Evil in the world poses two problems for theism. According to the logical problem, evil is logically inconsistent with the existence of an all-perfect God. According to the evidential problem, there is more evil than can be justified as necessary to secure some greater good; such evil is gratuitous and hence would not be allowed by an all-perfect God.Footnote 1
It is generally thought that the logical problem is easily rebutted. As a logical matter, evil can coexist with an all-perfect God insofar as there could be a morally respectable reason for allowing evil. The existence of God seems consistent with any evil necessary to achieve a greater moral good. Thus, if there is a logically coherent story about some greater moral good that cannot be achieved without evil, then God's existence is consistent with the occurrence of evil.
The free-will argument (FWA) provides such a story. According to FWA, it could be true that the existence of free beings is a greater moral good that cannot be achieved by even an omnipotent God without allowing some evil, because it is nearly inevitable that free beings will do evil. Accordingly, the moral value free beings add to the world is a possible justification for an all-perfect God to allow evil.
While FWA has been traditionally directed at the logical problem of evil, some philosophers believe it can also ground a solution to the evidential problem of evil – which, as a claim about the evil in this world, requires a solution that shows that God's allowing such evil is actually justified. Augustine, for example, argues that ‘[n]either the sins nor the misery are necessary to the perfection of the universe, but souls as such are necessary, which have the power to sin if they so will’.Footnote 2 On this line of analysis, the claim that free beings add moral value to the world is not just possibly true; it is actually true and hence would justify God's creating a world in which evil occurs. If this is correct, then FWA functions not only as a defence, but also as one component of a theodicy that shows what evil there is in this world is compatible with God's existence. If successful qua theodicy, FWA rebuts both the logical and evidential problems of evil.
I argue that FWA, as formulated, cannot adequately ground a theodicy that is sufficient to alleviate the doubts about God's existence to which the problems of evil give rise. The problems of evil cause doubts about God's existence precisely because it seems, from the standpoint of ordinary intuition, that a morally perfect, omnipotent being would never allow so much evil (which usually takes the form of great suffering). What is needed to alleviate these doubts is a theodicy that is grounded in intuitions that are at least as forceful and widely shared. I argue that the claim that, other things being equal, a world in which free beings exist is morally preferable to a world in which free beings do not exist (hereinafter ‘VT’ as shorthand for the ‘value thesis’) has no foundation in common intuitions about morality.Footnote 3 Without some sort of intuitive support, VT lacks the resources to serve as the foundation for a theodicy that addresses the powerful intuition, which affects believers and non-believers alike, that a perfect God would not allow so much evil.
FWA and VT
FWA is based on the claim that an all-perfect God would be willing to allow some evil if necessary to achieve a greater moral good; after all, a morally perfect God would want to create, if not the morally best of possible universes, one that is morally worthwhile. Thus, if it could be true that God cannot secure some more important moral value without allowing some evil, then God could be justified in allowing such evil. Accordingly, the occurrence of evil per se is not problematic for classical theism; what would be problematic for classical theism is the occurrence of evil that could not be justifiably allowed by an all-perfect God (i.e. evil that is not justified at any possible world as necessary to secure a greater moral good).
FWA reconciles evil with a morally perfect God by pointing to the moral value free beings introduces into the world. As Michael Tooley describes FWA in the Stanford Encyclopedia of Philosophy:
A second important approach to theodicy involves the following ideas: first, that libertarian free will is of great value; secondly, that because it is part of the definition of libertarian free will that an action that is free in that sense cannot be caused by anything outside of the agent, not even God can cause a person to freely do what is right; and thirdly, that because of the great value of libertarian free will, it is better that God create a world in which agents possess libertarian free will, even though they may misuse it, and do what is wrong, than that God create a world where agents lack libertarian free will.
Likewise, in the Routledge Encyclopedia of Philosophy, Marilyn McCord Adams states:
Free-will approaches contend that: (A1) created free will is a very great good, whether intrinsically or as a necessary means to God's central purposes in creation; and (A2) God cannot fulfill his purposes for and with free creatures without accepting the possibility that some will misuse their freedom, thereby introducing evil into the world …. The introduction of evil into the world is explained by the doctrine of the Fall …. God wanted them freely to choose what is right and good, but some angels and the primordial humans Adam and Eve chose what is wrong.Footnote 4
That these are encyclopedia entries suggests that the most common formulation of FWA relies heavily on VT. In any event, the reader should understand this paper as evaluating only VT.
The value thesis
Although FWA purports merely to show that evil is not logically inconsistent with God's existence by showing how it could be true that an all-perfect God would allow evil to secure a greater moral good, FWA is typically expressed in terms that suggest that VT is actually true.Footnote 5
There are good reasons for this. First, VT is not restricted to some proper subset of possible worlds. The claim is not that given any two possible worlds α and β instantiating properties P1, …, Pn, if α contains free beings and β does not, then α is morally preferable to β. The claim is rather that, given any two possible worlds α and β, if α and β resemble each other in every respect compatible with there being free beings on α but not on β, then α is, as an objective matter, morally preferable to β. By its own terms, then, VT states a claim that purports to be true of every possible world: the existence of free beings makes any logically possible world morally better than it otherwise would have been.
Second, VT is most plausibly construed as asserting an objective, and hence necessary, moral truth. In this connection, it is important to note that the various atheistic arguments from evil presuppose the truth of moral objectivism. The claim is not that most people believe either that an all-perfect God would not, as a moral matter, allow the existence of evil or that there is gratuitous evil in the world that cannot be justified by any greater moral good – as would be relevant if normative ethical relativism were true. The claim is rather that, as an objective moral matter, what evil there is in the world cannot plausibly be reconciled with the existence of an all-perfect God. Given that the problem of evil presupposes moral objectivism, any response to the problem of evil must also be grounded in claims about morality that are objectively true. It is, in part, for this reason, that VT is typically expressed in terms that assert that it is necessarily true and hence true in this world.
Motivating the value thesis
In one respect, VT is very natural. Most people believe we have a moral status unique among material things and that we count for more, morally speaking, than any other known material being in virtue of this unique status. Given that what distinguishes us from every other known material being is that we instantiate the capacity for free and rational choice, it is quite natural to think that our existence makes this world more valuable, morally speaking, than it would have been without us.
But even if the capacity for free choice explains our unique moral status in the world, it does not follow that our existence adds moral value to the universe. While we are unique in having both moral rights and moral obligations, neither claim obviously entails that the world is a morally better place because we exist. The claim we are unique in this respect entails only that there are moral constraints on what we can do and moral constraints on what moral agents can do to us. But neither claim clearly implies our existence adds moral value to the world or makes the world a morally better place.
Indeed, VT is far from obvious. To see why, let us begin by considering a couple possible worlds. Let ω1 and ω2 be two possible worlds defined as follows: (1) human beings exist and have free will in both worlds; (2) ω1 has considerably less suffering than ω2 (because, say, genocidal maniacs in ω2 take out half the world's population every 200 years or so while this sort of event never takes place on ω1); and (3) ω1 and ω2 resemble each other as closely as is consistent with it being true that ω1 has considerably less suffering than ω2.
It seems clear ω1 is morally preferable to ω2 in the sense that an all-perfect being would choose, other things being equal, to create ω1 over ω2 – if it is simply a matter of that being's choosing which world to bring into existence.Footnote 6 Clearly, suffering results, so to speak, in moral disvalue; and this is true of suffering in any being capable of experiencing it. Regardless of what position one takes with respect to the permissibility of eating meat, we should avoid causing unnecessary pain to animals. Causing unnecessary pain to non-human animals constitutes cruelty, and cruelty to animals is morally impermissible.
Notice, however, that, in the relevant sense, ω1 remains morally preferable to ω2 even if we assume that human beings are not free on either world. It seems obvious that if, (1) these are the only two possible worlds; (2) an all-perfect being must, for whatever reason, bring one of them into existence; and (3) it is simply a matter of choice which of these two possible worlds comes into existence, then such a being would bring ω1 into existence – and not ω2. Suffering is clearly a morally relevant characteristic of a universe without regard to whether there exist free beings in that universe. Though no one can be blamed for the suffering on either ω1 or ω2 because there are no free beings on either world, a morally perfect being would regard such suffering as having presumptive moral disvalue.
These reactions are grounded in intuitions that theists, atheists, and agnostics have in common. Most, if not all, conscientious sceptics have a strong intuition that, other things being equal, suffering should be minimized. For this reason, the judgment that people should do what they can to alleviate unnecessary suffering needs no argument in most contexts. Indeed, it is precisely because conscientious sceptics share such intuitions that they regard the problems of evil as sufficient to refute classical theism.
To see that, in contrast, VT needs an argument, let us consider two additional possible worlds, ω3 and ω4, defined as follows: (1) human beings exist in both worlds; (2) human beings in both worlds cause tremendous suffering to humans and non-humans that does not result in anything justifying the infliction of such suffering; and (3) the two worlds resemble each other in every respect consistent with humans being free on ω3 but unfree on ω4.
While there are a number of shared moral intuitions about free beings, none of them is relevant with respect to VT. Most people would agree ω3 is morally worse off than ω4 in that much of the suffering on ω3 has a morally culpable origin in the behaviour of free human beings, but this is not sufficient to secure VT: the claim that ω3 is morally worse off in some respect than ω4, by itself, does not imply that ω3 is morally preferable in another respect to ω4. Similarly, most people would agree it is morally good that free beings be allowed to exercise their freedom within reasonable limits, but this tells us only that a world in which free beings are allowed to exercise the capacity for choice is, other things being equal, morally preferable to a world in which they are not allowed to do so. But these judgments, which are shared by theists and non-theists alike, say nothing about whether the instantiation of that capacity per se has moral value – and this is true of any standard that tells us how free beings should be treated.
It is just not intuitively clear that the existence of free beings, by itself, makes a universe morally better than it would otherwise have been. Whereas one can sensibly contest this claim from an objectivist perspective, this is clearly not the case with respect to the claim that suffering results in presumptive moral disvalue. If ordinary intuitions are correct, no-one could sensibly contest that suffering results in presumptive moral disvalue and hence must be justified.Footnote 7
To say, of course, that VT is not intuitively clear simply tells us that VT does not express the content of some fundamental moral principle that is widely accepted; it does not tell us VT cannot be derived from other widely shared moral commitments. In the next section, I consider a number of possible arguments that attempt to ground VT in other moral intuitions and argue that none of these arguments succeeds.
The value of free beings: the arguments from love
The existence of beings capable of loving God
The first argument grounds the value of free will in the value of the capacity to love God. On this argument, the existence of beings capable of loving God adds moral value to any world in which they exist. But the capacity to love adds moral value to the world only insofar as this capacity is combined with free will. Since, as before, God cannot create beings with free will without allowing the evil that such beings will sometimes choose, God cannot secure the moral value brought into the world by beings capable of freely loving God without allowing such evil. Since the capacity freely to love God is a greater moral good that outweighs such evil, God is morally justified in allowing such evil.
This approach cannot logically ground VT because the argument is structured to support a narrower conclusion than VT. This argument, if sound, supports the conclusion a world in which there exist free beings capable of love is morally better than it otherwise would be. That is, this argument purports to show that it is the existence of beings that instantiate both the capacity to love and the capacity for free choice that makes a world morally better than it would otherwise have been. But VT asserts that the existence of the capacity for free choice, by itself, makes a world morally better than it would otherwise have been.
But it isn't clear why the existence of beings capable of love adds moral value to the world only insofar as they also instantiate free will. The instantiation of free will is clearly not a necessary prerequisite for the capacity to love. After all, if our folk theories are correct, infants and young children have this capacity despite lacking a developed capacity for free choice. Further, many domestic animals are capable of profoundly bonding with human beings despite lacking free will. Indeed, it is precisely because such animals are capable of something closely resembling love that so many of us regard them as important members of our households.
John Hick argues that love that isn't freely chosen is not maximally valuable for the same reason that the acts of someone carrying out a post-hypnotic suggestion are not maximally valuable – namely that such behaviours would be pre-selected by something other than the agent:
Just as the patient's trust in, and devotion to, the hypnotist would lack for the latter the value of a freely given trust and devotion, so our human worship and obedience to God would lack for Him the value of a freely offered worship and obedience. We should, in relation to God, be mere puppets, precluded from entering into any truly personal relationship with Him.Footnote 8
While God could consistently construct human nature so that human beings always love God, this would result in our being ‘mere puppets, precluded from entering into any truly personal relationship with Him’.Footnote 9 And love that is ‘truly personal’, on Hick's view, has considerably greater moral value than love that is not.
Hick's argument assumes too much. First, while it might be true that a puppet's love lacks moral value, it is not clear that creating human nature such as to assure that all human beings love God results in our being ‘puppets’ in a morally relevant sense. A child will almost inevitably respond with love to sustained interactions with other loving human beings, especially her parents. But it is certainly not true that this fact makes children puppets incapable of having genuinely personal relationships with us that have tremendous moral value.
As it turns out, the same is true of pets. It is almost inevitable that a dog will respond with something that, on folk theories, counts as something akin to love to sustained kindness from a human companion. Although dogs lack free will and hence are incapable of freely choosing love for human beings, this neither makes them puppets nor precludes our having relationships with them that have tremendous moral value to us; the sheer devotion of most persons to their pets is, I think, testimony to how much value they assign to the love of an animal.
Indeed, for many people, the value of even a pet's love is moral in character. I have frequently heard people remark that someone who does not have any feeling in response to the qualities of affection and loyalty in such animals is, on this common reasoning, capable of less emotion and warmth than someone who does respond to such qualities. But failure to respond to such qualities indicates a moral defect only to the extent that such qualities have moral value: if failure to respond to the devotion of an animal is morally problematic in some way, it is because such a person is failing to acknowledge moral value that ought to be acknowledged in a certain way – and, in this case, it is the love a pet feels that has moral value.
This is not counterintuitive. It is not altogether uncommon to encounter someone who believes that the love a dog feels for its human companions is morally more impressive than the love that adult human beings feel for each other. A dog's love, on this intuition, has great moral value for being utterly unconditional in a way that, say, the love that a parent ought to feel for her child is all too frequently not. The unconditional quality of its love evokes the sort of respect from human beings implicit in platitudes like ‘a man's best friend is his dog’.
One could intuitively respond that our relationships are more deeply personal and hence morally more valuable because of our greater intellectual abilities. As lovely as a dog's love is, there are obvious limits on a dog's ability to interact with other beings. Lacking the capacity for thought and language, a dog cannot communicate much with any other being and is hence limited in what it can reciprocate in a relationship with another being. In contrast, our relationships with one another are capable of profound depth and intimacy precisely because our feelings and ideas are deeper, more intimate, and more complex and because we can communicate these feelings and ideas to other people. This ability sets inter-human relationships on a much higher level, morally speaking, than any relationship involving a non-rational animal.
But notice that free choice doesn't play any role in this. What makes our love so much more valuable, on this line of argument, than a domestic animal's love is that we are capable of experiencing and expressing deeper, more intimate, and more complex content in virtue of our superior intellectual abilities. Our superior intellectual abilities facilitate morally more valuable relationships and interactions than non-rational beings are capable of. Regardless of whether human beings are free, then, it seems reasonable to think that our relationships and interactions are morally more valuable because our intellectual complexity makes us capable of deeper and more complex intimacy.
Hick would not deny this. As he concedes, ‘[t]here might, indeed, be very great value in a universe of created beings who respond to God in … love and trust and worship which He has Himself caused to occur by His initial formulation of their nature’.Footnote 10 Indeed, it seems counterintuitive to deny that deeply felt unconditional devotion and love has presumptive moral value wherever it occurs – including in animals.
Even so, Hick insists freely chosen love has greater moral value than love that flows from a nature that makes it inevitable. As he puts the point,
… if human analogies entitle us to speak about God at all, we must insist that … a universe [in which loving God is part of a being's nature] could be only a poor second-best to one in which created beings, whose responses to Himself God has not thus ‘fixed’ in advance, come freely to love, trust, and worship Him.Footnote 11
Hick's view seems quite intuitive. As a general matter, a freely chosen gesture of affection means considerably more to us than one that is not freely chosen. An apparently affectionate gesture or gift motivated by a spontaneous and free volition means much more to me than a gesture or gift motivated by some other reason. The view that freely chosen love has greater moral value than love that is a psychologically inevitable response harmonizes very nicely with these intuitions.
But there are important disanalogies between love of God and gifts and apologies. Regardless of whether we come equipped with a strong predisposition to love God, we face tremendous temptations to pursue our own self-interest to the detriment of our relationship with God. We are born with intensely egoistic desires and urges strong enough to test the wills of even the most devoted Christians. It is difficult for anyone always to resist the temptation to objectify other people in ways that range from pre-judging them as unworthy to lusting after them. And resisting temptation is difficult (and morally praiseworthy) precisely because doing so involves a ‘sacrifice’ from the perspective of our own passionately felt material desires.
Indeed, one could argue that what really matters, morally speaking, with respect to what we bring to our relations with God is the way in which we express our love and faith in freely chosen and deeply felt sacrifices. Freely giving up a self-centred perspective for an other-centred perspective grounded in love of God involves a sacrifice of tremendous significance. It is a profoundly difficult task for beings that come into the world with the kinds of desire we have to subordinate (and even renounce) satisfaction of certain desires to a loving commitment to do God's will. Mother Theresa is justifiably revered precisely because she was willing and able to subordinate selfish desires consistently over the course of her life to do God's will.
This seems true of romantic relationships. Whether infatuation, attraction, and romantic love are freely chosen does not seem relevant in assessing the moral quality of a romantic relationship; what matters in assessing the moral quality of a romantic relationship is how the two people freely treat each other. Freely chosen honesty, respect, affectionate gestures, and sacrifice for the well-being of one's partner seem to be what determines the moral value of a relationship. Even if romantic attraction and love are largely responses conditioned by factors beyond immediate volitional control, this would not obviously diminish the moral value of an enduring romantic relationship characterized by a freely chosen mutual commitment in which two people evince greater concern for the other's interests than for their own. If ordinary intuition harmonizes with the claim that freely chosen love is morally more valuable than love that isn't freely chosen, it harmonizes at least as well with the claim that whether love is freely chosen has little to do with the moral quality of a personal relationship.
In any event, there is a deeper problem here. To the extent that the claim that love has moral value can be grounded in ordinary moral judgments and intuitions, it gets us at most that, other things being equal, a world in which free beings love each other is morally preferable to a world in which free beings do not love each other. The judgments that it is morally good for P to do x and that P has a duty to do x imply only that a state of affairs in which P does x, other things being equal, is morally preferable to a state of affairs in which P does not do x; such judgments do not imply that a state of affairs in which P exists and can do x is, other things being equal, morally preferable to a state of affairs in which P does not exist.
Consider, for example, the duty to refrain from violence. This duty implies that, given any two moral agents A and B, the state of affairs in which A refrains from committing acts of violence against B is morally preferable, other things being equal, to a state of affairs in which A does not refrain from committing acts of violence against B. But this does not imply that a state of affairs in which there exist moral agents capable of refraining from violence is morally preferable, other things being equal, to a state of affairs in which moral agents do not exist. Just as the claim that it is good for moral agents to refrain from violence doesn't imply the claim that the existence of beings capable of refraining from violence adds moral value to the world, so too the claim that it is good for moral agents to love one another, whether freely chosen or otherwise, doesn't imply the claim that the existence of beings capable of love adds moral value to the world.
The most serious problem, however, is that love is not something that ordinary mortals have the ability to choose freely. Love is an emotion over which we have at best indirect volitional control. It is not possible for any of us to look at someone for whom we feel no love and will ourselves into that joyous emotional state whenever we encounter her. We can freely take steps to try to get to know her better and perhaps to begin to develop affection for her. But the intense emotion, love, is simply not the kind of thing that can be directly willed by anyone. Romantic, maternal, paternal, and Platonic love are emotions that happen to us, rather than are chosen. Now it might be true that the kind of love we are expected to direct to God is not merely an emotion; however, it seems reasonable to think that it consists, at least in part, of an emotional state, which might be different from the other forms of love we feel (as romantic and Platonic love are different) but is nonetheless an emotional state that bears sufficient family resemblances to these other forms of love to be plausibly characterized as ‘love’. Either way, love of God will contain elements over which we do not exercise direct volitional control.
The existence of subjects for God's love
The second version of the argument infers the moral value of free will from God's omni-benevolence (construed to include perfect lovingness). Since one characteristic associated with our capacity to love is that we actively seek out beings that we can love, it is reasonable to think that a perfectly loving God would desire a world in which there exist appropriate subjects of His love. But since it is considerably more gratifying to love a being capable of deliberative free choice than a being incapable of deliberative free choice (because the potential for meaningful interaction and reciprocity is greater in such a being), a perfectly loving God, other things being equal, would prefer to create a world in which there exist free beings who are appropriate subjects of God's love over a world in which such beings don't exist.
One might think that this version of the argument from love avoids the most serious problem with the last version of the argument. On this line of analysis, this argument is distinguishable from the last one in that it does not rely on the intuitively questionable claim that a state of affairs in which beings capable of love exist is, other things being equal, morally preferable to a state of affairs in which beings capable of love do not exist; as we saw in the last section, our ordinary moral intuitions and judgments do not seem to support this claim. Instead, this argument relies on the claim that, for any existing being B capable of love, a state of affairs in which B loves someone, other things being equal, is morally preferable to a state of affairs in which B does not love anyone. Insofar as this latter claim seems to be a general principle that would explain and justify the ordinary intuitive judgment that we ought to love our fellow beings, one might conclude that a state of affairs in which there exist appropriate free subjects of God's love is morally preferable to a state of affairs in which there do not exist appropriate free subjects of God's love.
Nevertheless, while the relevant principle is intuitively plausible, it does not imply that the existence of free subjects adds moral value to the universe. Most theists and conscientious non-theists would agree on the principle that if A is a being capable of love and B is another being, then the state of affairs in which A instantiates an attitude of love towards B is morally preferable, other things being equal, to the state of affairs in which A does not instantiate an attitude of love towards B. But this eminently intuitive claim does not imply the stronger principle that if A is a being capable of love, then a state of affairs in which there exists a B towards which A can instantiate an attitude of love is morally preferable to a state of affairs in which there does not exist a B towards which A can instantiate an attitude of love.
Again, the duty to refrain from violence implies that, given any two moral agents A and B, the state of affairs in which A refrains from committing acts of violence against B is morally preferable, other things being equal, to a state of affairs in which A does not refrain from committing acts of violence against B. But it should be clear that this latter claim does not imply the stronger principle that if A is a being capable of refraining from violent acts, then a state of affairs in which there exists a B against whom A can refrain from committing violent acts is morally preferable to a state of affairs in which there does not exist a B towards whom A can refrain from committing violent acts. Like the principle that posits non-violence as a moral good, the principle positing love as a moral good governs our behaviour towards beings that already exist and does not, for example, imply that the existence of the relevant class of recipient beings have the sort of moral value that makes the world a morally better place than it would otherwise be.
It is true, of course, that we regard as morally unfortunate any situation in which some person has no one to love, but this judgment is grounded in more fundamental concerns about human well-being. What we find particularly regrettable about a state of affairs in which any person wholly lacks companionship and loved ones is the unhealthy effects of such a deprivation. Among other effects, persons who continually lack companionship tend to experience more loneliness and unhappiness on a persistent basis; to become insensitive to human suffering; and to become bitter; indeed, it has been shown that persons who chronically lack meaningful companionship, on average, have a significantly shorter lifespan than persons who do not lack such companionship.
These concerns about human well-being, unlike the previous judgments we have considered, do seem to provide intuitive support for a principle that assigns greater moral value to a world in which a person has a loving companion than to a world lacking any potential companions for that person. It seems clear that a morally perfect being would prefer, other things being equal, to create a world in which there exist two human beings to a world in which there exists only one human being – if these are the only two choices. We need to love (and to be loved): it is vital to our happiness and to our ability to thrive, both psychologically and physically.
But these intuitive considerations commit us to a somewhat weaker principle than would support FWA. What is needed to support FWA is the principle that if A is capable of love, then a state of affairs in which there exists a B towards which A can instantiate an attitude of love is morally preferable to a state of affairs in which exists no B whom A can love. What the intuitive considerations above show is the principle that if A needs to love, then a state of affairs in which there exists a B whom A can love is morally preferable to a state of affairs in which there is no one A can love.
But this latter principle will not justify FWA because the antecedent of the conditional principle is not satisfied. While an all-perfect God clearly wants love, it would be problematic to characterize an all-perfect being as needing love. To the extent that God is, as is commonly thought, utterly self-sufficient, God cannot have any interests rising to the level of needs: God's well-being does not, unlike human well-being, depend on having loving companions. To assert otherwise is to make God's well-being dependent on the existence of human beings – a claim that does not appear compatible with the traditional conception of God as self-sufficient.
Although a loving God prefers a world with appropriate subjects for His love, this is not sufficient to show that the existence of such subjects adds moral value to the world. The problem is that the claim that a world in which free beings exist is preferable from the standpoint of a perfectly loving being that wants to love to a world in which such beings don't exist shows only, to put it awkwardly, that a world with free beings is lovingly preferable to a world without free beings. It does not imply that a world in which free beings exist is morally preferable (or preferable from the standpoint of a morally perfect being) to a world in which free beings don't exist.
The reason is that the demands of love and morality often diverge; what love asks of a person sometimes seems to conflict with what morality asks of a person. For example, the demands of love and morality seem to conflict with respect to punishment. According to one well-known argument, the concept of an all-perfect God is incoherent because the claim that a being B is morally perfect is inconsistent with the claim that B is omni-benevolent. On the one hand, B cannot be morally perfect without being perfectly just; and a perfectly just being always gives a person no more and no less than what she deserves in the way of reward and punishment. On the other hand, B cannot be omni-benevolent (construed to include perfect lovingness) without also being merciful; and a merciful being sometimes gives a person less punishment than she deserves. Thus, the argument concludes, it is logically impossible for a being to be morally perfect and omni-benevolent.
Debates about the doctrine of hell also suggest that moral perfection and omni-benevolence sometimes conflict. Proponents of the traditional doctrine of hell believe that all unsaved persons are eternally consigned to hell and that many persons will die without having been saved. To justify this doctrine, it must be shown that eternal consignment to hell of unsaved persons is deserved and hence consistent with God's moral perfection. But it must also be shown that the eternal consignment of some persons to hell is consistent with God's omni-benevolence. As Michael J. Murray describes the worry, even if unsaved persons deserve hell, we can all feel the ‘tug’ of the intuition that ‘a loving God would do whatever is necessary to prevent those who are destined to hell from going there’.Footnote 12
George Schlesinger argues that God possesses each of the perfections only to the extent that it enhances His excellence.Footnote 13 This opens up a couple of possibilities. First, one could argue that the extent of God's love is qualified by His moral perfection; on this line of analysis, God does not love in a way that would be inconsistent with His moral perfection and justice. Second, one could argue that the extent of God's moral perfection is qualified by God's perfectly loving nature; God's preference for morally sound states of affairs extends only so far as can be reconciled with his perfectly loving nature.
Either way, though, this offers no support for VT. Insofar as Schlesinger's view explicitly concedes that the demands of love and morality sometimes conflict, it entails that we can't validly infer the claim that a state of affairs S is morally desirable from the claim that S is lovingly desirable (i.e. desirable from the standpoint of a perfectly loving being). Regardless of whether God's loving nature is qualified by His moral perfection or conversely, just knowing that a perfectly loving being, other things being equal, would prefer to create a world where free beings exist to a world where free beings don't exist, by itself, does not tell us whether a morally perfect being, other things being equal, would prefer to create a world where free beings exist to a world where free beings don't exist – which is what is needed to vindicate VT. Love and morality might disagree on the value of free beings: while love might favour the creation of free beings, morality might be opposed to or, more likely, neutral with respect to the creation of free beings.
One might think that all that is needed here is the claim that a world in which God has a free being to love is morally preferable to a world in which God lacks a free being to love. The problem with this weaker principle is that it is satisfied in a world in which only God exists. The triune God is constituted in some mysterious way by three distinct divine persons with three distinct loci of consciousness: God the Father, Jesus, and the Holy Spirit. As is clearly evident from scripture, God the Father loves Jesus, the only Son of God, and conversely. Presumably, the same holds true with respect to relations between the Holy Spirit and God the Father, as well as between the Holy Spirit and Jesus Christ. Even in a universe without human beings, then, there are appropriate objects of divine love.
Here it is crucial to note that a universe in which only the triune God exists will contain love without any evil. The weaker principle described above cannot justify the creation of human beings because it is satisfied without adding evil to the world by a world in which only the triune God exists. Adding free human beings will add other subjects for God to love, but it also adds evil to the world. The weaker principle described above will not justify adding free human beings to the world because it is satisfied without the introduction of evil into the world. VT and FWA explain the existence of evil as being an evil that inevitably accompanies the greater moral good of having free beings in the world. The weaker principle cannot do this work because God does not have to allow evil in order to get the greater moral good of having appropriate subjects of His love. The argument from love fails to vindicate VT.
The value of free beings: the argument from resemblance
One might ground a defence of VT in the claim that human beings resemble God because they are free. The argument begins with the claim that a possible world in which an all-perfect God exists is morally preferable to a possible world in which an all-perfect God doesn't exist. What accounts for the fact that the existence of an all-perfect being adds moral value to the world is that it is in possession of all the perfections. Thus, it is the instantiation of these perfections that adds moral value to any universe in which there exists an entity that instantiates them. Since free will is one of the perfections that God instantiates, it adds moral value to any world in which God exists. Since moral persons instantiate the perfection of free will, a universe in which persons exist is, other things being equal, morally preferable to a universe in which persons do not exist.
Nevertheless, assuming that the existence of God adds moral value to the universe in virtue of God's instantiating free will, it is not clear that God has free will in even remotely the same sense that human beings have it. Free choice as it pertains to ordinary human contexts entails the ability to do otherwise. On this line of analysis, the claim that a person P freely performs an act a implies that P could have done other than a.
Compatibilists and libertarians disagree, of course, on what is involved in free will but neither account applies to God. Whether libertarian or compatibilist, the claim that an agent P freely does a at time t entails, at the very least, that there is a logically possible world in which P does other than a at t.Footnote 14 But classical theism appears to be inconsistent with the claim that God ever freely does anything in even this very weak sense. Insofar as it is necessarily true that God is morally perfect, it follows that there is no logically possible world in which God acts immorally. Further, insofar as it is necessarily true that God acts in accordance with an immutable nature that includes all the perfections, it seems to follow that there is no possible world, for any divine act d at time t, where God exists and does other than d at t. Assuming that it makes sense to think of the God of classical theism as having free will, it is of a qualitatively different kind than human free will. If this is correct, then one cannot soundly infer the claim that the sort of free will possessed by human beings has moral value from the claim that the sort of free will possessed by God has moral value.Footnote 15
More importantly, the argument from resemblance begs the question insofar as it assumes the very thing that is in question, namely that free will has moral value. Even if a universe in which an all-perfect God exists is, other things being equal, morally perfect to a universe in which an all-perfect God does not exist, it does not follow that God's instantiation of free will adds value to the universe. One might plausibly argue that God's existence adds value to the universe in virtue of God's simultaneously instantiating all the perfections; the existence of any lesser being does not add such value to the universe. On this line of argument, the divine instantiation of free will, in and of itself, does not add moral value to the universe. The argument from resemblance cannot vindicate VT.
The value of free beings: the argument from intrinsic value
A third argument for VT is grounded in a distinction between two kinds of value. A being, entity, property, or state of affairs has instrumental value if and only if it is valuable as a means to an end. In contrast, a being, entity, property, or state of affairs has intrinsic value if and only if it is valuable for its own sake or as an end-in-itself. Human beings have both instrumental and intrinsic value. We have instrumental value because we can be useful as means to secure other ends; thus, for example, a doctor has instrumental value as a means to the end of continuing good health. We also have intrinsic value because we have worth independent of any use to which we can be put and hence are valuable for our own sakes.
This distinction has powerful normative consequences. Things like money that have only instrumental value can permissibly be treated as objects or resources and hence can permissibly be used. In contrast, a thing with intrinsic value is a moral subject entitled to respect and is thus the beneficiary of direct duties. In consequence, it is impermissible simply to use a thing with intrinsic value. This does not imply it is impermissible to seek out instrumental value in human beings; rather, it implies only that it is impermissible to treat human beings as if they were nothing more than objects or receptacles of instrumental value. Even when one person A employs another B to benefit from B's instrumental value, A must treat B with respect for her dignity as a moral subject. Thus, on this line of reasoning, intrinsic value constitutes or gives rise to moral value.
The argument from intrinsic value derives the claim that a universe with free beings is, other things being equal, morally preferable to one without free beings from the claim that free beings have intrinsic value. Since free beings are intrinsically valuable, such beings add moral value to any universe in which they exist. Further, it is a conceptual truth that, given any two worlds ω1 and ω2, ω1 is morally preferable to ω2 if and only if ω1 has greater moral value than ω2. Since a world with free beings has greater intrinsic and hence moral value, other things being equal, than a world without free beings, a world with free beings is, other things being equal, morally preferable to a world without free beings. Thus, a morally perfect God would, other things being equal, prefer a world with free beings to a world without free beings.
The success of this argument depends on the plausibility of its underlying claims about intrinsic value – and, as it turns out, there is considerable controversy about such claims. While many philosophers accept that there are some things or states that are intrinsically valuable, they disagree about what those things or states are. Indeed, many utilitarian-minded philosophers regard pleasure as the only intrinsic value and pain as the only intrinsic disvalue in the world (hereinafter the hedonistic principle). As Mill famously put the point, ‘pleasure and freedom from pain are the only things desirable as ends; and … all desirable things … are desirable either for pleasure inherent in themselves or as means to the promotion of pleasure and the prevention of pain’.Footnote 16 Insofar as such philosophers are correct, it follows that the instantiation of free will by a being adds nothing of moral value to the world.
In response, one might try to reconcile FWA with the hedonistic principle in the following way. Presumably, rationality is a necessary (but not sufficient) condition of free will; after all, free choices are supposed to be related in a certain way to the result of agent deliberation and only rational agents are capable of deliberation. But, as Mill argued, rationality makes possible the experience of higher intellectual pleasures that are qualitatively different from the sensual pleasures of which even brute animals are capable, such as those associated with sex and eating. Thus, one might conclude that it follows that the existence of free beings adds moral value to the universe insofar as it makes possible the experience of such higher pleasures.
Even assuming the soundness of Mill's distinction between the two kinds of pleasure, this reasoning is unpersuasive. The problem is that the hedonistic principle doesn't imply that the existence of beings capable of pleasure constitutes an intrinsic good. For the claim that pleasure is intrinsically good assumes the existence of beings capable of experiencing it and hence implies that:
(1) A world with beings capable of experiencing pleasure that contains more pleasure is, other things being equal, intrinsically more valuable (and hence morally preferable) to a world with beings capable of experiencing pleasure that contains less pleasure.
But (1) does not imply:
(2) A world containing beings capable of experiencing pleasure is, other things being equal, intrinsically more valuable (and hence morally preferable) to a world that doesn't contain such beings.
The difficulty arises because there are no grounds under the hedonistic principle on which to discriminate between a world in which beings capable of experiencing pleasure exist but never experience pleasure or pain and a world in which such beings do not exist. Thus, even if the hedonistic principle is correct, it provides no reason to think that the existence of free beings, by itself, adds moral value to the world.
But disputes about intrinsic value are not limited to disagreements about which entities or states of affairs are intrinsically valuable; many philosophers reject altogether the idea that there is anything in the world that is intrinsically good.Footnote 17 Judith Jarvis Thomson considers a number of different ways to flesh out the content of the concept of intrinsic value and finds them all problematic; thus, for example, she rejects the idea that intrinsic value can be fruitfully defined in terms of ‘what a person … would value for its own sake if he or she were fully informed, free of neuroses, and assessing the matter in a cool hour … [because we can't] show that people really would not love the nasty under this constraint’.Footnote 18 Philippa Foot argues that we can't make sense of the notion of intrinsic value if defined as ‘a valuable state of affairs from an impersonal or moral point of view’, and concludes that there is nothing intrinsically valuable in a sense that is external to moral theory.Footnote 19
To deny there is anything intrinsically valuable in the world, of course, is not to deny the normative claim that human beings have ‘moral standing’ in virtue of being able to choose their behaviour freely, but this latter claim, by itself, implies only a conditional claim about moral standing. In particular, it implies only that:
(3) For every being x, if x's behaviour is freely chosen, then x has a moral claim to be respected as a moral subject and treated with dignity.
This conditional claim, however, falls well short of imputing the sort of value to free beings that makes a world a morally better place in virtue of containing such value. At most, (3) implies:
(4) A possible world in which free beings exist and are treated with respect is, other things being equal, morally preferable to a possible world in which free beings exist and are not treated with respect.
However, this doesn't imply that a possible world in which free beings exist is, other things being equal, morally preferable to a possible world in which such beings don't exist; as is readily evident, (4) asserts nothing about the moral value of a possible world that lacks free beings. And a claim that compares the moral value of two possible worlds in which free beings exist tells us nothing about the moral value of a possible world in which free beings exist relative to a world in which free beings do not exist.
What is ultimately at issue is not the claim that beings have moral standing in virtue of having characteristics like sentience, rationality, or free will. It is rather the idea that to have moral standing and hence a claim to decent treatment is a matter of instantiating the sort of value that can be aggregated to make the universe a better place in virtue of containing it. To deny this is not, however, to deny either that sentences expressing moral judgments are sometimes objectively true or that there exist moral properties that, so to speak, form part of the furniture of the universe. But it is to deny both that the sentence ‘the existence of free beings is an intrinsically good property of the universe’ is objectively true and that a being instantiates some real property of intrinsic goodness in virtue of having free will.
While the argument from intrinsic value is thus problematic insofar as it relies on contestable claims about the existence and nature of intrinsic value, there is a more fundamental problem with it. Even if one accepts the critical claims that intrinsic value exists and that human beings are receptacles of intrinsic value in virtue of having free will, it simply doesn't follow that the existence of free beings adds moral value to any universe in which they exist. It should be clear, for example, that the claim that human beings have intrinsic value in virtue of having free will doesn't imply that Earth, other things being equal, is morally preferable to Mars or that the Milky Way is morally preferable to some galaxy in which free beings don't exist. Indeed, assuming we can even make sense of this sort of judgment in the contexts of planets or galaxies, it seems pretty clearly implausible. But if having free beings doesn't make Earth morally preferable to Mars, it is hard to see why it would make this possible world morally preferable to one that lacks free beings.
Certainly, it seems reasonable to think that the claim that some thing has intrinsic value entails, as a matter of definition, the claim ‘there is a moral reason for preserving it for its own sake independently of whatever uses it serves’.Footnote 20 As we have seen, the claim that a being B has intrinsic value arguably entails that B is a moral subject that has value as an end-in-itself that ought to be respected by other moral agents. And that duty to respect B arguably entails a duty to preserve B's existence and life (which makes possible the exercise of its deliberative capacities).
But it is one thing to claim that a duty (or reason) to respect an intrinsically valuable being entails a duty to preserve it; it is clearly another to claim that the duty to respect an intrinsically valuable being entails a duty (or reason) to bring it into existence. After all, there can't be a duty (or reason) to respect an intrinsically valuable being until it exists and instantiates that value – and that is true no matter whom the duty is owed to. Non-existent things instantiate no properties and hence have no value whatsoever. Accordingly, if it is the instantiation of intrinsic value by a being that gives rise to a duty (or reason) to respect that being, then that duty (or reason) does not entail an obligation (or reason) to bring intrinsically valuable beings into existence. Whatever duty (or reason) there is to respect beings that instantiate intrinsic value, then, it provides no grounds for thinking that it is morally preferable to bring intrinsically valuable beings into existence.
One might think we have a duty to bring free beings into existence; after all, scripture commands us to be fruitful and multiply, and even agnostics and atheists tend to regard news of new pregnancies as good news. But, understood as a consequence of the argument from intrinsic value, it is implausibly strong. To the extent that we ought to bring free beings into existence because this brings intrinsic value into the world, this normative claim cannot be defeated by considerations having to do with instrumental value. Intrinsic value is a different kind of value than instrumental value; while a cost-and-benefit analysis is appropriate in deciding how to allocate purely instrumental value, it is not appropriate in making decisions about, so to speak, allocating intrinsic value (i.e. whether to bring it into existence, preserve it, or extinguish it). Insofar as rights protect intrinsic value and the infringement of a right can't be justified by considerations of instrumental value, it follows that considerations of instrumental value are utterly irrelevant with respect to decisions about whether to bring an intrinsically valuable being into the world.
This means, of course, that the claim that Jack and Jill ought to bring a free being into existence can't be defeated by instrumental considerations having to do with what sort of life they would be able to provide for themselves or their children. But if this is true, then ordinary moral judgments about postponing having children until one can provide them with a reasonable quality of life are simply wrong. One ought never to postpone introducing intrinsic value into the world for such instrumental considerations – and if such considerations wouldn't justify putting off a child, it is not clear what considerations would.Footnote 21 If these claims don't refute the key claims about intrinsic value, they certainly diminish their plausibility.
Conclusions
In this essay, I have argued that FWA fails as a theodicy in the absence of a plausible defence for the crucial claim that a world in which free beings exist is morally preferable to a world in which free beings don't exist. To address adequately the problems of evil, FWA must provide a plausible reason for thinking that a morally perfect God would allow the evil and suffering we experience in this world because there is no other way to secure the greater moral good of free will. But this requires an argument rooted in some reasonably uncontroversial moral principle that shows that the existence of beings with free will is, indeed, a moral good that adds value to a world.Footnote 22
