Schellenberg's argument from horrors explained

Schellenberg's argument takes its start by noting several considerations neglected by contemporary discussions of the problem of evil. The first is that a God who is both perfect and personal (the notion of God presupposed by Schellenberg's argument is that of a perfect and personal creator of any universe there may be: necessarily omnipotent, omniscient, and perfectly good and loving) would have maximally complete knowledge by acquaintance of every possible occasion of horrific suffering (where he follows Marylyn Adams in thinking of horrific suffering as suffering of such a magnitude as to give the perpetrator and/or the victim a prima facie reason for thinking her life not worth living).

Note the strength of the claim. Schellenberg isn't proposing that God has maximally complete knowledge by acquaintance of every actual experience of horrific suffering. The claim, rather, is that God has knowledge by acquaintance of every possible experience of suffering, where that is a matter of his having undergone courses of experience qualitatively indistinguishable from every possible horrific experience. Only then, he thinks, would God possess the sort of ‘maximally rich and penetrating and meaningful understanding of suffering that we must surely associate with divine perfection’.Footnote ² By way of brief complaint, this isn't obvious. True, human ability to attain depth of understanding into the nature of suffering seems to require having oneself experienced a certain amount of suffering, but it's not clear why that would be true of God. Why not think hisFootnote ³ ability to project himself imaginatively into a course of suffering without actually experiencing that suffering would be sufficient for him to possess that maximally rich and penetrating understanding we would expect of the divine? Perhaps he doesn't even need that: perhaps he has maximal depth of understanding of every possible experience without needing to either have the experience or imaginatively project into it. Why couldn't this be?

The second neglected consideration concerns the relationship between outweighing goods and the deepest good of personal creatures. Schellenberg notes, quite plausibly, that these seem conceptually distinct. It could well be that some bit of horrific suffering is (a) required by some good whose goodness outweighs the evil of the suffering, but is nevertheless (b) wholly unnecessary for the sufferer's deepest good.

The third neglected consideration is that the deepest good of any possible finite personal beings created by God is known to us and is an unending and ever-deepening relationship with God: ‘If indeed God is to be construed as perfection personified, then what could be better for finite creatures than to enter ever more fully into the maximally great richness and beauty and glory of God?’.Footnote ⁴

The fourth neglected consideration is that there would seem to be nothing to prevent those created persons who experience no horrific suffering from realizing the just-described good, for ‘if God is unsurpassably deep and rich, then even where no horrors are experienced there must be an infinite number of ways of developing a relationship with God – an infinite number of possible journeys into self, the world, and God that realize the ultimate in meaning and goodness for finite created persons’.Footnote ⁵

So far, then, four neglected considerations in the vicinity of the problem of horrors.Footnote ⁶ Schellenberg turns next to some implications.

With respect to the first consideration, he proposes that, in the case of humans, the greater someone's empathy (where depth of empathy, as he is thinking of it, is a matter of depth of compassion and depth of acquaintance with horrific suffering), the stronger her opposition to horrific suffering – the stronger her disposition to seek to eliminate it wherever she can. But God, if there is such a being, would be maximally empathetic (maximally compassionate and, as per the first neglected consideration, maximally acquainted with horrific suffering). So we must expect, then, that were there such a being as God, he would be maximally opposed to horrific suffering and thus maximally disposed to seek to prevent or eliminate it wherever possible.

Of course, it doesn't follow so far forth that God would prevent or eliminate all horrific suffering, for there might be outweighing goods which justify God in permitting such suffering. This takes us to Schellenberg's second neglected consideration on the distinction between outweighing goods and the deepest goods of created persons. He posits here that a perfectly empathetic being of the sort God would be (were there such a being) would permit a given bit of horrific suffering only if the outweighing good for the sake of which the suffering were allowed was the realization, by the sufferers, of their deepest good. Put differently:

And here the third and fourth considerations become relevant, for we know what the deepest good of finite created persons would be, if there were such a person as God: an unending and ever-deepening relationship with God. And it is implausible in the extreme, thinks Schellenberg, that that good should require horrific suffering. A God who is infinitely deep and unsurpassably rich would have available to him, surely, an infinite number of ways of growing someone into wholeness and fulfilment in relationship with him, without the need for horrific suffering.

Pulling all the pieces together, then, we get this: were there such a being as God, he would be maximally empathetic (maximally compassionate and maximally acquainted with horrific suffering) and thus maximally opposed to horrific suffering. And were he thus maximally opposed to horrific suffering, he would permit it in the lives of his creatures only if it were necessary for an outweighing good which constituted their deepest good.Footnote ⁸ But were there such a being as God, the deepest good for any finite created person would be unending and ever deepening relationship with God, and it is evident that horrific suffering would not be necessary for the realization of that good in any creature's life. Wherefore, were there such a being as God, he would not permit horrific suffering in the lives of finite created persons. Since there is plenty of horrific suffering in the lives of finite created persons, we may conclude that there is no such being as God.

Schellenberg's argument from horrors examined

I develop in what follows a two-part reply to Schellenberg. First, I put an epistemic objection against his argument, urging that no one could be justified in accepting its conclusion on the basis of its premises, and that therefore, the argument fails as an argument for atheism. Second, and more interestingly I think, I develop a conception of God a bit different than that at the heart of classical theism, and urge that Schellenberg's argument makes no trouble for belief in God so construed. This is an interesting result. The notion of God I'll be developing is one that, in certain ways, fits better with the God depicted in the Hebrew and Christian scriptures than does the God of classical theism, and describes a creator of enormous power, knowledge, love, and goodness. It's an interesting possibility, I think, that philosophical atheological arguments from evil, of which Schellenberg's is a species, are driven by features of the classical theistic conception of God which were, in effect, philosophical add-ons to the Semitic conception of the divine presupposed by the writers of the Hebrew and Christian scriptures, and that these arguments have no traction against other conceptions of the divine, conceptions on which God is a personal creator of enormous power, wisdom, love, and goodness, but differs in interesting ways from the God of classical theism. It's beyond the scope of this article to argue that no philosophical atheological argument from evil makes trouble for this alternative conception of God, but if I'm right that Schellenberg's (very powerful) argument makes no such trouble, it's a good sign. (It's a good sign, anyway, for those of us attracted to variations on the conception of the divine I'll sketch.)

Philosophical scepticism and the greatest good for humans

The target of my first objection is this:

It is the conjunction of two key premises in Schellenberg's argument. I think this thesis false, but I won't argue that here. I'll argue, rather, that no one could justifiably accept the thesis as a premise in an argument to atheism, and that therefore Schellenberg's argument fails as an argument for atheism: it is not an argument on the basis of which one could rationally hold atheistic belief.

My argument takes its start by noting that the Deepest Good Thesis is a recondite philosophical claim: a deep and profound claim about matters far removed from the everyday concerns of life and about which there is serious philosophical debate. That it is a deep and profound claim about matters far removed from the everyday concerns of life is, I take it, clear enough. As to its being subject to serious philosophical debate, consider this alternative proposal defended, in various versions, throughout the history of the discussion.

On this proposal, the deepest good for created persons is union with God, where this is at least partly a matter of having joined one's will to God's – having autonomously willed oneself to become the sort of person who freely wills the same things God wills, where one wills a certain course of action autonomously if and only if (a) one is not caused to will thus, but is oneself the source of the willing, and (b) one does so from a position of significant responsibility, the sort of responsibility one possesses if and only if it is within one's power to realize seriously good and seriously evil courses of action. It is a further claim of the proposal that, depending on how creatures use their autonomy, it may sometimes happen that God's best hope for bringing a creature into union with God is to permit horrific suffering in the creature's life, horrific suffering functioning here as kind of medicine for the soul, painful in its administration, but the best (and maybe only) means of inducing the creature to seek union with God.Footnote ⁹

Here, then, is an alternative view of our deepest good which is incompatible with the Deepest Good Thesis and has been defended by able philosophers.Footnote ¹⁰ I bring it up not to defend it. My point is just that the Deepest Good Thesis is subject to serious debate.

To summarize: the Deepest Good Thesis is a recondite philosophical claim, a deep and profound claim about matters far removed from the everyday concerns of life and about which there is serious philosophical debate.

Next let me briefly rehearse an argument I have developed elsewhereFootnote ¹¹ for the claim that atheists – for present purposes, those who deny the existence of an extremely powerful, wise, good, and loving creator of all – have good reason to doubt the deliverances of those portions of their cognitive endowment responsible for belief about recondite philosophical matters.

The argument has three steps. The first step is to point out that, if you are an atheist, you have powerful reason to accept this

There is powerful scientific evidence for this claim, and though some religious believers are sceptical of certain parts of the evolutionary story, I take it that those doubts wouldn't be operative for most atheists.

The second step is to defend this

Briefly, is there any reason to expect processes of the sort described by contemporary evolutionary theory to have endowed us with much by way of reliable insight into recondite philosophical matters, given that they were unguided by any sort of theistic intelligence? Surely the answer here is ‘no’. According to the usual evolutionary story, human cognitive faculties – of the same basic sort we possess today – appeared during the Pleistocene era, during the period lasting from about 2.5 million years ago to about 12,000 years ago. By the end of that period, our ancestors possessed brains of the same basic architecture and cognitive capabilities as our brains. The main explanation why they evolved these faculties, on the usual story, is that having such faculties was adaptive in Pleistocene environments: useful for feeding, flying, fighting, and reproducing on the plains of Pleistocene Africa. But why would cognitive faculties selected for their success at those tasks in those environments have required the ability to theorize reliably about recondite philosophical matters, matters which would have been wholly irrelevant to life on the plains of Pleistocene Africa? From a fitness point of view, such cognitive capability seems wholly unnecessary. But if so, the probability that it should have evolved seems low.

There is of course the possibility that reliability on recondite philosophical matters far removed from the everyday concerns of life is a ‘spandrel’ – a non-adaptive by-product of some adaptively selected trait, in the way that, for example, abstract mathematical abilities could be non-adaptive by-products of the adaptive ability to do simple arithmetic and geometry. That could be, but as I read the cognitive science literature, no one has been able to find good reason for thinking so. The right thing to say here, I think, in light of this possibility and the uncertainty of our evidence, is that the probability that we should have got cognitive faculties capable of reliability on recondite philosophical matters, given atheism and the usual evolutionary story, is inscrutable.

(Note well that I am not here reproducing Plantinga's famed evolutionary argument from naturalism.Footnote ¹² Although my argument is inspired by and similar to his in various ways, my argument differs in this key respect: whereas he argues that the probability, given atheistic evolution, that we'd have got overall cognitive reliability is low or inscrutable, I am arguing for something much weaker and much more easily defended. Critics of his argument have replied that, given the usual evolutionary story, we should expect general cognitive reliability. I needn't dispute that. I claim only that, given atheistic versions of the usual evolutionary story, we should be agnostic about philosophical reliability.)

The third step is to defend this

where I take it for granted here that you have good epistemic reasons for agnosticism about the reliability of some source of information S if you have good epistemic reasons for agnosticism about the probability that S is reliable.Footnote ¹³

Why accept my Principle? Briefly: suppose, for example, that you have good epistemic reasons for agnosticism about the reliability of a certain pool thermometer, and no reason to discount or ignore those reasons for agnosticism. Then, one thinks, you should not believe what the thermometer says about the water's temperature unless you have some other source of information to go on. Reflection on this and like cases, I suggest, lends strong support to the Principle.

Our two Theses and Principle in hand, the argument that, if you are an atheist, you have good reason to doubt the deliverances of those portions of your cognitive endowment responsible for belief about recondite philosophical matters is easily stated. For if you are an atheist, then for reasons I have just laid out, you have good epistemic reasons to think that those of your cognitive faculties responsible for recondite philosophical belief (your philosophical faculties, henceforth) are the product of atheistic evolutionary processes of the sort described by contemporary evolutionary theory, and you've also good epistemic reasons for agnosticism about the reliability of faculties so produced. Wherefore, provided you have no good epistemic reasons for discounting or ignoring these reasons, it follows by our Principle that the rational attitude to take towards matters about which your philosophical faculties are your only source of information is agnosticism.

Since (so I have argued elsewhereFootnote ¹⁴) there aren't any good reasons for discounting or ignoring the above reasons for doubting the reliability of our philosophical faculties (given atheistic evolution), our reasoning leads us to this conclusion: if you are an atheist, the rational attitude to take towards matters about which your philosophical faculties are your only source of information is agnosticism.

Now for the application to Schellenberg's Deepest Good Thesis. For it is clear, I take it, that your philosophical faculties are your only source of information on that Thesis. It follows, therefore, that, if you are an atheist, the rational attitude for you towards that Thesis is agnosticism. But if so, then you can't sensibly accept the conclusion of Schellenberg's argument on the basis of its premises: belief in its conclusion renders belief in the conjunction of two of its key premises unjustified for you. And if so, the argument fails as an argument for atheism.

Such is my first objection to Schellenberg's atheological argument from horrors.Footnote ¹⁵ I recognize that there is something deeply unsatisfying about this style of reply to Schellenberg's and like arguments. Even if it is right, it sheds no light on the ways of God, offers no help in fathoming how an allegedly wholly good and loving being of enormous power and wisdom could countenance a world of horrible suffering. It would be nice if we could do that.

In the remainder of the article, I sketch a conception of God which, if correct, would shed light on why God allows horrible suffering. It is a conception of God different in important ways from the classical theistic conception, though compatible, I think, with the depiction of God in the Hebrew and Christian scriptures. Showing as much would take some work, as would properly developing and defending the view. I shall not attempt that here. I put it forward, instead, as a defence of broad theism (the thesis that the world owes its being to a personal creator of enormous power, wisdom, goodness, and love; for short, ‘theism’) against Schellenberg's argument from horrors. Here, I propose, is a picture which entails both theism and the existence of much horrific suffering, is accurate for all anyone knows (given that there is a God), and is such that Schellenberg's argument from horrors gives no good reason for doubting it. If I am right, it follows that Schellenberg's argument from horrors gives no good reason for doubting theism. (I put the picture forward, then, as a defence, claiming no more for it here than that it is epistemically possible and that Schellenberg's argument gives no reason to doubt it. I lack space to explore the idea in this article, but I am attracted to the possibility that, suitably nuanced, fleshed out, and defended, this way of thinking about God might well shed light on why it is that God, a being of enormous power, wisdom, goodness, and love, countenances the world of horrible suffering we inhabit.) To the picture, then.

God as non-coercive love

I draw heavily in what follows on the theological writings of Thomas J. Oord.Footnote ¹⁶ I'll develop things in my own way, but I am borrowing generously from Oord.

I develop the picture in stages. We start with the idea that God is non-coercive love, or less metaphorically, that God is a person among whose chief attributes is non-coercive love. Let me say a bit about that.

God, on our picture, is a person: a being with beliefs, desires, aims, and intentions. He is essentially characterized by the property of love, and in particular, by what Nicholas Wolterstorff calls the love of benevolence, the sort of love that seeks the good, the well-being, the flourishing of the beloved. God, on our picture, is essentially such that he always pursues the good, the well-being, the flourishing of all things.Footnote ¹⁷

I build into the picture a trinitarian component, on which God is tripersonal, comprising the Father, who is everlastingly the ontological source of the Son, which two together are everlastingly the source of the Spirit. Part of the way God's love is manifested is in relationship among the members of the Trinity: each is essentially such that he always seeks the well-being of the other.

There are different ways of seeking the well-being of a beloved. One thinks here about different styles of parenting. There are those parents who, out of concern for a child's well-being, push and prod her into courses of behaviour conducive to well-being by application of various pressure tactics: criticism, guilt, anger, dissembling, threats of violence, actual violence. All such tactics deploy a coercive approach to love, where I follow Aquinas in thinking of coercion as a sort of necessity, wherein one person's actions make necessary those of another.Footnote ¹⁸

God, on the picture I am developing, is essentially opposed to coercion: the members of the Trinity are necessarily such as to take a non-coercive approach in their love for one another and in their love for created beings.

It would be nice to be able to say something more precise about what non-coercive love comes to. Let me try. As I am thinking of things, to seek the well-being of the beloved in a non-coercive way is, first, to seek to bestow on it goods constitutive of its well-being without necessitating that it behave in any particular way. Non-coercive treatment of a thing, then, requires that the behaviour of the thing not be determined by your actions towards it.

Of course, you can treat a thing coercively without necessitating its behaviour. Threatening someone with horrible violence may not necessitate that he behave as you'd wish, but is coercive in the usual case. Part of the explanation why such treatment would ordinarily be coercive is that, though it doesn't necessitate a given course of behaviour, it nevertheless exerts causal pressure on the coerced of a sort whose effect is to strongly incline the coerced to behave in ways that go contrary to how she would have behaved if left to her own devices.

If we think of necessitating causal pressure as causal influence that necessitates behaviour, and strongly inclining causal pressure as causal influence that strongly inclines behaviour of a sort that runs contrary to what would have transpired had the pressure not been applied, then we may say that non-coercive love for a thing seeks the good of the thing without exerting necessitating or strongly inclining casual pressure. It seeks the good of the beloved, often by application of causal pressure, but causal pressure of a sort that weakly inclines – gently influences, woos – and co-operates – working with the native powers of things, amplifying those powers in so far as they are already autonomously tending in the direction of God's weakly inclining influence. (The occurrence of miracles would be a matter of such amplifying divine causal influence.)

Such, then, is the non-coercive love of God: it seeks the good, gently influences, woos, co-operates, but never forces.

Next, we add the idea that God, by nature, interacts in this way with everything, personal or no. He loves galaxies, stars, electrons, mountains, rocks, tomatoes, dogs, and humans in this way: gently influencing them, wooing the unfolding of their potentials in the direction of beauty and goodness, joining his power to theirs so as to magnify their potential for beauty and goodness, but never necessitating or strongly inclining.

Next is the idea that God is essentially and everlastingly creative: he is the creator of all that is non-divine; he is everlasting (has always existed), and is essentially such as to have always been creating. As we'll think of things, he is essentially and eternally such as continuously to create matter of certain sorts, so that whether and what types of matter he creates is not subject to divine choice. Divine choice, on the picture, is limited to the directions in which he shapes, woos, and moulds this continuously created matter. Even here, though, his creative activity is limited by his nature to non-coercively wooing the unfolding of potentials inherent in matter in directions I shall describe in the next paragraph. (So unlike most versions of process theology and Oord's open theism, God, on the picture I am sketching, does engage in creation ex nihilo, but in a very limited way: his creating ex nihilo is restricted to the matter he creates, and it is not a voluntary creation.)

Next, a point about the deepest good of finite created persons, which, says our picture, is participation in shalom community: community comprising God, created finite persons, and non-personal created things, characterized by non-coercive, loving relationship between God and creatures, creatures and other creatures, and any one creature and itself, in which the livelihood, well-being, and dignity of every creature is secured in co-operative endeavour between God and creature.Footnote ¹⁹ Such is what God, by nature, non-coercively draws all things towards and such is our deepest good.

Finally, a point about the existence of horrors. God, by dint of his nature, continuously creates certain sorts of matter, and by nature, continuously seeks to shape, mould, and woo this matter in the direction of shalom community. But his nature and powers render him incapable of coercive interaction with his creation, restricting him to weakly inclining and co-operating causal pressure, to allowing the potentials in things to unfold mostly in their own way, in their own time, wooing the unfolding of those potentials in the direction of shalom community, co-operating with them as they unfold in that direction, but never doing so in such a way as to necessitate or strongly incline. Because the potentials in things unfold autonomously (without necessitating or strongly inclining causal pressure from God) and, I shall suppose, indeterministically, they often unfold in directions opposed to shalom community. They often unfold in the direction of horrific suffering. When this happens, God grieves, suffering with the sufferer, and exerts weakly inclining causal pressure so as to draw those creatures and processes causing the suffering into different directions, towards shalom, towards communities of non-coercive love. And he relentlessly pursues victims and perpetrators of horrors into pathways of healing, wholeness, forgiveness, and joy, and is limitlessly patient and persistent in this pursuit. But he is incapable of enforcing his will or strongly inclining non-horrific outcomes; that is beyond his power.Footnote ²⁰ Consequently, though the long-term direction of things inclines towards shalom, the process is slow, roundabout, and detours through much suffering. In Martin Luther King's (and Theodore Parker's before him) powerful phrase, the moral arc of the universe is long, but bends towards justice.

That completes our picture. What I claim of this picture is that it is accurate for all anyone knows (given that there is a God), and that Schellenberg's argument from horrors gives no good reason for rejecting it. I close with a brief defence of the latter claim.

Return to Schellenberg's neglected considerations.

The first was that a perfect and personal God would have knowledge by acquaintance of every possible occasion of horrific suffering. You will notice that the picture I have developed says nothing about perfection or maximality. That was by design. I think of that as one of those philosophical add-ons to the Hebrew conception of God which eventually came to be central to the classical theistic conception. The being I describe may or may not be the being more perfect than which none can be conceived; it doesn't much matter. Still, I suppose God to be a being of enormous knowledge, power, and love, and it's not at all inconsonant with my picture to suppose he possesses the sort of maximal empathy Schellenberg describes.

The second was the suggestion that there is a distinction between outweighing goods and deepest goods. I would be happy to take that on board too.

The third was the claim that the deepest good of any possible finite personal beings created by God is ever-deepening relationship with God. Not so, on our picture. Participation in a shalom community of non-coercive love with God and other creatures is our deepest good. That would certainly carry in its train ever-increasing relationship with God, but it is much more besides.

The fourth was that there would seem to be nothing to prevent those created persons who experience no horrific suffering from realizing the good of ever-deepening relationship with God. I agree. Likewise, I think there is nothing in principle to prevent created persons who experience no horrific suffering from realizing the good of participation in a shalom community of non-coercive love with God and other creatures.

Let us turn next to the implications Schellenberg draws from his neglected considerations.

First, there is his proposal that a maximally empathetic being would be maximally opposed to horrific suffering – maximally disposed to prevent or eliminate it wherever possible.

Perhaps so, but what is possible for God is a function of his nature and powers, and on the picture I have been sketching, God's nature and powers restrict him to weakly inclining and co-operating casual interaction with created things, to allowing the potentials in things to unfold mostly in their own way, on their own time, wooing the unfolding of those potentials in the direction of shalom community, but never forcing. That being so, says the picture, though God is maximally opposed to horrific suffering, it is not within his power to prevent or eliminate it.

Schellenberg next proposes that a maximally empathetic God would permit a given bit of horrific suffering only if the outweighing good for the sake of which the suffering was allowed was the realization, by the sufferers, of their deepest good, which for Schellenberg, is a matter of ever-deepening growth in relationship with God.

Things look different from the perspective afforded by our picture. Though God, as we have depicted him, would be maximally empathetic (maximally acquainted with actual and possible suffering and maximally compassionate), he would be unable to assure (or even render it highly likely) that horrific suffering occurred only in so far as it conduced to the deepest good of the sufferer. He would no doubt weakly incline the processes of creation, endeavouring to bring it about that horrific suffering occurred only in so far as it was necessary for the deepest good of the perpetrators and victims of that suffering, wooing those processes and their participants away from gratuitous horrorsFootnote ²¹ and towards shalom community. And he would no doubt relentlessly pursue the victims and perpetrators of those horrors into pathways of healing and wholeness. But he would be unable to guarantee or strongly incline in such a way that horrific suffering was never gratuitous; such would be beyond his power.

Note then that the question why God permits suffering admits of a complex answer on our picture. In some cases, God would no doubt refrain from exerting any sort of countervailing causal pressure at all on processes tending towards suffering because he would see that those processes and their consequent suffering conduced to the deepest good of those involved. In other cases, his permission of suffering would have nothing to do with pursuit of an outweighing good; it would be a matter of his lacking the power to prevent the suffering, since doing so would require coercive intervention in the unfolding of creaturely potentials, of which God would be, by nature, incapable. Not all suffering, on our picture, would subserve an outweighing good.

Finally, there is Schellenberg's point that the deepest good of finite creatures is ever-deepening relationship with God, a good realizable in any created person's life without permission of horrific suffering in her life. Our picture holds, by contrast, that the deepest good of finite creatures is participation in shalom community, and while it is perhaps logically possible that such community be realized without horrific suffering, it would not be within God's power to assure or even render highly likely the actualization of that possibility.

By way of conclusion, then, here is a picture of God on which God is a personal creator of enormous power, wisdom, goodness, and love, and on which there is much horrific suffering. I claimed that the picture is accurate for all anyone knows (if indeed there is a God), and that Schellenberg's argument from horrors gives no good reason for doubting it. If I'm right, it follows that Schellenberg's argument from horrors gives no good reason for doubting theism.Footnote ²²