtug’s argument can be summarized as such: since the hypothetically actual life of a person most likely would have been ‘worth living’ had she been born, she would have been benefited by being brought into existence. Quite simply, if some actual person prefers existence to non-existence, she has been benefited by being brought into existence: ‘Since, then, Jeremy prefers existing to never existing, he has benefited [sic] from coming into existence. Had he preferred never to exist, he would have been harmed instead’.33 Put in utilitarian terms, the expected utility of existence can be used to determine whether some possible person’s creation would benefit her. Holtug calls this ‘the Value of Existence View’. His comparison between existence and non-existence is founded on the view that most lives have positive value for the existent: their lives are such that the good usually outweighs the bad; they are lives worth living. ‘[…] I should emphasize that I am claiming here that having a life in which the good outweighs the bad can be better than never existing. I am not claiming that existence as such can be better than never existing’.34 The value of a life is its value for the person whose life it is. This is important to have in mind when we assess theodicies, as well. It makes limited sense to speak of the value of existing if the positive effect of someone’s existence can only be apprecriated externally, outside the life in question. Our concern should be the same as Holtug’s, namely ‘the effect in terms of benefits and harms on the person whose existence is at stake’.35 31 32 33 34 35 Parfit (1984), p. 394. Arrhenius, Ryberg, & Tännsjö (2010). Holtug (2001), p. 365. Holtug (2001), p. 364. Loc. cit. 155 Holtug establishes a symmetry between harm and benefit in creation. It seems, on his account, that you cannot have the possibility of benefit in creation without its converse: being brought into existence can potentially both harm and benefit. Luckily, in the eyes of Nils Holtug, most lives are worth living. It can be instructive to compare Holtug’s optimistic account with an argument put forward by Richard Swinburne: since so few commit suicide, evil seems not to affect most people’s lives to the degree that they don’t have positive overall value.36 Against similar claims, it has been argued that one can coherently regret one’s own existence and still have a desire to keep on living.37 One of the virtues of Holtug’s view is that it allows us to assess certain situations that arise from the non-identity problem in person-affecting terms. If we are to decide on some course of action that will have consequences for future individuals, it doesn’t matter who these individuals will be, even if their identity is influenced by our choice. The question is which state of affairs will be most hospitable to any future person or population: which lives will be most worth living? 38 However, Holtug’s account has some weaknesses that eventually render his views utterly implausible. Although he holds that non-existence is a harm, I would still be surprised if he grieves the non-existence of a potentially happy person with an intensity similar to that with which he regrets the suffering of actual persons. Besides the non-identity problem, there are related problems which completely escape Holtug’s horizon. Since there are infinitely many possible persons who will never come into existence, grief on their part should also be infinite. The number of actually existing persons is necessarily finite, and any jubilation at their happiness must be moderated by the fact that their number could have been higher. This is absurd. If non-existence really is a harm, there would also exist a moral obligation to create as many people as possible whose lives are worth living (‘the mere addition paradox’). Parfit devotes much space to the idea, but rejects it as a ‘repugnant conclusion’, since it entails that an immense population whose lives are barely worth living would be preferrable to a relatively small population whose quality of life is very high.39 36 Swinburne (1998), pp. 251–252. Smilansky (2007), pp. 100–112. 38 This is a typical example: a woman has a temporary affliction that will cause her child to be deaf if she conceives now. If she waits five months, the chances that her baby is born healthy will be back to average. Since a baby conceived in five months will end up as another person than a baby conceived now, one cannot say that her choice benefits the baby whom she decides not to have. That person never exists. To some extent, one could say that Holtug is able to resolve this particular problem. 39 Parfit (1984), pp. 381–390. 37 156 Elizabeth Harman’s approach is similar to that of Holtug, but with the crucial difference that harms count more than benefits. For this reason, her theory fares better in the context of different people choices, but she is unable to avoid all of the problems that haunt Holtug’s article. Harman understands harm as ‘the causing of bad states’ and benefit as ‘the causing of good states’.40 By employing this definition, it is rather straightforward to claim than we can harm and benefit in creating, but it also causes some difficulty. Harman calls the article in which she discusses these problems ‘Can We Harm and Benefit in Creating?’, but interestingly, she devotes the bulk of the argument to the harm part of the question. Although many miserable lives will be said to be ‘worth living’ once a person exists, this person’s estimate of the value of her own existence is a criterion too precarious to decide whether it is morally advisable to create more persons with similar lives. Take a severely disadvantaged person. This person’s quality of life is very low compared to yours, but she feels that her own life is worth continuing (and you may also think it is worth continuing). You appreciate her perseverance and courage, but may still think it unjustified deliberately to create more persons whose quality of life will be similarly low. This is a common intiution, and shows some of the difficulties that arise if one holds that existence is a benefit. If we benefit in creating and a severely disadvantaged person prefers her life to non-existence now, it seems that this person must be among those benefited by being brought into existence.41 On this view, will not the strongest reasons count in favour of creating more lives similar to the one under consideration? Harman finds her way out of the tangle by positing a normal, healthy life as a threshold for morally justified procreation. To an extent, everyone will be harmed by existence, but some more seriously so: ‘I propose that for persons, there is a point of comparison that involves a healthy bodily state’.42 Only when a life can be expected not to fall below this threshold do we have a reason in favour of creating it. The repugnant conclusion lurks in the background, and Harman admits: ‘On my view, we do have some reasons to create every happy 40 Harman (2004), p. 105. Holtug, for instance, must embrace this consequence. 42 Harman (2004), p. 96. To avoid complete arbitrariness, healthy mental states should also be included in such a standard of comparison. Harman agrees (Harman (2004), note 18). In addition to that, attention should be brought to what kind of prospects for a ‘normally happy’ life one has in the society into which one is born. 41 157 child we could create’.43 That we have reasons to create these happy children does not, however, amount to an obligation: ‘The couple who does not create a happy child does not do anything bad; they merely fail to do something that would be good’.44 In other words, the act of creating (normally) happy persons is construed as supererogation. Harman has a strong reluctance against weighing harms and benefits equally. Harms count more. This provides further protection against the complaint that her views engender a general obligation to make life on the planet maximally abundant. The intuition that harm avoidance is more basic to morality than satisfaction of desires is a universal one. It is, in a sense, an element of our ‘common morality’. Since any sentient being will experience some amount of suffering throughout its life, wouldn’t the primacy of the harm avoidance principle lead one to think that the strongest moral reasons favour non-creation? Surely, the least harmed (whether ‘seriously’ or not) will be the non-existent. If one, like Harman, holds that benefit quite simply is the causing of good states, the moral implications of instigating a transition from one good state of affairs to another become unclear. Viewed in person-affecting terms, it may be difficult to claim that a person is better off existing than not; for if existence is better than non-existence, it seems that non-existence must be worse than existence. This can hardly be the case, since in non-existence there is no-one to be worse off. If instead we approach the problem as one of impersonal goodness, we must ask: can causing a happy person to exist (a good state of affairs) improve a world in which that person didn’t previously exist (another good state of affairs)? But, again, this position is threatened by the repugnant conclusion. To avoid this obstacle, one could say that there is no improvement involved. What, then, is the nature of the benefit in question? If we have the option of choosing freely between two good states of affairs, it seems that our choice will be beneficial no matter what we do. Such an account of beneficence risks making the whole concept redundant. In my judgement, Harman shows, convincingly, that we can harm in creating. However, she infers from the possibility of harm a parallel possibility of benefit. This is fallacious. As will become apparent below, one can hold that there is harm but no benefit in creating. 43 44 Harman (2004), p. 98. Loc. cit. 158 4.3 Ascribability Derek Parfit admits that, morally speaking, causing someone to exist is ‘a special case’. He does insist, however, that it is possible to say that the created person is benefited thereby: ‘Some argue that this is not benefit because it lacks some feature shared by all other benefits. But this begs the question’.45 This particular way of putting it may be question-begging, but I find little satisfaction in stopping here. It seems natural to ask what requirements an action must meet to qualify as a benefit. Creation of new beings differs importantly from other instances of benefiting on at least one point: the ‘position’ of the beneficiary. Mustn’t a benefit necessarily be directed at someone? Bringing someone into being does not meet this requirement. Instead, it causes the subject's existence, a subject who in turn can be harmed or benefited. I contend that existence is a necessary condition for the possibility of being benefited. We can call this the ascribability condition. My suggestion is related to Melinda Roberts’ Variabilism theory. Both accounts allow for the moral status of a person to be connected to her modal status. According to Roberts, a possible person’s ‘loss’ in welfare or benefits has no moral significance in a world where that person never exists.46 Her theory entails a variant of proposition (5) in my argument above: the never-existent cannot be deprived in any morally relevant sense. By choosing not to consider such objections, Parfit leaves one of the biggest questions unanswered. 4.4 The Advantages of Non-Existence Barring all other conditions, it is good when a preference47 is satisfied. But is a preference plus its satisfaction better than no preference? In ‘A Pareto 45 Parfit (1984), p. 490. The author doesn’t say whom he criticizes here. ‘The loss of wellbeing incurred at a world where the person who incurs that loss does or will exist has full moral significance for purposes of evaluating an act performed at a given world that imposes that loss and any alternate act performed at any alternate world that avoids that loss, while a loss incurred by that very same person at a world where that person never exists at all has no moral significance whatsoever’ (Roberts (2011), p. 773). For a fuller discussion of the assumptions implicit in Roberts, I would have had to address notoriously knotty questions about ‘transworld identity’. That belongs to a paper whose modal status is yet undetermined. 47 Fehige’s term is ‘preferences’, while Benatar likes to speak of ‘desires’. In this paper, I use the two 46 159 Principle for Possible People’, Christoph Fehige forcefully denies that it is so.48 He describes his position as ‘anti-frustrationism’. What makes one world better than another is not its amount of preference satisfaction, but the avoided preference frustration: ‘we have obligations to make preferrers satisfied, but no obligations to make satisfied preferrers’.49 On Fehige's account of anti-frustrationism, a maximum level of utility can be ascribed to everybody whose preferences are all satisfied. This covers three cases: ‘the individual exists and has preferences and they are all satisfied, or it exists but has no preferences (say, it's a stone, or a chair), or it does not exist’.50 As a consequence, all merely possible people have maximum utility. Their welfare wouldn't be higher if they were brought into existence. Nobody can be better off than a ‘non-preferrer’.51 Among its virtues, anti-frustrationism rescues total utilitarianism from the repugnant conclusion. If utility is measured by the principle of harm avoidance instead of aggregated preference satisfaction, utilitarianism does not, as the accusation often goes, entail that it is better the more (acceptably) happy lives there are. Fehige stresses, however, that anti-frustrationism is not motivated by ‘the premiss that the Repugnant Conclusion is repugnant’, but on ‘general reflections about the concept of welfare’.52 David Benatar’s Better Never to Have Been: the Harm of Coming into Existence builds upon Fehige’s article and has received much attention since it was published in 2006. Like Fehige, Benatar argues that the absence of pain is good, but, at the same time, that the absence of pleasure isn’t necessarily bad – as long as there is no-one around to be deprived. From this asymmetry he deduces a moral impermissibility of procreation. Benatar would say that Holtug conflates ‘a life worth continuing’ and ‘a life worth starting’.53 When the exister claims to have a life worth living, this says very little about existence as a benefit, which it isn’t. Most lives are worth living from the viewpoint of the interchangeably. The term ‘interest’ is wider than ‘preference’/’desire’; you can have interests but no perferences/desires. Young children usually have no articulated preferences regarding their parents’s financial dispositions, but it can still be in a child’s interest that her parents don’t squander money that would otherwise go into a savings account. 48 The opposite view, that we do good by creating satisfied preferences, Fehige calls orexigenesis (from Gr. orexis, ‘appetite’). 49 Fehige (1998), p. 518. 50 Fehige (1998), p. 524. 51 Fehige (1998), p. 525. 52 Fehige (1998), p. 536. 53 Benatar (2006), p. 22–28. 160 living person, but according to Benatar, none are worth starting. His position is often referred to as ‘anti-natalism’. Two crucial steps can be identified in Benatar’s argument for anti-natalism. Step number one is to formulate the asymmetry independently of ‘worseness’ and ‘betterness’. Even when an outcome of some action cannot be said to be worse for a person than its alternative, it can still make sense to say that it is bad for her. Similarly, the possibility that an outcome is good for a person is not logically precluded by its not being better for her. This may seem pedantic, but I believe it rescues Benatar’s argument from one important objection, namely the complaint that non-existence cannot have any value for the non-existent, and thus cannot be better than existence. Krister Bykvist stresses that the only value one can plausibly ascribe to non-existence is preventive: ‘We could then say that non-existence can be preventively good for a person, to some degree, in the sense that things would not have been bad, to the same degree, for the person, if he had not existed. This also shows that it can make sense to wish that one had not been born’.54 Benatars agrees.55 Instead of restricting his assessment of procreation to a life’s value for the the exister, he combines a personal and an impersonal approach. He emphasizes that his comparison should be viewed not as one between different states of the person, but between states of affairs. Step number two in the argument is to endorse anti-frustrationism: ‘According to this view, a satisfied preference and no preference are equally good. Only an unsatisfied preference is bad’.56 Per Algander identifies the principle involved as one saying that there is a moral reason in favour of an act ‘only if the consequences of not performing the act would be bad for someone’.57 I think this distorts Benatar’s position. When a person exists, there are reasons to promote this person’s interests, viz. the person’s preference for her interests being promoted. Even when failure to promote a person’s interests cannot be said to be bad for this person, the very existence of these interests provide a reason in favour of their being promoted.58 There are, however, no 54 Bykvist (2007), p. 355. The sense in which non-existence can be better than existence cannot be called a benefit: ‘Since no one can be in a state of non-existence, no one can enjoy the preventive goodness or betterness of non-existence. This means that if you believe that preventive goodness of non-existence is a kind of benefit, you would also have to believe the absurd view that there are benefits that no one can enjoy’ (loc. cit.). 55 Benatar (2006), p. 4. 56 Benatar (2006), p. 54. 57 Algander (2012), p. 150. 58 If no-one opens a store selling high-end Italian road bikes, it seems odd to say that this is bad for 161 reasons to promote the interests of the never-existent, for if a person never exists, no interests can be ascribed to that person. A person who will come into existence, on the other hand, may have future interests that count morally, even now. Anti-frustrationism doesn’t preclude this. We can, and should, distinguish between an existent person’s future interests and a future (possible) person’s interests. The former count unconditionally. The latter count (and exist) only if an actual person comes into existence. Fehige emphasizes a similar point: ‘[A]ntifrustrationism should not be confused with the grotesque view that a person's future preferences do not count. Of course they do – just not orexigenically’.59 It is good when a desire is satisfied, but morally speaking, there is nothing inherently better about a desire being created and then satisfied than a comparable situation where that same desire never arises: ‘What matters is not that people have satisfied desires but that they do not have unsatisfied ones. It is the avoidance of frustration that is important’.60 The upshot is that the fact that a desire will be satisfied, doesn’t count in favour of creating it.61 The inevitable frustration of some desires and interests once you have sentient life, on the other hand, counts fully against their creation. In other words, anti-frustrationism entails a presumption against creating new life. A similar point, although more directed at formulating the problem than arriving at this particular conclusion, was made by Derek Parfit: ‘When we consider the badness of suffering, we should claim that this badness has no upper limit. It is always bad if an extra person has to endure extreme agony: And this is always just as bad, however many others have similar lives. The badness of extra suffering never declines’.62 Extra happiness, however, is not good in itself. Parfit echoes Jan Narveson’s famous distinction between ‘making me. However, if such a store opens in my neighbourhood, it would be good for me. Why? Because I already have interests on which basis the creation (and subsequent satisfaction) of new preferences would be good. My objection may be based only on the suspicion of sloppy usage of the words ‘good’ and ‘bad’, although I have the impression that it influences Algander's conclusions. He formulates the asymmetry by speaking of procreation as an act that ‘would be good for a person’ (Algander (2012), p. 152). This is tantamount to holding the premise that existence is a benefit, which is begging the question. 59 Fehige (1998), p. 518. 60 Benatar (2006), p. 55. 61 An entertaining example found in Fehige: ‘[S]uppose we paint the tree nearest to Sydney Opera House red and give Kate a pill that makes her wish that the tree nearest to Sydney Opera House were red – have we really done her a favour?’ (Fehige (1998), pp. 513–514) Naturally, Fehige denies that this is so. 62 Parfit (1984), p. 406. 162 happy people’ and ‘making people happy’: ‘On his view, it is not good that people exist because their lives contain happiness. Rather, happiness is good because it is good for people’.63 It has been argued that before conception, bestowal of existence cannot harm or benefit in purely person-affecting terms. If your intuitions tell you that any advantage of non-existence over existence cannot be explained through its effects on persons, you will still be likely to think that, at least in some cases, non-existence would be the better scenario. Think of the prematurely born infant who is given a few days of extreme agony if she is artificially held alive. Benatar will come to aid you. By supplementing the person-affecting view with an impersonal one, one can say that non-existence has an advantage over existence. Non-existence can be better, not because it contains more or greater benefits than existence, but because more harm is avoided.64 Benatar, of course, also says that existence never has any advantage over non-existence.65 It is easy to see how anti-frustrationism points towards a rejection of the moral permissibility of procreation. It should be noted that no prohibition follows by necessity from a principle’s entailing a presumption against something. The principle in question and the presumption which it entails must be worked into a larger moral scheme for there to be a prohibition. Anti-frustrationism implies that bringing someone into existence cannot coherently be said to benefit the created person. It does not follow that creation cannot benefit anyone. In many cases, there will be existent preferrers who will be greatly benefited by another person’s coming into existence. ‘Making happy people’ can make people happy. Although the satisfaction of a preference cannot, by itself, count in favour of creating it, we shouldn't say that there cannot be any reasons for creating it. Already existent preferences often give us reason to create new ones. New and satisfied preferences may be instrumentally good. They may contribute to the satisfaction of some other, already existent preference or some preference that will, no matter what, exist in the future. If you induce in me a desire to have a particular painting and I end up buying that painting, this may satisfy my already present desire for having a beautiful living room. My hitherto vague sensation that ‘something is lacking’ in my living room 63 Parfit (1984), p. 394. I believe it is unnecessary to scrap the person-affecting intuition to be able to say this. The betterness of non-existence can be founded on the negative effects that existence would have had on persons. But I am not going to press the point. 65 Benatar (2006), p. 14. 64 163 is now stilled. My desire for the painting is not an intrinsic good, but, as it turns out, the overall utility of my life is higher with that desire created and satisfied than without. The same is true for procreation: there usually are preferences around to begin with. Although procreation cannot coherently be said to benefit the person whose existence is at stake, it may benefit her parents and their community. The newborn may in turn enjoy benefits – but only if she exists and and has interests.66 Anti-frustrationism doesn’t directly entail anti-natalism, but anti-frustrationism in combination with the principle against treating other persons as a means to an end does. David Benatar bites the bullet and thinks that his conclusions are consistent with the widest range of moral intiutions. 67 Whereas someone like Holtug simply ignores ‘the mere addition paradox’, anti-natalism as championed by Benatar provides an answer: the optimal size of the world’s population is zero.68 Although a coherent philosophical position, as a programme with ambitions outside the domain of personal choice, anti-natalism is entirely unpracticable. It seems unlikely that it will gain popular support. This is lamentable, because much avoidable suffering will be caused by people not seriously considering the alternative of not having children. 5. Conclusion I have argued that God Himself is the only being who can coherently be said to have been benefited through His decision to create. Few standards of morality would teach that this is a sufficient reason. In any event, it is not one that springs from benevolence. If my argument is sound, a world without sentient beings would be preferable to the actual world. A benevolent God alone in an otherwise empty world would have kept it that way.69 66 Stones and rivers exist without interests. Since these are non-sentient entities, they cannot appreciate the goodness of grass growing in their vicinity. Pure existence – without interests – never happens to people. 67 Both Benatar and Fehige stress that their conclusions are compatible with Parfit’s elusive ‘Theory X’, that theory of beneficence which solves the non-identity problem, but avoids, among other related problems, the repugnant conclusion (Benatar (2006), p. 178; Fehige (1998), p. 540). 68 Animal suffering is no different in this respect. However, human beings have the unique ability to understand their predicament and make morally informed choices accordingly. 69 I leave the question unanswered whether God should have created suns and lifeless planets for company. As long as no other beings come into existence, God would perhaps be free to create according to His own whim. If so, wouldn't He be required to create such that it would be impossible for sentient life ever to evolve? It seems that a world with a risk (however minimal) of suffering is not 164 By my reading, the free-will defense implicitly relies on the highly problematic premise that there is extra value in new and satisfied preferences (orexigenesis). In other words, the theodicist needs to show that anti-frustrationism is false. I find it puzzling that most philosophers of religion don’t pay attention to the potential advantages of non-existence – at least for the sake of rigorous argument. They could end up with the parent analogy turned upside down: as any existent person has been wronged by her parents by being brought into existence, creation itself deserves only our resentfulness. References Algander, Per (2012). ‘A Defense of the Asymmetry in Population Ethics’, Res Publica, 18, pp. 145–157. Arrhenius, Gustaf; Ryberg, Jesper; & Tännsjö, Torbjörn (2010). ‘The Repugnant Conclusion’, in Zalta (ed.) Stanford Encyclopedia of Philosophy, fall 2010 edition [http://plato.stanford.edu/archives/fall2010/entries/repugnant-conclusion, accessed Dec 12, 2012]. Saint Augustine (1996). The Confessions; The City of God; On Christian Doctrine. Chicago: Encyclopœdia Britannica, Inc. Benatar, David (2006). Better Never to Have Been: The Harm of Coming into Existence. New York: Oxford University Press. Bykvist, Krister (2007). ‘The Benefits of Coming into Existence’, Philosophical Studies, 135, pp. 335–362. Fehige, Christoph (1998). ‘A Pareto Principle for Possible People’, in Fehige & Wessels (eds.) Preferences. Berlin: Walter de Gruyter, pp. 508–543. Feinberg, John S. (2004). The Many Faces of Evil: Theological Systems and the Problems of Evil, revised and expanded edition. Wheaton: Crossway. Flew, Antony (2005). God and Philosophy. Kindle edition. Amherst: Prometheus. Gert, Bernard (2004). Common Morality: Deciding What to Do, Kindle edition. New York: Oxford University Press. Hare, R. M. (1952). The Language of Morals. London: Oxford University Press. as good as one without that risk. 165 Harman, Elizabeth (2004). ‘Can We Harm and Benefit in Creating?’, Philosophical Perspectives, 18, pp. 89–113. Helm, Paul (1999). ‘Goodness’, in Quinn & Taliaferro (eds.) A Companion to Philosophy of Religion. Malden: Blackwell Publishing, pp. 243–249. Hick, John (1990). ‘Soul-Making and Suffering’, in Adams & Adams (eds.) The Problem of Evil. Oxford: Oxford University Press, pp. 168–188. Holt, Jim (2012). Why Does the World Exist? An Existential Detective Story, Kindle edition. New York: Liveright Publishing Corporation. Holtug, Nils (2001). ‘On the Value of Coming into Existence’, The Journal of Ethics, 5, pp. 361–384. La Croix, Richard R. (1974). ‘Unjustified Evil and God’s Choice’, Sophia, 13, pp. 20–28. Mackie, John L. (1982). The Miracle of Theism: Arguments for and against the Existence of God. New York: Oxford University Press. Maring, Luke (2012). ‘A new problem of evil: authority and the duty of interference’, Religious Studies, 48, pp. 497–514. Parfit, Derek (1984). Reasons and Persons, Kindle edition. New York: Oxford University Press. Penelhum, Terence (1990). ‘Divine Goodness and the Problem of Evil’, in Adams & Adams (eds.) The Problem of Evil. Oxford: Oxford University Press, pp. 69–82. Phillips, D. Z. (2004). The Problem of Evil and the Problem of God. London: SCM Press. Pike, Nelson (1963). ‘Hume on Evil’, The Philosophical Review, 72/2, pp. 180–197. Plantinga, Alvin (1974). The Nature of Necessity. London: Oxford University Press. Plantinga, Alvin (1977). God, Freedom, and Evil. Grand Rapids: Eerdmans. Rehnmann, Sebastian (2004). Tänkesätt: studier i Alvin Plantingas filosofi. Skellefteå: Norma bokförlag. Roberts, Melinda (2009). ‘The Nonidentity Problem’, in Zalta (ed.) The Stanford Encyclopedia of Philosophy, fall 2009 edition [http://plato.stanford.edu/archives/fall2009/entries/nonidentity-problem, accessed Mar 2, 2013]. Roberts, Melinda (2011). ‘An Asymmetry in the Ethics of Procreation’, Philosophy Compass, 6/11, pp. 765–776. 166 Smilansky, Saul (2007). ‘Preferring Not to Have Been Born’, in Ten Moral Paradoxes. Malden: Blackwell Publishing, pp. 100–112. Swinburne, Richard (1993). The Coherence of Theism, revised edition. Oxford: Clarendon Press. Swinburne, Richard (1998). Providence and the Problem of Evil, Kindle edition. Oxford: Clarendon Press. Swinburne, Richard (2001). ‘Some Major Strands of Theodicy’, in Rowe (ed.) God and the Problem of Evil. Malden: Blackwell Publishers, pp. 240–264. Søvik, Atle Ottesen (2009). ‘The Problem of Evil and the Power of God’, dissertation. Oslo: MF Norwegian School of Theology. 167