The Problem of Evil, Skeptical Theism and Moral Epistemology

Abstract

This paper argues that skeptical theism isn’t susceptible to criticisms of the view presented in James Sterba’s new book on the logical problem of evil. Nevertheless, Sterba’s argument does serve to underscore the unpalatable moral-epistemological consequences of skeptical theistic skepticism (STS): for precisely the reasons that STS doesn’t succumb to Sterba’s critique, STS threatens to undermine moral knowledge altogether.

1. Introduction

I will use the term ‘skeptical theist’ in referring to one who embraces both theism and skeptical theistic skepticism; and I will define ‘skeptical theistic skepticism’, or ‘STS’, as the conjunction of the following skeptical theses:

• ST(a)    We shouldn’t think that the possible goods that are known to us are (probably) representative of those there are; and
• ST(m)  We shouldn’t think that the entailment relations we believe to obtain between possible goods and the prevention or permission of possible evils are (probably) representative of those there are.1

Where ‘E’ stands for some paradigmatic instance(s) of inscrutable evil, like the abduction and murder of a child (or an aggregate of likewise inscrutable evils), skeptical theists claim that ST(a) and ST(m) undermine inductive inferences of the form:

• We don’t know of any morally justifying reason for God to permit E.
• So there probably isn’t any morally justifying reason for God to permit E.

Although skeptical theists intend STS to undermine inductive arguments from evil—in particular, those that hinge on an inductive inference like the move from (1) to (2)—Sterba insists, plausibly, that he must contend with STS since the same skeptical strategy might be directed against his own account.2

The thrust of STS vis-à-vis the argument from evil is this. We shouldn’t affirm an inference like the move from (1) to (2) unless we have reason to be confident that the relevant goods and entailment relations that are known to us are representative of those there are. But in light of STS, we shouldn’t think that the relevant goods and entailment relations that are known to us are representative of those there are: for all we know, there’s some good that we don’t know about, such that its realization is logically incompatible with God’s prevention of E. Thus we shouldn’t affirm any inference like the move from (1) to (2).

2. Sterba’s Reply to STS

Sterba’s primary quarrel with skeptical theist’s account is this. If God’s permission of E were justified by the kind of good to which STS alludes, then God would have to be far less powerful than an average human—which is totally implausible. As Sterba argues,

Here we are dealing with situations where we lack the causal power to prevent the evil consequences of both immoral actions, and we appeal to that lack of causal power to justify why we permit the lesser evil consequence to prevent the greater evil consequences. Now, for just such situations, we are imagining that it is logically impossible for God to prevent the consequences of both immoral actions that are just causally impossible for us to prevent. Right off, that would make God impossibly less powerful than ourselves.3

Since logical impossibility is far more stringent than causal inability, the notion that it might be logically impossible for God to prevent E seems to imply that God is far less powerful than humans, who are merely causally unable to prevent E. This, argues Sterba, is wildly implausible.

I agree that it’s implausible to suppose that, in certain cases, it may be logically impossible for God to prevent a single instance of evil that humans are merely causally incapable of preventing—e.g., the abduction of a child. But I don’t think that’s really an implication of STS. We might put the skeptical theist’s point this way.4 For all we know, there’s some good, G, such that the value of realizing G outweighs the disvalue of allowing E. And for all we know, G stands in a logical relationship to E that makes it impossible for anyone (God or human) to prevent E in any state of affairs in which G is realized. So it’s logically impossible for God (or anyone else) to prevent E in any state of affairs in which G is realized. Thus, with respect to the logical possibilities surrounding the realization of G vis-à-vis the prevention or permission of E, God’s power is equal to ours—just as God’s power is equal to ours with respect to the inability to make 2 + 2 amount to something other than 4. As it happens, humans are also causally incapable of preventing E (let’s say). God has the ability to prevent E (causally speaking); but God chooses to refrain from preventing E, in favor of realizing G, given that the realization of G is logically incompatible with the prevention of E. In terms of causal powers, then, God is more powerful than we are. However, since it’s logically impossible for anyone to prevent E in a state of affairs in which G is realized, God lacks the ability to both prevent E and preside over a state of affairs in which G is realized. The inability to do logically impossible things is a standard limitation on the power of even an omnipotent being. So the scenario I’ve just described is not only consistent with God’s being vastly more powerful than we are, it’s entirely consistent with divine omnipotence.

Perhaps Sterba would reply that a good like I’ve just described would be truly foreign to us. But this is precisely the skeptical theist’s point: we shouldn’t think the possible God-justifying goods or entailment relations that are known to us are representative of those there are. So, for all we know, there’s some good or entailment relation that’s unknown to us—perhaps a good or entailment relation that’s totally unlike the goods or entailment relations that are known to us—such that God’s failure to prevent E would be morally justified by that good. So I don’t think that Sterba’s answer to STS succeeds. Nevertheless, as I’ll now argue, we should reject STS altogether in light of its moral-epistemological consequences. So I agree with Sterba’s assessment that STS isn’t a problem for his argument, even though I disagree with his reasons for believing this to be so.

3. The Moral-Epistemological Implications of STS

It is a matter of some controversy whether or to what extent God would be morally motivated to realize sentient flourishing generally or human flourishing in particular.5 Nothing in my argument hinges on the outcome of that debate. That said, we’ll need a term to describe whatever it is that God would be morally motivated to realize—whatever that may be—and, given a few qualifications, ‘flourishing’ strikes me as an eligible candidate. So let ‘S’ be any human; and let ‘F’ be any (aggregation of) feature(s) of S’s life, such that: ceteris paribus, by virtue of God’s moral perfection, God would be motivated to realize F.6 If and insofar as F is realized in the life of S, we will say that S flourishes. Note that it needn’t be the case that S or any other human recognizes the flourishing of S as such in order for it to be the case that S flourishes. For all I intend to say about flourishing, S and every other human might fail to recognize that S flourishes either because we fail to recognize that some known feature of S’s life constitutes a form of flourishing, or because we fail to recognize that some known form of flourishing is realized in the life of S, or both.

I now argue that the skeptical theist is in no position to deny that God (exists and) has realized a morally optimal pattern of flourishing in the actual world. And if we shouldn’t deny that God has realized a morally optimal pattern of flourishing in the actual world then we should embrace skepticism about the possibility of moral knowledge.

My argument begins with a fact, a conjecture and an implication. The fact that I have in mind is this. Every single day, according to relatively recent estimates, roughly 29,000 children under five years of age perish for want of life-sustaining necessities like food, shelter and basic medical remedies.7 We will call this The Fact.

Here I should preempt two potential objections that might otherwise appear down the line. First, humankind’s ability to prevent most of the suffering implicated in The Fact would do no more to diminish God’s culpability in permitting it than God’s ability to prevent all of this suffering would diminish humankind’s culpability in allowing it to continue. Second, I will not address the assertion, even should anyone be so bold as to make it, that the agonizing deaths of more than 10.5 million children per year falls within the range of that which, relative to God’s moral situation, would be a matter of indifference.

Our conjecture calls for a couple of terminological conventions. Let Set PW be the set of all possible worlds. Imagine that each member of Set PW has a value, and that the value of each member of Set PW corresponds to its level of per capita human flourishing: The gross quantity of human flourishing in that possible world, divided by its total human population. Suppose that the values of all the members of Set PW were added together, and call the sum T. Finally, if T were divided by the population of Set PW, the result would be the mean level of per capita human flourishing across all possible worlds, or ‘μ’. For ease of expression, we’ll stipulate that possible worlds with a level of per capita human flourishing more than n standard deviations above μ are in a subset of Set PW that we will call possible worlds with the most per capita human flourishing.

Now, for all we know, possible worlds with the most per capita human flourishing are ones in which relatively few humans have the opportunity to flourish a great deal, while many humans flourish only some and many others flourish not at all. We might say that such worlds would have an undistributed arrangement of human flourishing. At the opposite extreme, for all we know, the possible worlds in which human flourishing is distributed absolutely equally may be ones in which all humans flourish some but no human flourishes a great deal, and the level of per capita human flourishing is far less than it might otherwise be. We might call this a distributive arrangement of human flourishing. For present purposes, we needn’t speculate about the extent to which God, qua morally perfect person, would aim to bring about a distributive or undistributed arrangement of human flourishing—or whether, qua omniscient, omnipotent being, God’s agency would be limited by the sorts of tensions implicated in these epistemically possible extremes. Instead, we will stipulate that an arrangement of human flourishing is morally optimal if it strikes a morally appropriate balance between distributive and undistributed human flourishing (if there is such a balance, whatever it may be).

The conjecture runs as follows. It seems plausible to suppose that God would have an interest in the flourishing of sentient creatures; and it doesn’t seem totally unlikely that bringing about human flourishing in particular would be among God’s highest priorities. One might even suppose that a person like God would bring about the most human flourishing that is logically consistent with a morally optimal arrangement of human flourishing.8 A subtler, slightly stronger form of this supposition is the view that God would bring about a morally optimal arrangement of the most human flourishing that is logically consistent with a morally optimal arrangement of human flourishing. For ease of reference, I will use the phrase optimal pattern to describe a morally optimal arrangement of the most human flourishing that is consistent with a morally optimal arrangement of human flourishing. And I will refer to the assertion that ‘God would bring about an optimal pattern of human flourishing’ as The Conjecture.9

Two clarificatory remarks are in order here. Note first that The Conjecture is subjunctive: God would—i.e., if God exists—bring about an optimal pattern of human flourishing. So embracing The Conjecture doesn’t entail a commitment to theism. Second, The Conjecture should be read to imply that God would bring about an optimal pattern of human flourishing over any (aggregation of) good(s) the realization of which is incompatible with an optimal pattern of human flourishing. In other words, The Conjecture claims that no good is such that God would realize that good rather than realizing an optimal pattern of human flourishing. We will revisit these matters in more detail below.

Finally, the implication unfolds in the following way. Let α and β be any two possible worlds. Imagine that in possible world α, on average, 29,000 children per day perish for want of basic life-sustaining necessities; and in possible world β, on average, less than a single child per day perishes for want of basic necessities. Let ‘pattern α’ denote the pattern of flourishing exhibited in possible world α; and refer to the pattern of flourishing exhibited in possible world β as ‘pattern β’.

Provisionally, just for the sake of argument, let’s suppose that The Conjecture is true: God would bring about an optimal pattern of human flourishing. In light of The Fact, we know that the actual world exhibits pattern α rather than pattern β. So if God exists, given The Conjecture, it follows that pattern α is no less optimal a pattern of human flourishing than pattern β. (Absent The Conjecture, of course, it may be that pattern β is more optimal a pattern of human flourishing than pattern α, yet their respective logical entailments make pattern α morally preferable, human flourishing notwithstanding, to pattern β. But that would imply that there’s some (aggregation of) good(s) that a morally perfect God would realize over an optimal pattern of human flourishing. And that would contradict our provisional assumption that The Conjecture is true.) So it follows from The Fact, theism and The Conjecture that: A world in which 29,000 children per day perish for lack of life’s basic necessities conforms to a pattern of human flourishing no less optimal than that of a world in which, on average, less than a single child per day dies under such circumstances. Since this is a logical implication of conjoining {The Fact & The Conjecture & theism}, we’ll refer to it as The Implication.

Notice that The Fact is merely an observation about the world we inhabit; and the skeptical theist is committed to theism. So as far as it concerns the skeptical theist (qua theist), The Conjecture entails The Implication.10 And if The Conjecture entails The Implication, the skeptical theist cannot have more reason to deny The Implication than she has for denying The Conjecture.11 This yields the following dilemma. The skeptical theist is in no position to assert that The Conjecture is false. Yet if The Implication is true (or true for all we know) then we shouldn’t think that we know very much at all about the moral status of human conduct.

We have defined ‘flourishing’ as nothing other than some feature(s) of human life that God, by virtue of moral perfection, would be motivated to realize. It follows that human flourishing is a morally valuable good. Thus an optimal pattern of human flourishing is, ceteris paribus, a morally appropriate arrangement of a morally valuable good—viz., the good of human flourishing. (It’s important to observe that this doesn’t entail the truth of The Conjecture, which posits that an optimal pattern of human flourishing is a morally appropriate arrangement of human flourishing whether or not all other concerns are held equal). It follows that God, qua morally perfect person, would bring about a less-than-optimal pattern of human flourishing only if there is some (aggregation of) good(s), G, such that: The realization of G is at least as morally valuable as an optimal pattern of human flourishing; and the realization of G is logically incompatible with the realization of an optimal pattern of human flourishing.12 We’ll refer to such a good as ‘a good that would morally justify God in failing to bring about an optimal pattern of human flourishing’.

For argument’s sake, suppose that the skeptical theist wishes to deny The Conjecture. To that end, the skeptical theist might point to some good, G, and claim that G would morally justify God in failing to bring about an optimal pattern of human flourishing—and that, moreover, God would in fact realize G instead of bringing about an optimal pattern of human flourishing. (Perhaps, e.g., the skeptical theist claims that the realization of G is not only as morally valuable as the realization of an optimal pattern of human flourishing, but in fact more so.) It would follow that The Conjecture is false.

Notice that any argument along this line would constitute a theodicy: “I know of some good that would morally justify God in realizing a less than optimal pattern of flourishing (e.g., free will, soul-making, etc.); and indeed, God has realized that good instead of realizing an optimal pattern of flourishing. Hence, The Conjecture is false.” I take my 2015 paper on skeptical theism and theodicy to have demonstrated that STS is incompatible with theodicy.13 Broadly, where ‘G’ is any given (aggregation of) good(s), I argue that those who embrace STS cannot consistently claim to know that G would morally justify God in realizing one state of affairs rather than another. My account would apply, mutatis mutandis, to the skeptical theist’s reasoning about whether some or other good would morally justify God in realizing a state of affairs in which the pattern of flourishing is suboptimal. Rather than rehearsing the details of that argument here, I will proceed on the assumption that my argument on that point succeeds. Given that assumption, the skeptical theist cannot, with consistency, point to a particular good and deny The Conjecture on the grounds that God would realize (or has realized, or will realize) that good instead of bringing about a state of affairs that includes an optimal pattern of human flourishing.14

The skeptical theist might go on to claim that we needn’t know which good morally justifies God in failing to bring about an optimal pattern of human flourishing in order to know that some good morally justifies God in failing to bring about an optimal pattern of human flourishing. However, this only gives us a reason to deny The Conjecture if we have some reason for thinking that, in point of fact, God (exists and) hasn’t brought about an optimal pattern of human flourishing.15 Presumably, we’d come to know that God hasn’t brought about an optimal pattern of human flourishing by pointing to some feature of the world—e.g., The Fact—and adducing that an optimal pattern of human flourishing wouldn’t have that particular feature. But the skeptical theist claims that STS undermines precisely that line of reasoning. So at most the skeptical theist can claim to know that if (God exists and) the pattern of flourishing in our world is less than optimal then there is some good that would morally justify God in failing to bring about an optimal pattern of human flourishing. The point to notice is that the skeptical theist cannot claim to know that the pattern of flourishing in our world is in fact less than optimal. Thus the skeptical theist has no reason for thinking either that God wouldn’t or that God hasn’t brought about an optimal pattern of human flourishing. So, for all the skeptical theist claims to know, The Implication is true.

One might think that this result is reason enough to reject skeptical theism. “It’s just obvious,” the argument would begin, “that The Implication is false. (Perhaps there is a good that would morally justify God in failing to bring about an optimal pattern of human flourishing. Be that as it may, features of our world like The Fact make it clear that our world doesn’t conform to an optimal pattern of human flourishing.) Yet skeptical theism entails that The Implication is true or true for all we know. So we should reject skeptical theism.” I would expect the skeptical theist to reply along the following line. “For all we know, humans can flourish in ways that are unknown (to us), some of which may be involved in The Fact. Furthermore, there may, for all we know, be unknown ways of realizing human flourishing. So we shouldn’t doubt that The Fact stands in unknown entailment relations to known or even unknown ways of human flourishing. Accordingly, it’s not at all obvious that the pattern of human flourishing that we see in the world is less than optimal.”

Here’s the difficulty with that response. Suppose that X is a way of human flourishing that’s totally unknown (to us). What are X’s properties? Unless I assume that properties of unknown ways of flourishing would resemble properties of known ways of flourishing, I have no idea. (And surely the skeptical theist denies that we are entitled to such an assumption.) So let F_X be a feature of X in virtue of which X is a way of human flourishing, such that F_X is unlike any known feature of any known way of human flourishing. What reason do we have for thinking that F_X does not, unbeknownst to us, adhere to some or other way(s) of human life? Given STS, none at all: We haven’t a clue what F_X might be. Nor can I imagine, given STS, a good reason for supposing that F_X—whatever it may be—wouldn’t be just one among many unknown features of human flourishing that might, for all we know, attach to many known ways of human life. So if we endorse STS then we shouldn’t doubt that there are many unknown features of human flourishing that adhere to all manner of known ways of human life. I will hereafter use the phrase deep axiological skepticism to describe the view that, for all we know, there are many unknown ways of flourishing that adhere to all manner of known ways of human life. Thus we’ve just established that STS implies deep axiological skepticism.

Now let W be a way of human life that we least associate with flourishing—for instance, a relatively brief life, the entirety of which is spent languishing from debilitating illness and lack of adequate nutrition. And let F_X be some unknown manner of human flourishing that, unbeknownst to us, adheres to some known way(s) of human life. Given deep axiological skepticism, I cannot think of any reason for supposing that F_X isn’t a feature of W: As we have no idea what F_X could be, we have no reason for thinking that F_X does not, unbeknownst to us, adhere to W. So for all we know, W constitutes robust human flourishing no less than, say, a life of healthful leisure and high-minded contemplation, replete with the affections of family and friends who enjoy the same. Accordingly, we have no moral basis for thinking that we should orient our conduct toward realizing one of those ways of life rather than the other. Nor, it seems, would we have any principled moral basis for choosing one rather than another from among the many ways of human life that we take to be more constitutive of human flourishing than W. So deep axiological skepticism leaves us without any basis for directing our actions toward one telos rather than some (or any) other. Therefore, STS implies skepticism about moral knowledge.

4. Objections and Conclusions

I’ll close by considering a couple of objections, the first of which is this. I would expect the skeptical theist to object that we needn’t know anything at all about an optimal pattern of human flourishing, or whether or how God might go about realizing an optimal pattern of human flourishing, in order to have a grasp of human flourishing that’s sufficient for the purposes of directing our own actions toward that end. “While it may be morally appropriate for God to bring about (an optimal pattern of) the flourishing of all humankind,” the argument would contend, “It is morally appropriate for us to bring about the flourishing of ourselves and those to whom we stand in special relationships—and, perhaps, once we’ve satisfied those obligations, we morally ought to do our best to bring about the flourishing of some subset of those who remain. We needn’t know the details of God’s moral situation in order to know exactly what we morally ought to do.”

This objection misses the force of my argument: For all the skeptical theist claims to know, The Implication is true. If The Implication is true (for all we know), then how are we to regard The Fact? Is the apparent suffering of millions of starving children constitutive of human flourishing (as a soul-making theodicist might claim—cf. Stump; Hick)?16 Or is it merely a regrettable but ultimately necessary condition for the realization of an optimal pattern of human flourishing? Could it be some combination of the two, or something else entirely? I don’t see a reason, consistent with skeptical theism, for rejecting any of these possibilities. And if, for all we know, the human experiences implicated in The Fact are constitutive of human flourishing, I don’t see any reason at all for thinking that we know what flourishing consists in or how flourishing is to be achieved. Since flourishing is, by definition, a morally significant good, our ignorance about what constitutes flourishing and how to achieve it should undermine any confidence we have in the deliverances of moral cognition.

Another objection is that deep axiological skepticism needn’t bother the skeptical theist much, at least insofar as the skeptical theist holds theistic background assumptions about morality. As Bergmann and Rea note,

Skeptical theists are, after all, theists. Thus, when they consider the bearing of skeptical theism on their moral practice, they will inevitably and quite sensibly do so in a way that takes account of other things that they believe. But once this fact is appreciated, it is clear that most skeptical theists will find themselves completely untouched… The reason is simple: theists very typically believe that God has commanded his creatures to behave in certain ways; and they also very typically believe that God’s commands provide all-things-considered reasons to act.17

The thrust of the objection is that, at most, deep axiological skepticism threatens to undermine the moral knowledge of those who do not hold theistic background beliefs about morality.

Note that this objection is predicated on theistic belief. So if this is the skeptical theist’s only prospect for laying claim to moral knowledge, it follows that only theists who embrace STS can consistently claim to possess moral knowledge. Thus, on skeptical theists’ evident supposition that moral skepticism is a damnable implication for their view, the specter of moral skepticism would present the non-theist with a compelling reason to reject STS. In short, the skeptical theist’s skepticism would be highly unattractive to any non-theist that isn’t content to embrace moral skepticism. Moreover, the skeptical theist hopes to persuade others that her skepticism is plausible or even commonsensical. Since a great many philosophers reject both theism and moral skepticism, that hope is frustrated if those who embrace STS must thereby assent to the disjunction ‘theism or moral skepticism’. So this objection, even if it succeeds, should be of small consolation to the defenders of skeptical theistic skepticism.