A New Defence against the Problem of Evil

Abstract

In this paper, I propose a defence against the problem of evil. This defence does not involve either free will or soul-making, but, rather, is intended as a replacement for the traditional theodicies. The defence will have two components: firstly, a proposal for why a good God would not intervene to eliminate the evil (natural or moral) in the world; and second, a proposal for why a good God would need to allow evil to exist in the world in the first place. I identify four desiderata for defences against the problem of evil and I argue that this new defence achieves all of these at least as well as traditional defences involving free will and/or soul-making.

1. Introduction

Why is there so much pain in the world? Why is there so much moral evil? Why so much natural evil? Do these facts show that there is no morally good, omniscient, and omnipotent God? If there is a morally good, omniscient, and omnipotent God, why do they not intervene to lessen the amount of evil? And if there is such a God, why did they choose to embed pain in the evolutionary process? All these questions are familiar ones, and in this paper, I will propose a new set of answers to them.

The answers I offer involve the following three key theses:

(1)

Humanity, collectively, has some feature φ, the having of which entails that, in general, it would be morally wrong for God to intervene in human affairs.

(2)

For any human, h, being human is an essential property of h.

(3)

The identity of any given species is, in part, fixed by its evolutionary prehistory (i.e., the evolutionary history leading up to the first generation of the species). It is metaphysically impossible for any species to have had a substantially different evolutionary prehistory from that which it in fact did have.

In this paper, I will argue that a defence of theism (by which I will mean the thesis that there is a morally perfect, omniscient, omnipotent God) built around (1)–(3), each of which I will argue is plausible, can answer both the natural and moral problem of evil without the need for any appeal to either the free-will or soul-making theodicies, thus avoiding the problems that come with these established theodicies.

2. Some Desiderata

Why bother with a new defence against the problem(s) of evil (generally, I will speak of the “problem of evil”, singular, but it will be important to what follows, that it comes in several different forms), when we already have so many? The answer is that defences against the problem of evil are subject to a number of desiderata, and at least arguably no position currently on the market can satisfy them all. In contrast, I think the novel defence I offer here can. The following are the desiderata for a defence against the problem of evil that I acknowledge:

(a)

The anti-Leibizian constraint: a successful defence against the problem of evil must explain how it can be that, although God’s actions are all morally optimal, we nevertheless do not live in the best of all possible worlds.

(b)

The Pauline constraint: a successful defence against the problem of evil cannot involve God in violating the “Pauline Principle”, namely, “do no evil that good may come of it”.

(c)

The efficiency constraint: a successful defence against the problem of evil must show that God could not have achieved her or his purposes without allowing the possibility of some equally bad or worse evil than that which actually exists.

(d)

The metaphysical parsimony constraint: a successful defence against the problem of evil must keep dubious metaphysical posits to a minimum.

The last of these desiderata differs from the others in being a matter of degree. The notion of “metaphysical dubiousness” is vague and any metaphysician is free to deny that there is anything dubious about their own posits. All I intend with (d) is to reflect the fact that one of the criteria by which an attempted defence against the problem of evil will be judged is the plausibility of its metaphysical commitments. Ceteris paribus, a defence against the problem of evil with more plausible metaphysical commitments, is to be preferred to one with less plausible metaphysical commitments.

The reason I am proposing a new defence against the problem of evil is that I think a plausible case can be made that the well-known free-will and soul-making theodicies both fail to satisfy at least one of (a)–(c). In the next two sections, I outline these concerns before proposing a new solution which I think does satisfy (a)–(c) and satisfies (d) at least as well as free will and soul-making.

3. Problems with the Soul-Making Theodicy

Let us start with soul-making. I think the soul-making theodicy is at its most plausible when combined with something like virtue ethics, the general idea being that some virtues or goods of the soul are of intrinsic moral worth and that certain of these can be acquired by the experience of evil. The virtues of perseverance, courage, patience, and moral rectitude all jump to mind. Of course, the intrinsic value of these virtues must outweigh the disvalue of the evil suffered to achieve them. Moreover, in order to satisfy desideratum (c), it will have to be said that only by actually experiencing evil are these virtues (or the highest degree of these virtues) attainable. Otherwise, it would be possible for God simply to give individuals these virtues without making them suffer for it, and the problem of evil could then be reformulated in this way: why does God make humans suffer for benefits she or he could have easily given them without allowing them to suffer?

It is debatable whether there are virtues (or certain degrees of virtues) that can only be had through experience. If virtues are just dispositions to manifest certain patterns of behaviour in specific circumstances, then whether these circumstances are ever realized is irrelevant to the issue of whether an agent has the given virtue. The defender of the soul-making theodicy will, therefore, have to deny that virtues are just dispositions, and if this denial cannot be independently motivated, their commitment to intrinsically valuable experience-only-attainable virtues places the satisfaction of desideratum (d) at risk. This is a central concern with the soul-making theodicy. Do we really have good reasons to think that there are goods of the soul that are worth suffering evil for and which are unrealizable, even by an omnipotent being, in some way that does not involve evil? Perhaps, perhaps not, but it is a worry for the soul-making theodicy and not the only one.

James Sterba’s recent work (Sterba 2019) raises a problem for many traditional defences against the problem of evil. According to Sterba, if God is a moral agent, he is bound by the Pauline Principle: do no evil that good may come of it. Sterba takes the general intuition expressed in the Pauline Principle, as just stated, to motivate three more specific moral requirements which any good God would have to meet. These three requirements, all of which focus on moral evil, are as follows:

Evil Prevention Requirements

• Prevent, rather than permit, significant and especially horrendous evil consequences of immoral actions without violating anyone’s rights (a good to which we have a right), as needed, when that can easily be achieved.
• Do not permit significant and especially horrendous evil consequences of immoral actions simply to provide other rational beings with goods they would morally prefer not to have.
• Do not permit, rather than prevent, significant and especially horrendous evil consequences of immoral actions on would-be victims (which would violate their rights) in order to provide them with goods to which they do not have a right, when there are countless morally unobjectionable ways of providing those goods (Sterba 2019, p. 184).

I think that the Evil Prevention Requirements (I)–(III) are all true moral principles: no perfect moral agent violates them.1 In any case, it seems pretty clear that if the only reason that God allows a great deal of moral evil to befall humans is in order for them to build moral characters, then God is in violation of Evil Prevention Principle I. Thus, a defence against the problem of evil which depended on the soul-making theodicy alone would not satisfy desideratum (b), at least not if Sterba is right in deriving Evil Prevention Requirement I from the Pauline Principle, as I think he is.

Of course, soul-making is often assumed to work in conjunction with the free-will theodicy, to which we now turn.

4. Problems with the Free-Will Theodicy

The free-will theodicy, of course, suggests that God might allow evil in order to allow humans to have free will. For any defence against the problem of evil which encompasses the free-will theodicy to satisfy desiderata (c), it must be the case that God could not have simultaneously made humans free and at the same time created humans who were predetermined to behave virtuously. The sort of free will that the theodicy requires is, therefore, of a libertarian kind. As with soul-making, there is a trade-off here between the satisfaction of desiderata (c) and the degree to which desiderata (d) is satisfied. In order to satisfy (c), more controversial metaphysical commitments must be taken on. This must be kept in mind in evaluating the traditional theodicies against the defence I will offer shortly.

The free-will theodicy, on the face of it, does enjoy one major advantage over the soul-making theodicy. The advantage is that it clearly satisfies desideratum (a), in a way that soul-making does not. It satisfies (a) because if human free will is the reason for the amount of evil that there is in the world, then there is an explanation for how God’s choices may all have been optimal (in the sense of directly resulting in only as much evil as is necessary for bringing about even greater goods) and yet there be worlds which are better than the actual one. Specifically, there are worlds in which God’s actions are just what they are in the actual world and in which human free choices cause less evil than they do in the actual world. On the other hand, the free-will theodicy seems to account only for moral evil, not natural evil.2 This, once again, speaks in favour of running the two theodicies together, and we must return to the question of whether this allows for the satisfaction of desiderata (b).

I believe that even when run together, the traditional theodicies (assuming they are intended as a complete explanation for God’s allowing evil) would still involve God in violating the Pauline Principle. Far from disarming the Evil Prevention Requirements, I think that introducing the free-will theodicy only serves to make the threat they pose worse, because whereas a defence based solely on the soul-making theodicy would seem to conflict with Evil Prevention Requirement I, a defence which also encompasses the free-will theodicy would seem to conflict with Evil Prevention Requirements I and II.

Evil Prevention Requirement II does not seem to be in obvious tension with the soul-making theodicy. It says that a perfectly moral agent would not allow evil consequences of immoral actions purely for the sake of benefitting some other rational being (i.e., apart from the subject of the evil consequences) with goods that that being would morally prefer not to have. In characteristic applications of the soul-making theodicy to explain instances of evil, it is the subject of the evil consequences who is benefitted (e.g., with the opportunity to build a virtuous character), and so the tension here is not apparent. It is otherwise with the free-will theodicy. The benefit of free will is, in at least many possible applications of the free-will theodicy, enjoyed by the perpetrator of moral evil, rather than the victim. The idea is that God allows rational beings to harm others because God wants the former to enjoy free will. The problem is that it is not difficult to imagine cases, and surely some such cases actually occur, where a perpetrator would rather their free will be violated in order that they are prevented from some immoral action. Indeed, most repentant evil-doers, whose evil-doing leads to horrendous consequences for others, would likely rather that the good of free will had been denied them in the cases where it led them into these occasions of evil-doing. Many drink drivers who cause accidents no doubt genuinely wish that someone had intervened to prevent them from getting behind the wheel. Moreover, surely this desire is an entirely moral one. Given that there are plenty of repentant evil-doers whose evil-doing led to horrendous consequences, it would seem that God frequently ensures that people who would rather be denied free will (at least in the relevant occasions) are, nevertheless, given free will even though it leads them to act in way that results in horrendously evil consequences for others. This is a violation of Evil Prevention Principle II.

Sterba’s work has resulted in a large number of responses, and I do not intend to engage with the possible objections that can be made. My view is that Sterba’s arguments do demonstrate that there is a serious prima facie problem for the traditional theodicies. It is for this reason that I would like to propose an alternative to the free-will and soul-making package, one which, I will argue, enjoys the advantages of that package (including the satisfaction of desiderata (a) and (c)), while also satisfying desideratum (b), which the free-will and soul-making package does not satisfy. I will leave it to the reader to consider how well what I offer here satisfies desideratum (d), but I think it does so at least as well as the traditional theodicies do.

5. Human Self-Determination3

I now offer a replacement for the free-will and soul-making theodicies. I will split the exposition of this replacement into three parts. In the first part, I will defend the plausibility of the aforementioned thesis:

(1)

Humanity, collectively, has some feature φ, the having of which entails that, in general, it would be morally wrong for God to unilaterally intervene in human affairs,

and I will argue that if this thesis is true, then the problem of moral evil is dissolved. In the second part, I will argue that this same thesis also takes the force out of the problem of natural evil, provided that some explanation is to be had, consistent with the assumption of theism and the divine omni-properties, for why the very possibility of evil exists in the first place. In the final part, I will reintroduce the aforementioned theses (2) and (3) to argue that such an explanation is available.

We start, then, with thesis (1) and the problem of moral evil. The inspiration for (1) comes, perhaps ironically, from Sterba’s own work and, more particularly, from the “just state” analogy that he repeated appeals to in order to motivate the applicability of the Pauline Principle to God. Sterba asks us to consider the following:

The analogy of a political state that is aiming to secure a significant degree of justice for its members. Such a state would not try to prevent all the moral evil that occurs in its domain, even if that were within its power to do so. Instead, a just state would focus on preventing significant and especially horrendous moral evils that impact people’s lives … Similarly, God, like a just state, should be focused on preventing (not permitting) just the consequences of significant and especially horrendous moral evils which impact on people’s lives.

(Sterba 2019, p. 51)

Naturally, Sterba’s “just state” analogy has been the source of many objections (Hasker (2021) and Reichenbach (2022) are two among many who have objected), and, indeed, I agree that the analogy between God’s relationship to humans and the relationship of a state to its citizens breaks down. However, I also think that, in some important ways, God is like a sovereign authority.

First let us note some problems with Sterba’s analogy. There are any number of disanalogies between God’s relationship to humans and the relationship of a state to its citizens, but what is key to recognize is that it is in the very nature of a just state that part of its business is to play a role in preventing both moral and natural evil, and acting for the benefit of its citizens who are impacted by both moral and natural evil. It is not merely a consequence of a just state’s being a moral agent that it engages in these activities; it is quite literally part of the definition of statehood that these responsibilities are part of its role. Now, to simply assume that the same goes for God is, in effect, to beg the question against the theist. Theists do not conceive of God as a political authority responsible for preventing crime and responding to natural disasters. For this analogy to serve its intended purpose, Sterba would first need to show that a morally good God ought to be involved in these sorts of human-benefitting activities. I do not believe there is a good reason to think that this is so, however. I intend to appeal to thesis (1) to support this opinion.

Let us note that our intuitions about human-benefitting interventions vary considerably depending on the nature of the intervening agent. Imagine an attempted murder that is witnessed and then prevented by a passing moral agent. If the intervening agent is a human being, intervening to prevent a murder would normally be praiseworthy. Indeed, if the intervening agent were an alien from another galaxy, the intervention would still likely strike us as praiseworthy. But these are not the only sorts of possible moral agents. If the intervening agent were a nation state, then our intuitions about the intervention are likely to become considerably more complicated. If a foreign state intervenes to prevent murders within the borders of its neighbours, without the prior agreement of the local government, then many will feel at least in some circumstances (for example, where the local government is legitimate and at least somewhat functional and opposed to foreign interference) that the intervention is a morally reprehensible violation of sovereignty.

As I said above, I think that, in certain respects, Sterba’s just state analogy is fair. Like a state, God judges and condemns, God is the source of laws, and God wields power well in excess of that of individual humans. Where I think Sterba’s analogy goes wrong is in assuming that, because there are points of analogy between God and a state, a further analogy can thereby be sustained between God’s relationship with humanity and a just state’s analogy with its citizens. But the relationship between a just state and its citizens depends on some features which are noticeable absent in the from the relationship between God and humanity. For one thing, under the influence of Locke and Rousseau, it is still the dominant view in political philosophy that the legitimacy of a governmental authority depends on the consent of the governed (Nozick (1974) and Simmons (2001) are examples of recent Lockean accounts of legitimacy). In short, granting that an analogy can be drawn between God and a just state, is there any more reason to think that the relationship between God and humanity is more like the relationship between a just state and its citizens than the relationship between a just state and the citizens of an independent state? It is clear that the usual grounding for the relationship between a just state and its citizens does not exist between God and humanity. In light of this, I suggest that an analogy between God and a (just) foreign state is at least as well motivated as Sterba’s analogy. But, as we have seen, intuition (and, for that matter, much political theory) tells us that states are subject to moral restrictions on their interventions respecting foreign citizens. It is wrong, in many cases, for a state to intervene for the benefit of citizens of a foreign country. Thus, I think that the analogy between God and a just state offers as much support to (1) as it does to the claim that a good God would be in violation of the Evil Prevention Requirements.

If (1) is true, then God should not, in general, intervene either to prevent moral evil or to prevent, respond to, or manage the risk of natural evil. This, on its own, would be sufficient to answer the problem of moral evil and go some way to answering the problem of natural evil, but it does not go the whole way. It does not go the whole way to answering the problem of natural evil because it does not explain why there is natural evil in the world in the first place. Moral evil is explicable simply because humans choose for it to be there. But God, whatever his or her role is in the world, created a world with nociception and noxious stimuli. Why? Is this something a good God would do? In what follows, I propose an answer to this question based on (2) and (3) above.

6. Essences and Evolution

I will now suggest an answer to the question “why would a good God create a world with natural evil in it?”. I will not try to defend the answer I propose in detail, which would take considerably more space than is available here, given that the issues turn on quite deep topics in metaphysics and philosophy of science. My aim, instead, is simply to open a new avenue for exploration. That being said, I think the answer I propose has the virtues of simplicity and neatness and there is much to be said for it. The answer I propose to the question “why would a good God create a world with natural evil in it?” is, in brief, that God created such a world with evil in order to create humans. The idea is that even an omnipotent God could not have created humans without creating evil. But, more particularly, I propose that the humans, for whose sake God had to create a world with evil in it, are such that they had to be human beings if they were to exist at all, and, if any humans were to exist at all, it could only be because the human species was generated by an evolutionary process which included natural evil. Let us look at this proposed answer in a bit more depth.

So, the key claims, once again, are these:

(2)

For any human, h, being human is an essential property of h.

(3)

The identity of any given species is, in part, fixed by its evolutionary prehistory (i.e., the evolutionary history leading up to the first generation of the species). It is metaphysically impossible for any species to have had a substantially different evolutionary prehistory from that which it in fact did have.

Claim (2) is a popular position among contemporary metaphysicians. “Human” is often thought of as a paradigm case of a “substance sortal” (as opposed to a “phase sortal” like ‘boy’ or “adult”), a notion which is often tied to the thesis of individual essentialism. Individual essentialism, the view that individuals have essential properties, saw a revival from the mid-1970s onwards, connected with the popularization of quantified modal logic and the view that proper names are rigid designators. One obvious candidate for an essential property of an individual is its species. A human cannot have been anything other than a human, in this view.

Now, individual essentialism has its opponents, as does the view that “human” is a substance sortal, but it also has many contemporary defenders (see Kripke (1980), Salmon (1981) and Forbes (1985) for influential arguments in its favour), and that is sufficient for my current purposes. Once again, I am only intending to develop a plausible account; attempting to give a thorough defence of all its component parts will have to wait for future research.

We turn, then, to (3). This will require a little more explanation. It is a commonplace that species are units of evolution. There is significant disagreement in biology and philosophy of science over the nature of species, but that much is agreed on by most of parties to these debates. Beyond this, another majority, but not consensus, view is that the thesis of species essentialism is false. Species essentialism, which is quite independent of the aforementioned individual essentialism, holds that for some species there are nontrivial properties which, as a matter of necessity, are such that any member of that species has those properties. The thesis had previously been very popular through much of the history of science, as evidenced by the many attempts to define various species by the properties of their members. Thus, humans have been said to be essentially rational animals, for example. The historical popularity of species essentialism notwithstanding, it is currently overwhelmingly unpopular. The reasons for this unpopularity need not delay us here (see Hull (1965) for some influential arguments). Similarly, we will not consider the arguments of current defenders of species essentialism. What is important for our present purposes is what the rejection of species essentialism has meant for our best theories of the concept of a species.

The rejection of species essentialism led to a renewed interest in the metaphysics of species. The traditional view is that species are natural kinds. Natural kinds are kinds that play a role in the laws of nature. From the 1970s on, this view was attacked by those who argued that species are individuals rather than kinds (beginning with Ghiselin (1974) and Hull (1978)), a view which continues to be very popular. One of the main motivations for the species-as-individuals view is that it is claimed by its defenders to better accord with biologists’ notions of species. In particular, it is claimed that biologists implicitly take species to be “historical” entities, in that historical facts are part of the identity conditions for any given species. As David Hull states the point:

One final parallel between organisms and species warrants mentioning. Organisms are unique. When an organism ceases to exist, numerically that same organism cannot come into existence again. For example, if a baby were born today who was identical in every respect to Adolf Hitler, including genetic makeup, he still would not be Adolf Hitler. He would be as distinct and separate a human being as ever existed because of his unique “insertion into history,” to use Vendler’s propitious phrase… But the same observation can be made with respect to species. If a species evolved which was identical to a species of extinct pterodactyl save origin, it would still be a new, distinct species.

(Hull 1978, p. 349)

Therefore, defenders of the of the species-as-individuals view hold that species are historical entities in the above sense, and further hold that the compatibility of these two positions speaks in favour of the former. Not everyone has accepted the arguments for the species-as-individuals thesis (e.g., Reydon 2003; Crane 2004; Crawford 2008), but even among those who reject the thesis, there is general recognition that historical properties, like lineage, must play a role in the metaphysics of species. Therefore, for example, whereas traditional essentialists had focused on intrinsic properties as the essential characteristics of members of a given species, contemporary essentialists look to relational properties such as being descended from a, where a is some common ancestor of the species (for example, Okasha 2002). In short, there is at least something of a convergence, towards recognizing the importance of species’ history in the metaphysics of species.

Similarly, there is at least some degree of convergence among the various sides of the debate over the status of species names. Specifically, it is a popular view that species names are rigid designators (Crane (2004) argues that taking species names as rigid designators offers some support to the species-as-individuals view, but notes that many philosophers of science who reject that view nevertheless think that species names are rigid), that is, that they pick out the same thing in every possible world in which they pick out anything at all. The significance of this broad convergence of views is that the views held in common by both many defenders of species-as-individuals and their opponents provide support for (3), or so I shall argue.

If species names are rigid designators, then every species must have some metaphysically necessary properties which ground their transworld identity. And if species are historical entities, it would be natural to look to the historical features of species to find these properties, especially if traditional essentialism is rejected so that being a member of a species does not guarantee the having of some particular intrinsic property. We may illustrate with Hull’s earlier example: the same species cannot reappear later in history; rather, a new set of organisms can evolve which are characterized by the same genetic make-up as the members of the now-extinct species. This would constitute the emergence of a new species, not the reappearance of a disappeared species. This is because the new species would lack some essential historical property that the original species had. What is this property? Presumably species origin, for example, being descended from some particular first generation. In other words, perhaps the human species is, as a matter of metaphysical necessity, just the lineage (or set of individuals) that descends from the first generation of humans. This is a plausible starting point, but there is just one problem with taking this to give us the whole of the identity conditions for the human species though. It leaves this question unanswered: What are the identity conditions for that first generation? A popular view with which I agree, is that any individual organism must have had the parents it actually had. Assuming that is right, we can add that to our identity conditions for the human species that it must have been preceded by the specific set of (nonhuman) parents it was in fact preceded by. But this is not all, either.

Although, plausibly, it is necessary that each of the original humans had the parents they in fact had, it is not necessary that those parents had the offspring they had. They might have had no offspring or different offspring. Indeed, if the evolutionary process had been somewhat slower than it was (which is presumably possible), then they might have had nonhuman instead of human offspring. Surely there are metaphysically possible worlds in which our nearest nonhuman ancestors were followed by further generations of nonhumans. Given this possibility, the question is, in virtue of what, were the first humans human? We could, of course, simply refuse to answer the question, maintaining that humanity is a matter of either being descended from this first generation or being a member of this first generation and leave it at that, but I suggest that we want some explanation for why some possible offspring of our most recent nonhuman ancestors are human and others are nonhuman beyond mere stipulation. I propose the following answer: our earliest human ancestors must, as a matter of metaphysical necessity, have had the parents they actually had and that the evolutionary process by which our earliest human ancestors were generated by our most recent nonhuman ancestors be, largely, what it in fact was.

To state the final point in another way, if our most recent nonhuman ancestors were subject to very different evolutionary factors, such that very different traits were selected for in their offspring, then the resulting generation would not have been human. That, at least, is what I am suggesting. This, of course, is just thesis (3). Humans had to have had the immediate evolutionary prehistory that they actually had, not just having the same nonhuman ancestors, but also being subject to the same broad evolutionary forces. There could have been some small differences in the evolutionary story that led from nonhumans to humans, but no major differences. I cannot say exactly how much difference is allowable and the thesis will have to remain vague (though the allowable differences will need to be minimal enough to support our intuitions about which possible offspring of our most recent nonhuman ancestors count as human and which do not). Nevertheless, on this basis, I think that (3) is a plausible thesis that accords well with our best metaphysics of species, as it offers an explanation of why the first generation of humans should be counted as human, when other possible offspring of the same parents, intuitively, would not.

7. In Summary

In the previous section, I offered (2) and (3) as plausible metaphysical theses. What do they have to do with the problem of evil? In short, (2) and (3), taken together, explain why it would be necessary for a God to allow evil into the world in the first place, assuming that such a God was determined to bring about the existence of the human species or, indeed, determined to bring about the existence of any one or more of the humans who have or will actually exist.

If (3) is true, then there could not have been humans if the evolutionary process which, in fact, led to their generation was substantially different. It is well known that pain plays an important evolutionary role. The evolutionary story of the origin of humans would be very different indeed if it did not feature pain, and, thus, according to (3), this alternative evolutionary story is metaphysically impossible. Therefore, for example, although God could have created ex nihilo a species of sentient beings that were genetically identical to humans in every respect, it is metaphysically impossible for God to have created the human species in this way. The only way to create the human species is via the evolutionary process that biologists study. Thus, the existence of pain (and therefore evil) is a necessary cost that must be paid, even by an omnipotent creator, for the existence of humans.

Moreover, according to (2), every human could only have been a human. Thus, it is impossible for any one of us to have existed without this same evolutionary process, complete with pain, having been realized. The existence of pain in the evolutionary story leading up to humans was necessary for the existence not just of the species, but also of each and every human, if (2) and (3) are true.

8. Conclusions

The proposed solution to the problem of evil offered in this paper is intended as an alternative that avoids the necessity of appealing to either the free-will or soul-making theodicies. It involves two central components: first, a defence against the claim that a good God would have to intervene to prevent evil (natural and moral); and second, an explanation for why a good God might allow evil to exist in the first place. The general idea of the first component was that humans might have some feature which makes it the case that a good God would not intervene in human affairs without their consent. Problem of evil arguments, like Sterba’s, fail to acknowledge this possibility and offer no argument that this possibility is ruled out or even unlikely.

Indeed, I have suggested that Sterba’s own analogy between God and a just state offers us a plausible model of the sort of human feature that might render divine intervention (in at least some circumstances) immoral, namely, where we take humans to be analogous to the citizens of a foreign country. On this model, humans would have a right to self-determination, which a good God would not violate. Moreover, this also satisfies desiderata (a) without the need to appeal to any particular notion of contra-causal freedom. If a good God would not violate human self-determination, then, of course, it is possible that optimal divine actions do not guarantee that the best of all possible worlds is realized. Human actions will be suboptimal, but a good God would not intervene to prevent this. Thus, I think my account satisfies desiderata (a). Moreover, this would not involve God in violating any of the Sterba’s Evil Prevention Requirements, or the Pauline Principle more generally (at least not under any plausible interpretation of the principle). In this story, God is not allowing humans to harm others so that the former can enjoy some benefit, like having free will; rather, humanity has a collective right which a good God would not violate, thus desiderata (b) is similarly satisfied.

If this story is right, if there is some moral constraint on divine interventions, then (c) plausibly follows as well. Unless, that is, God could have set up the world in such a way that violations of the right to self-determination were unnecessary to make the world better than it is. An obvious question here is whether God could have simply created the world without the possibility of evil. The second component of the defence answers this worry. The general idea here is that, as a matter of metaphysical necessity, a world with humans must be a world with evil, specifically with pain. This depends on a metaphysical claim about individual humans (that they are necessarily human) and a metaphysical claim about the human species (that it could not have existed without the evolutionary prehistory that it in fact had, which includes the presence of pain). If these claims are true, then even an omnipotent God could not have created any one of the humans who has ever existed without introducing evil into the world in the first place.

This just leaves (d). Is the existing of something like a collective human right to self-determination plausible? How about the metaphysical theses (2) and (3)? I think these are all plausible enough but defending them would require much more work than can be undertaken here. I leave it for future discussion. Similarly, I will leave for future discussion a number of other questions that arise from the defence proposed here. For example, although the above might explain why pain was metaphysically necessary in the evolutionary prehistory of humanity, why would a good God choose to allow pain to remain after humans came into the world? Moreover, do we not have evidence that there was more pain in the evolutionary prehistory of humanity than was metaphysically necessary for the generation of the species?4 Finally, if the creation of humans metaphysically necessitated pain, and there are other metaphysically possible human-like species whose creation did not necessitate pain, would a good God not create the latter, rather than the former?5 Each of these is an interesting question, but they will have to wait.