Fr Kimel recently released When God becomes Jack Bauer, a characteristically generous and detailed defense of David Bentley Hart’s universalism against Mats Wahlberg’s greater-good theodicy of hell. The argument is compelling. If God creates ex nihilo under no external constraint, then the eschatological outcome is revelatory of who God freely is, and any eternal remainder of unredeemed evil reduces divine goodness to a merely relative good. Such a God, Hart argues, has “made a bargain” with evil rather than conquered it. I find myself in deep sympathy with nearly everything Fr. Kimel has written, and I share his conviction that the infernalist tradition cannot withstand Hart’s argument against it.

What I address here is not the theodicy question as such (not the question whether evil can be justified, which Hart and Fr. Kimel have answered, rightly, in the negative) but something internal to Hart’s universalism itself: whether the metaphysical commitments Hart brings to the question of creation can actually sustain the strongest claim his universalism requires. This is a question about what universalism needs in order to be fully coherent, not a reopening of the debate Hart has already closed. Specifically, I want to draw attention to a fault line that Fr. Kimel’s essay, by its very success in presenting Hart’s argument, leaves hidden. A fault line not between Kimel and Hart, but within Hart himself. It is a fault line that matters considerably for how we understand what “answers to nothing” can finally mean.

Consider the two passages from Hart that Kimel quotes with approval. In The Doors of the Sea, Hart writes that “the entire history of sin and death is in an ultimate sense a pure contingency, one that is not as such desired by God,” and that suffering and evil have “no ultimate value or spiritual meaning at all… cosmic contingencies, ontological shadows, intrinsically devoid of substance or purpose.” These are among the most important sentences Hart has written on the subject of evil. They do not merely say that evil is ontologically privative (i.e. a lack of being rather than a positive force) though they say that too. They say something stronger and more theodically decisive: that evil is a pure contingency, that the history of suffering answers to nothing, serves no purpose, contributes nothing to any design. The word “pure” is doing real work here. It is unqualified. It doesn’t state the history of suffering is contingency within a necessary order, or privation alongside a gratuitous gift. It says pure contingency, period.

Now set beside this Hart’s mature metaphysical position, which Kimel also draws upon. In Light of Tabor (Hart’s most recent and in many ways most systematic volume) Hart writes that “the language of creation from nothingness is simply the most radical and rigorous formulation of the metaphysics of emanation imaginable,” that creation is “the total donation, out of the infinite plenitude of divine Being, of both essence and existence to creatures that are absolutely contingent on that infinite gift.” Creation, on this account, is the self-diffusion of the Good, the outpouring of infinite actuality into participated being, bonum diffusivum sui. The sun must shine. Finite being radiates necessarily from infinite Being because the infinite plenitude of divine goodness is, by its very nature, self-communicative, not because God is compelled by anything exterior to himself.

The question I want to raise is this: can these two positions be held together without remainder?

If creation is the necessary emanation of divine plenitude into participated being, then evil (however privative, however lacking in a nature and purpose of its own) is the “shadow side” of a necessary order. Necessary, not willed. Not desired. Not intrinsically purposeful. But structurally ineliminable from the kind of finite, participated, defectible existence that necessarily radiates from infinite actuality. The privation is real; the evil has no being of its own. But it is still the defect of an order that had to be this kind of order, given what it is: participated being, finite, capable of falling short of its source. On this account, evil is the shadow of necessary light. It answers to nothing in the sense that it has no substance or purpose of its own. But it cannot be a pure contingency, because the order whose defect it is was not itself contingent. Something cannot be the privation of a necessary order and also be purely contingent. The two claims, taken together, are incoherent.

There is a further complication on display in The Doors of the Sea itself. Hart there runs a free-will theodicy alongside his evil as privation account: God permits the history of sin and death because “the union of free spiritual creatures with the God of love is a thing so wonderful that the power of creation to enslave itself to death must be permitted by God.” Notice the force of “must.” This is not merely descriptive. It is a justification: a reason why the wager was worth laying. The horror was the acceptable cost of something God judged surpassingly good. And here, precisely, the child’s suffering is made to answer to something: the incomprehensible worth of free creaturely union with God. It is a greater-good argument which is carefully limited, temporally bounded, and ultimately overcome in the eschatological universalist consummation. Note that it is structurally the same move that Kimel, following Hart, so powerfully targets in the infernalist tradition — only deployed one level up, within the universalist framework itself. The permission has already done its theodicy work before the eschaton arrives to dissolve it.

Fr. Kimel’s essay does not notice this because it reads Hart as a unified position, and because the target is the infernalist greater-good argument rather than the universalist one. But the fault line runs through Hart’s position as such, present from the beginning, and unresolved across his work.

To be sure, I am not saying Hart is wrong to affirm universalism. He is right, and the argument Kimel presents on his behalf against Wahlberg is, I believe, essentially sound. Nor am I saying that Hart’s evil as privation is mistaken, it is largely correct as far as it goes. What I am saying is that “pure contingency” (the claim that the child’s suffering answers to nothing, serves no order, purchases no good) requires something Hart’s emanation logic does not provide. That the world itself, with the child’s suffering in it, need not have been at all. Not merely that God had no need of it (deflationary gratuity, a claim about God’s act). Not merely that evil has no substance (privation, a claim about evil’s ontological status). But that the whole order, including the space for privation within it, was genuinely received existence, contingent gift, the kind of thing that need not have been given.

This is the distinction I have been pressing in my own theodicy essays. The difference between what I call gratuity, which tells you something about the giver (i.e. that God had no need, that creation adds nothing to the divine fullness) and contingency, which tells you something about the given — that participated existence, as such, is received being, the kind of thing that need not have been. These are different claims, and their difference is precisely what the theodicy question turns on. Gratuity tells you that God overflows into creation from sheer plenitude, requiring nothing from the creature, the act adding nothing to God’s being. Contingency tells you that the creature’s very existence is the kind of thing that need not have been — that received being, as such, carries in its constitution the mark of genuine gift. Hart has the first in abundance. What his emanation logic resists giving him (and what ‘pure contingency’ requires) is the second. And the two are not the same, even when both are described as requiring nothing from God, even when both leave God’s life entirely unchanged by creation’s existence.

There is a passage in The Beauty of the Infinite which makes the tension evident with particular clarity, because Hart runs both moves in a single paragraph and treats them as compatible. He writes that “there might just as well have been no creation (for creation adds nothing to God, but only participates in him).” This is precisely the contingency claim, and stated with force. But the passage does not end there. It continues: “but assert nevertheless that creation is ‘necessary’ in another, aesthetic sense: it has been from eternity fitting to God’s goodness to be a loving creator, manifesting his trinitarian love in creatures.” Here are three moves: creation adds nothing to God (gratuity of act); there might just as well have been no creation (apparent contingency of object); but it has been from eternity fitting that God be a loving creator (aesthetic necessity re-asserting itself). The third move does not merely qualify the second: it retracts it. What is eternally fitting to God’s goodness is not something that might just as well not have been. “From eternity fitting” is a modal claim, and it is a necessitarian one, however softened by the word “aesthetic.” Hart gives the contingency with one hand and withdraws it with the other, in the same breath, apparently without noticing the tension. This is the fault line — running through a single paragraph of his earliest systematic work, and unresolved in his most recent.

One could say, I suppose, that Hart’s deflationary gratuity already covers this ground, that “there might just as well have been no creation” is all the contingency his position requires. I do not think it is. An overflow that proceeds necessarily from infinite plenitude (even an overflow that adds nothing to the source, even one described as fitting rather than logically compelled) is not the same as a gift that need not have been given. The sun adds nothing to its own light by shining. But the sun, in Hart’s own analogy, must shine. A gift that must be given — or that it has been from eternity fitting to give — is no longer, in the fullest sense, a gift. And an order whose existence is eternally fitting to divine goodness cannot generate pure contingencies.

None of this diminishes what Kimel has accomplished in his essay. The Jack Bauer analogy is, I think, more explosive than even he has recognized — because the God who permits the child’s suffering for the sake of freedom’s surpassing worth stands in the same structural position as the God who permits eternal damnation for the greater good of… something. Both Gods have made a bargain. Yes, the bargain’s terms differ enormously, but the structure is identical. The full force of Kimel’s argument, pushed to its limit, requires not just universalism but the claim that the history of suffering answers to nothing: no wager, no bargain, no worth-it. That is the claim. It is Hart’s own claim in The Doors of the Sea. It is also the claim Hart’s later emanation metaphysics struggles to sustain.

I raise this not to diminish Hart, whose work on these questions remains indispensable, but because the fault line matters and because Fr. Kimel’s essay has brought it newly to the fore. If evil is to answer to nothing (purely, without remainder, without having served any order or purchased any good) then the world’s existence must be contingent in the strong sense: genuinely received, genuinely a gift that need not have been given. That is the only ground on which Ivan Karamazov’s protest can be fully honored rather than merely absorbed. This is not a tension that emerged as Hart’s thinking developed. It is structural, present already in The Beauty of the Infinite, his earliest systematic work, and unresolved in his most recent. The ‘pure contingency’ that his universalism most needs is precisely what his emanation logic — consistent throughout — cannot supply. That is the tension Kimel’s essay has concealed. It is also the reason that universalism, if it is to carry the full weight Hart and Kimel rightly place upon it, requires a stronger doctrine of creation’s contingency than Hart’s metaphysics will bear.

Ivan’s second refusal, and why the horror did not have to be

Robert F Fortuin

The long version. A companion to “Returning the Ticket.”

I want to begin with a question that lies underneath every other question about evil — and it is not merely a philosopher’s question. It is everyone’s. Whether the world had to be: if it did, the suffering child had to be there; if it did not, nothing required her existence.

And consider what hangs on the answer. If the world had to be, if God could not have refrained from making it, then this world had to be, and the child in the cancer ward had to be, and what was done to her had to be done, or at least had to stand among the permanent possibilities of the only order there ever could have been. It would have been built into the grain of things like the way knots are built into oak. If the world did not have to be, then none of it had to be. Not the child, not the ward, not the hours.

The whole thing could simply have not happened, and nothing whatever would have been missing.

I mean to argue that the world need not have been, and that everything depends on it.

In “Returning the Ticket” I argued that theodicy fails at its foundation. Its answers are not the trouble. The question itself is malformed. Every theodicy presupposes that divine goodness and creaturely goodness share enough content for God’s reasons to be, in principle, weighed and found sufficient. It assumes we could occupy the bench and hear the case. The Christian analogical tradition denies exactly this. God is not good in a sense we can measure and approve; God is good in a way that is the transcendent source of the very goodness by which we would presume to measure him. So the justificatory question, is the child’s suffering justified? does not get answered. It dissolves. Ivan returns his ticket to a God the tradition never confessed, the God whose harmonies can be shown to be worth their price; and the tradition, from inside faith, does not defend that God either.^[1]

I hold all of that still. But I have come to think it answers only half of what Ivan refuses.

Two refusals, not one

Read the “Rebellion” chapter of The Brothers Karamazov again and you find Ivan refusing two things, and theodicy (by hearing only the first) never registering his second refusal.

The first is the one I addressed. Ivan will not accept a suffering that is justified, a harmony purchased at the price of a child’s tears and then pronounced a good bargain. That refusal the analogy meets by dissolution.

Beneath it runs a second refusal, and I would say a much harder one. Ivan will not accept a suffering that is necessary, a horror woven so tightly into the fabric of the created order that it could not have been otherwise, a discord without which the final chord would not resolve. He hands back the ticket to the God who justifies the child’s death, and hands it back a second time to the God who wanted it there, whose world could not have gone forward without it.

This is the difficulty I did not face the first time. The dissolving of the justification question (the apophatic move, which unsays our confident measurements of God rather than answering them) has no bearing on the second refusal. You can dissolve the justification question forever and the necessity question stands untouched, because the most serious theologies on offer do not justify evil at all. They do something subtler, and to Ivan worse. They fit it in. They give the horror a place in an order it could not have been absent from. And then — precisely because they refuse to call it justified — they present this as reverence.

So the necessity refusal needs its own answer, and it cannot come from apophasis. It has to come from somewhere else. It comes from the one place the Church has always been most careful, and I will get to this shortly. But first I draw our attention to the two accounts that sharpen the problem, because both are held by theologians I admire, and both are more interesting than any theodicy.

The first account: creation as radiance

David Bentley Hart has for some twenty years pressed a picture of creation that I find largely persuasive and share in most of its parts. The world is the diffusive expression of the divine nature: the sun that, being the sun, must shine. It adds nothing to God, completes nothing in God, contributes nothing to God. The dependence runs one direction only. As he put it recently in a Substack comment, the order of ontological power is infinitely asymmetrical. The photons do not make the sun the sun.

I grant every word of that asymmetry. I have defended it myself against those who would make creation constitutive of God’s identity, and will say without embarrassment that Hart’s formulations of it are better than mine and likely to outlast them.

This Substack is reader-supported. To receive new posts and support my work, consider becoming a free or paid subscriber.

I want to be exact about him, though, because he is usually read badly on this, and, I will confess, I have read him badly on it myself. In The Doors of the Sea, the most widely read thing he has written on suffering, Hart sounds very like what I am about to argue. Being, he says there, is “the first good, the first gift of God’s gratuitous love.” Though God “had no need of us, still he loved us when we were not.”^[2] Evil, suffering, death have no ultimate value or spiritual meaning at all; they are cosmic contingencies, ontological shadows, devoid of substance or purpose, making no contribution to anything. He goes further, and puts it in words I would happily make my own: “the entire history of sin and death is in an ultimate sense a pure contingency, one that is not as such desired by God.”^[3] And he refuses, as sharply as anyone has, “the God who needs suffering and death, who chooses them as instruments of a perfect and universal design that — but for them — could not be realized.”^[4]

So let no one say Hart is a theologian of necessary evil. He is not. The passage above is one of the great refusals in modern theology, and I would rather be found defending it than quarreling with the man who wrote it.

And yet.

Notice what kind of contingency that is. Hart is speaking of the fall, of the history of sin and death, as contingent, and he grounds that contingency where the tradition usually grounds it, in the free will of rational creatures. Creatures freely turn from God; their turning need not have happened; God permits it. What he is not saying there is that the world’s existence is contingent. And when he does address the world’s existence, in an essay written on precisely this subject, the language moves in a different direction: between the ontology of creatio ex nihilo and that of emanation, he tells us, “there really is no metaphysical difference — unless by the latter we mean some gross material efflux of the divine substance, which nobody except perhaps Milton ever believed.” The same identification recurs, more forcefully, in his most recent book, where the wish to keep a strict distinction in principle between creation from nothing and a metaphysics of emanation is called “an altogether curious conceptual error.”^[5]

I really do not think this is a change of mind, and I am not going to accuse him of one. (A word in passing on how easy such accusations are to make and how rarely they are true. A writer returns to a question across twenty years, the emphasis shifts because the interlocutors have shifted, a pastoral book written in the shadow of a tsunami says what that occasion required and a systematic book says what system requires, and the reader who comes to both at once mistakes a change of lighting for a change of furniture. I have done this to Hart in conversation and been corrected, which is how I came to check.) What I find is one position held consistently, in which “gratuitous” always meant God gains nothing, needs nothing, and adds nothing to himself by creating — and never meant God might have refrained. The world is absolutely contingent upon God, in his phrase — utterly dependent, holding nothing of its own. That is a claim about dependence. It is not a claim about whether the world might not have been.

And this is the fissure I want to open, because it is not a small one.

Here’s the rub: two entirely different things go by the name “contingent.” The first means depends wholly on another for its being. The second means might not have been at all. These come apart: a thing can be utterly dependent and yet unavoidable. That is precisely how a Neoplatonist thinks of the world, wholly dependent on the One and wholly unavoidable, like the light streaming from a lamp that cannot help itself. Dependence is one axis; might-not-have-been is another.

When Hart says creatures are absolutely contingent on the infinite gift, he is saying the first with great force.^[6] He is not saying the second, and so far as I can find, he nowhere affirms it.

Which brings me to a passage I cannot pass over, since it appears to say the very thing I have been claiming Hart does not say. Writing on precisely this question, the doctrine of creation and the problem of evil, he insists that while one must avoid the pathetic anthropomorphism of imagining God’s decision to create as an arbitrary choice made after deliberation among options, one must still affirm that it is free, that creation can add nothing to God, that God’s being is not dependent on the world’s, and that “the only necessity in the divine act of creation is the impossibility of any hindrance upon God’s expression of his goodness.”^[7]

Read the shape of that. He rejects the deliberating God, exactly as I do. He affirms that creation is free. And the only necessity he admits is that divine goodness cannot be impeded. That is a real position, carefully expressed, and it is nearer to mine than the emanation language would suggest.

I can imagine the obvious reply which deserves an answer. Free means the world need not have been. What else could it mean, once the anthropomorphic deliberation has been stripped away? You have invented a distinction so as to have something to disagree with. If that reply held, my quarrel with Hart would be merely verbal, and I would rather find that out than paper over it. But I do not think it holds, and the reason is in the sentence itself.

Look at how the sentence ends. Hart does not say that there is no necessity in the divine act of creation. He says that the only necessity in the divine act of creation is “the impossibility of any hindrance upon God’s expression of his goodness”. That is not a denial that anything must happen; it is a specification of what must. Had he meant that the world simply might not have been, the natural thing to write would have been that there is no necessity in the act at all. Instead he names one, and the one he names is expression. Goodness cannot be prevented from expressing itself. A freedom compatible with that necessity is a freedom from hindrance, which is precisely what the rest of the clause describes: no external restraint, no dependence, nothing added to God. A lamp that shines without impediment shines freely in that sense and cannot help itself in the other.

The surrounding argument leans the same way. On the same page Hart tells us that God’s moral venture in creating is infinite and that all contingent ends are “intentionally enfolded within his decision; a page later, that creation is an eternal act whose term is the divine nature.^[8] A wager laid, its costs enfolded, in an eternal act terminating in God’s own nature, from which goodness cannot be hindered from proceeding: none of that is the grammar of something that might simply not have occurred.

And then the identification of creation with emanation settles it, or as near to settled as I suppose these things get. Whatever else emanation has ever meant, it has meant procession from a nature. That is the one feature that distinguishes it from making. If Hart is right that creation from nothing and emanation are metaphysically the same thing, excepting only the crude material efflux, then the world proceeds from the divine nature in whatever manner emanation proceeds, and no one has ever supposed that what emanates might have failed to emanate. Hart is entitled to a purified emanation: no principle exterior to the source, no efflux of divine stuff, no dialectical necessity, no compulsion. He is not, I think, entitled to an emanation that might not have emanated, since that would abandon the doctrine while keeping its name.

So the difference between us is not verbal. What Hart secures is that nothing constrains God. What I need, and what I do not find him holding, is that the world’s existence might not have been: that there is a real sense, however analogically we must speak it, in which nothing whatever would have been missing had God not created.

And I would not press the difference so hard if Hart himself did not show, a few pages later, exactly how much rides on it, how much is at stake.

There is one more thing to notice in Hart. It is the thing that most awards our attention, because in it he poses my own problem more sharply than I have and then answers it in a way I cannot follow.

He is considering whether the autonomy of creatures gets God off the hook: whether, if no one goes to hell except by his own industry, the divine decision to create is thereby exonerated. His answer is that absolutely nothing changes. Suppose God created on the chance that some might be lost. Then morally he has purchased the revelation of his power at that price even if in the end no one at all is damned, “because what is hazarded has already been surrendered no matter how the dice fall. The wager is itself an irrevocable intentional decision in which every possible cost has already been accepted.” And then the sentence that ought to stop any reader: “to venture the life of your child for some other end is, morally, already to have killed your child, even if at the last moment Artemis or Heracles or the Angel of the Lord should stay your hand.”^[9]

I know of nothing else in modern theology quite like that, and I am in complete agreement. The venture is the moral fact. Outcomes do not retroactively cleanse an intention that accepted the cost. Call it the venture-principle: what is hazarded has already been surrendered, and the wager is the moral fact irrespective of how the dice fall.

But now apply it where Hart does not. He has said that God’s act of creation is free and unconstrained, and that all contingent ends are intentionally enfolded within his decision. Put those together with the venture-principle and the difficulty is acute: if the horror stands among the costs enfolded within a decision that was made, then on Hart’s own reckoning it has been accepted, and the accepting is the moral fact whether or not the horror is later undone.

Hart’s answer to this is eschatological and nothing else. Eternity must, he says, reduce the price of evil to absolutely nothing; for “if anything were to be eternally lost — the least little thing — then the goodness of creation could never be more in the end than a purely conditional goodness,” and God could not be the Good as such.^[10] That is the whole of the answer, and it is a magnificent one so far as it goes: no remainder, no loss, no soul spent.

So consider what has and has not been secured. At the end, the price is reduced to nothing, and I hold this too. However, at the origin the horror is still a cost enfolded within a decision. And Hart’s own principle says that enfolding is morally decisive independently of the outcome. The wager is the moral fact, whether or not the dice fall well. He has made evil answer to nothing at its finish while leaving it, by his own account, among the accepted costs at its beginning.

And he cannot reach the origin with the resources he has left himself. He has secured the finish: at the end, nothing is lost. But the origin is untouched, and an eschatology, however complete, works only on the finish; it cannot unmake a cost that was accepted at the start, because that is exactly what his own venture-principle forbids. What would reach the origin is the claim that no wager was ever laid, that the world’s existence was unentailed, so that the horror was never among the costs of any decision, there being no decision that had to be made. But that claim is the one his account of unimpeded expression, his eternal act terminating in the divine nature, and his identification of creation with emanation have between them removed. The eschatology is being asked to do the work of two answers, and it can only do the work of one.

There is one route left, and I want to bring it up because I think it is where Hart would actually go. He might say that the question is malformed: that “might not have been,” said of God, imports a creaturely picture of alternatives and forks and roads not taken, and that he is not denying the world’s contingency so much as declining a badly-posed question. That truly is a serious reply and certainly not an evasion. It is what anyone fully committed to the analogical tradition should say, if he means it.

But then the burden is his, and it is a heavy one. No one writing today holds the analogical tradition more rigorously than he does. He holds that “good” said of God, and “infinite,” and “wise,” all cross the interval and arrive with real content once their creaturely contrast has been stripped away. So he owes an account of why this predicate alone fails to cross, why “might not have been” is left with nothing when the deliberative content is removed, while the others are left with everything. I will try to show, in a moment, that something perfectly definite does survive the crossing. If it does, the question is not malformed, and declining it is declining an answer rather than a confusion.

There is another, smaller thing worth noticing on the way which shows the same pattern in miniature. Having said that evils are cosmic contingencies and ontological shadows, intrinsically devoid of substance or purpose, Hart adds, in the same breath, that God may “under the conditions of a fallen order, make them the occasions for accomplishing his good ends.”^[11] There it is again. Emptied of purpose, then made an occasion. It is the gentlest possible form of the move, and it is still the move.

The same structure governs his account of permission. Having removed evil’s substance by privation, Hart must still say why God permits it, and his answer is this: “the union of free spiritual creatures with the God of love is a thing so wonderful that the power of creation to enslave itself to death must be permitted by God.”^[12]

Must be permitted.

Read that slowly, because a great deal is folded into it. The horror is not called good, and it is not justified in any crude sense; but it is permitted for the sake of something, namely the wonder of free union, and a horror permitted for the sake of something is a horror that answers to that something. Whatever else such a horror is, it is not waste. Its permission carries a rationale, and the rationale is a good that the permission makes possible. That is a free-will theodicy in miniature, which is precisely what “Returning the Ticket” dissolved and what I refuse here. Privation gets evil out of God at the level of being; permission lets it back in through the side door, at the level of purpose.

I can hear the reply in the form of an objection that has pressed me hardest throughout this argument. The reply states that you are conflating two things that must be kept apart. No one says the child’s actual agony serves a purpose. What is said is that God’s permitting the possibility of such agony is justified — permitted, not willed; the possibility, not the actuality. One may permit what one categorically opposes. To license the possibility for the sake of free union is not to assign the child’s particular horror a use. Moreover (this is the sharper form of the distinction) the actual privative evil, when it comes, is entirely the creature’s own doing, not God’s. God permits the possibility; the creature realizes the actuality. What is conditionally necessary on this account is the permission, not the suffering. God no more needs the child’s agony than a father who gives his child genuine freedom needs his child to abuse it. The distinction is real, it is important, and I grant that without any reservations.

But watch what the permission does once it is granted for the sake of something. If God permits the possibility of the horror because free union is a good so wonderful that the power to fall must be allowed, then the actual horror, when it comes, arrives as the realization of a possibility that was licensed for that good. It is covered, retroactively and in advance, by the rationale that licensed the possibility. The permission does not float free of its instances. Its whole justification is that the good it secures is worth what it lets in, and what it lets in is precisely the actual agony, not the bare abstract possibility, which purchases nothing by remaining a possibility. A possibility permitted for the sake of a good is a possibility whose actualizations are already accounted for in the ledger that justified permitting it. So when the scream comes, it does not answer to nothing; it answers to the worth of the union for which its very possibility was admitted. The child’s actual suffering answers to that good — not because God assigned it a use, but because the permission that admits it was itself justified by a use, and the actual cannot be cheaper than the possibility that was priced to include it.

This is why the possibility/actuality distinction, real as it is, does not rescue the permission from the second refusal. Ivan is not confused about the difference between permitting and willing. He is refusing a world whose goodness was reckoned worth the admission of the scream. And a permission justified by a good is exactly such a reckoning. The only account under which the scream answers to nothing is one in which no good was ever purchased by admitting it: where the world itself, and so the whole space in which the possibility could be permitted, need not have been at all.

And here Hart has handed me the blade himself: for it is he who says that to venture the life of your child for some other end is, morally, already to have killed your child, even if at the last the hand is stayed. The permission is the venture; the venture is the moral fact; and a permission laid for the sake of a good has already accepted the scream among its costs, exactly as the wager has already surrendered what it hazards no matter how the dice fall. What I add is only the one move Hart does not make: that a venture is a moral fact because a decision was made that need not have been made, and where nothing had to be ventured, nothing was.

The second account: creation as incarnation

Jordan Daniel Wood’s proposal is deeper and stranger than either, and I take it to be the most interesting Christian metaphysics written in my lifetime. Created being, on his account, is the Word’s own self-utterance rather than a thing made alongside God: creation is incarnation, the eternal Son’s kenotic self-expression extended outward, so that there is no possible world without the Incarnation, and at the limit no world that is not already the Son’s own person in a creaturely key. The identity is hypostatic and not natural. Wood is emphatic on this point, and I want to be fair to him about it: the creature does not become God by nature, and the unity he is describing is a unity at the level of person.^[13]

His account of evil is neither privation-and-permission nor necessity, but freedom followed by providence.

Evil, for Wood, is the creature’s own fabrication: it is our personal freedom, he writes, that renders the created world a fallen world, since particular persons, as Wood writes, “freely prefer fantasy to the Word”^[14] and in preferring it lend reality to a way of being that is not God’s creation at all, a world that is, in his fine phrase, not yet God’s creation. Nothing here is necessitated, and it would be a caricature to say that Wood makes evil necessary; I have made that mistake in conversation and withdraw it.

Then comes the redemption, and it moves in two directions at once. In the one direction there is annihilation: following Gregory of Nyssa, Wood writes that “the final restoration will make the memory of evil things utterly vanish,” that the false world will “dissipate into the nothing that it is,” and that evil is a “positive nothingness not reckoned to exist at all.”^[15] This is the strand I love in him, and it is very nearly what I want to say myself. In the other direction there is providence: Christ, in Wood’s words, “endowed the very false principles our sin falsely incarnates — the law of death itself — with the deeper principle of providence,” bringing about “the complete deification of even this universe and of the ‘me’ I make in vain.”^[16]

Even the me I make in vain. Read that as slowly as the sentence from Hart, because the vainly-fabricated self, the false world, and the law of death are not merely destroyed here. They are endowed with providence, given a deeper principle, made to serve the complete deification of the universe.

And there I must part from him, though it takes some care to say why, because Wood has an answer ready and it is a good one. He would say: you have set up a false opposition. Incorporation is annihilation’s mode, not its contrary. The false principles are destroyed precisely by being endowed with providence; evil does not survive as evil inside the providential whole, it is unmade by being taken up. Annihilation through incorporation, not alongside it.^[17]

That is the strongest move available to him, and it does dissolve the tension as I first stated it. But it does not reach the objection underneath, which is not about whether evil survives but about whether it serves. A thing can be wholly destroyed and still have been used on its way out of existence.

Consider a timber thrown into a furnace. It is annihilated; nothing of it survives as timber; sift the ashes and you will find no beam. And yet the house was warm that night, and it was warm because the timber burned.

The timber was not preserved. It was spent.

Whether a thing survives and whether a thing serves are two questions, not one, and a horror can be perfectly annihilated and still have functioned, on its way out of existence, as the fuel of something.

That is what troubles me in the providence-strand. If the child’s agony is endowed with the deeper principle of the universe’s complete deification (if it is taken up and made to serve that end) then the agony has done work. It has heated the house. And a horror that heats the house is a horror the house could not, in that respect, have done without. Which is exactly what Ivan will not accept, since his objection is not to the resolution at all, only to a world in which the scream was used.

Wood will say, rightly, that he never calls the horror good, and I believe him; but the second refusal does not turn on whether the horror is called good, only on whether it served. And “endowed with the deeper principle of providence” is service, however reverently the service is described.

The shape of the problem

What we have, stated plainly, is this: the two accounts converge in a way that is easy to miss.

Both men are far too serious to say that evil is justified. Both remove evil’s substance, Hart by privation (evil is a lack, not a thing), Wood by annihilation (evil dissipates into the nothing it is). And then both, having emptied evil of being, give it back a use. Hart: the horror is permitted for the sake of the wonder of free union. Wood: the horror is endowed with providence for the sake of the complete deification of all. Neither says the child’s suffering was good, and both say, in different registers, that it was not for nothing.

That last is what I want to deny, not that evil has no substance, since I hold the privation doctrine as firmly as either of them, but the second move, the one that comes after. I deny that the horror, having no being of its own, nonetheless has a place. My claim is that this (the privative horror, not the formative cost) has no place, no function, no role, and no service, and that it answers to nothing at all.

A word about what I mean by evil, since without it the claim I am about to make will seem to prove far too much. I do not mean every pain, every loss, every hard passage of finite becoming. The scraped knee of a child learning to walk, the ache of a muscle growing strong, the griefs by which a soul is enlarged — these are the costs of a being that comes to itself in time, and they are not privations of anything; they are the price of a good, and they are not waste. By evil I mean privation proper: the surd, the turning-from, the cancer in the ward, the affliction that enlarges nothing and ripens into no good the sufferer could be said to gain. It is of this, and only this, that I say it answers to nothing. One may hold (I do) that the formative sufferings of a creature learning to be are gathered up and lose nothing; the question of this piece is whether the privative horror, which is no one’s formation and no one’s price, has likewise a place. I say it has none.^[18]

And that claim, I will now argue, requires something neither of them holds: that the world’s existence is genuinely contingent, that all of this, the whole order in which the horror occurred, might simply not have been.

Where contingency actually comes from

I have to be careful now, because there is a wrong way to say this that would wreck everything.

The wrong way is to say: God surveyed the possible worlds, weighed them, and chose this one, so it is contingent because he might have chosen another. That is a picture of God as an enormous deliberator with a catalogue, and it is a picture I reject as firmly as Hart or Wood do. It makes creation an arbitrary selection and God a chooser among options, which is the creaturely kind of freedom writ large, and not the divine kind at all. Worse, for my purposes, it is disastrous for the child: a God who picked this world, horrors included, from a shelf of alternatives is a God who elected the horror, and I have gained nothing.

So I want the contingency without the catalogue: the world might not have been, and yet God did not pick it out from among rivals. That combination sounds impossible, and the sense that it is impossible seems to me the deepest reason thoughtful people slide toward necessity, since if you must have one or the other you will take the one that does not make God a shopper.

It is not impossible, and allow me to explain why.

Ask where contingency comes from in the ordinary case. A chair might not have existed, and the reason it might not have is that a carpenter chose to make it and might have chosen otherwise; in our world “might not have been” and “was chosen among alternatives” go together so reliably that they look like a single thing. That is the whole trouble, because for creatures they very nearly are a single thing. A creature has only two ways of acting, deliberating and choosing, or emitting by nature as the sun emits its light, so that whatever a creature does not choose follows from it necessarily. For us, then, “not chosen” means “necessitated,” and contingency really does depend on election.^[19]

Now transpose that to God and watch what happens. If God has only those two modes, then denying that creation is a choice among options (which everyone in this conversation does deny) forces the other option: creation follows from the divine nature as light from the sun, necessarily. That is exactly the pressure that produces the emanation-picture. This is no mistake. It is the honest working-out of a hidden assumption: that the only alternative to choosing is being necessitated.

That assumption is the thing I deny, and my denial is a doctrine, not a dodge.

Consider what actually differentiates the two great processions the Church confesses. The Father eternally begets the Son, and God creates the world, and both are “from God.” So what makes them different?

The tempting answer is a modal one: the first is necessary, the second free. But that answer secretly imports our creaturely categories into God, and it will not survive scrutiny, since “necessary” and “free” in our mouths mean things drawn from deliberation and compulsion.

The Church’s actual answer is not modal at all. It concerns what comes forth. In the begetting of the Son what proceeds is God — the divine essence communicated entire, the Son homoousios with the Father, of the very same nature, God from God, not a lesser thing derived from a greater. In creation what proceeds is not God — the creature receives finite existence, which is not the divine essence, and holds it as something received. The Son is God. The creature participates God. That difference lies in what stands at the far end of the acting, not in how God acts.

Everything I need follows from that, and follows without smuggling a single creaturely modal concept into the divine life.

Because the Son’s procession terminates in God himself, it cannot fail or be otherwise, not because “necessity” is some property we have measured and assigned, but because for God to be God simply is for the Father to utter himself wholly in the Son. A self-expression that is identical with the one expressing cannot fail to occur. There is no gap in which failure could happen.

And because creation’s procession terminates in what is not God — in finite existence, which is not what the divine plenitude is, and nothing about God’s being God entails it. The divine self-expression is already whole and infinite in the Son and in the Spirit’s communion; it does not need to reach past the eternal processions to be complete. Only God can express God as God, and that is done, eternally, without us. So the world is overflow, not completion; and an overflow from what is already complete is precisely a gift, given and not owed.

That is where the world’s contingency lives — not in a divine deliberation, and not in a rejected catalogue of alternatives, but in the status of what was given. The creature holds its existence the way a gift is held, in an open hand: received rather than owned, participated rather than possessed. Its essence does not include its being. It is the kind of thing that can not-be, because it does not have being of itself. That is the whole content of saying the world might not have been. The phrase makes no claim about God’s decision-making. It claims something about the kind of being a creature has.

Two grounds, then, in two different places. Election, being chosen among options, is a matter of how an agent acts. Contingency, might-not-have-been, is a matter of what the acting produces and how that product holds its existence. In us these two share a single ground, and so they seem to be one thing. In God they come apart, because the contingency is grounded in the received and participated character of the not-God standing at the end of the act, and not in any choice.

Saying “might not have been” without saying too much

Those with an eye for the tradition’s care about God-talk will see a difficulty here, and meeting it head-on is what keeps my argument from collapsing into the very thing it rejects.

The difficulty is this. The whole case of “Returning the Ticket” was that our words do not apply to God in the sense they bear among creatures, that “good,” said of God, does not mean what it means when said of a neighbor, only larger. That is why theodicy fails: it weighs divine goodness on a creaturely scale. But if that is true of “good,” it must be true of “might not have been” as well. There are no exemptions. A word cannot be barred from univocal use when it would embarrass theodicy and then admitted through the back door when it is needed for a theodicy’s refutation. So what can “the world might not have been” possibly mean, if it does not mean what it means of the chair and the carpenter?

The answer is that it means what every other word means when it is said of God and his works: it reaches toward its object without capturing it, and it sheds the creaturely content that would distort it.

Watch what happens when other words make that passage. To say God is good is not to say that he is good rather than evil, as though good and evil were a scale along which he had scored well and might conceivably have scored otherwise; divine goodness is the source of the distinction between good and evil rather than a member of it, and the contrast lies beneath him, grounded by him, not ranged alongside him. To say God is infinite is not to say that he is infinite rather than finite, as though finitude were an alternative he had narrowly escaped; divine infinity is that of which finitude is a participation, not its contrary. In each case the word arrives stripped of the contrast it carries among creatures, and what remains is more, not less.

Now run “might not have been” through the same passage. Among creatures the phrase carries a contrast: this rather than that, chosen from a field, the road not taken. Strip that — and it must be stripped, since the field of unrealized possibilities God surveyed and declined is exactly the catalogue we have already refused — and what is left is not nothing. What is left is: the world is not entailed by God’s being God. Its existence follows from nothing in the divine nature. The plenitude that would have to be short of something in order to require the world is short of nothing, and is whole without it. That is the content, and it is entirely positive, and it says nothing about any deliberation.

Someone will press further: you have leaned on “whole,” on “complete,” and completeness is a contrast-word too. Complete rather than lacking. Have you not smuggled back in, at the foundation, the very thing you disallowed?

The same discipline answers it. Divine completeness is not completeness-rather-than-incompleteness, as though God had escaped a deficiency he might have suffered. It is the plenitude that is the source of the distinction between fullness and lack, and lack lies below it as finitude lies below infinity. And when you ask what that completeness is, the answer is the eternal processions themselves, and no quantity of being at all: the Father’s whole self-utterance in the Son, the Spirit’s communion, God expressed as God to God without remainder. That is what “complete” cashes out to. It is not a modal claim at all. It is the doctrine of the Trinity.

So the argument bottoms out where I said it would. The world’s might-not-have-been rests on the world’s not being God; and what makes it not-God is that the Son is God of the very same nature, homoousios — and the creature is not. That distinction is not a further concept needing a further ground beneath it. It is the difference between the source and what participates the source, which is the shape of every analogical statement anyone has ever made about God. You cannot dig beneath it, because it is the floor on which all the digging has been done. If that is not bedrock, nothing in Christian speech is.

And I want to be clear that none of this is a private metaphysical construction I am asking anyone to buy on my authority. The might-not-have-been is not an inference I have contrived; it is the content of the Christian distinction itself — what the Church has always meant by confessing that God made the world from nothing, ex nihilo. Robert Sokolowski put it as well as anyone: “the God of Christian faith is understood as one who could have been, in undiminished goodness and greatness, even had the world not existed.”^[20] That is not a metaphysical thesis laid on top of the doctrine of creation; it is the doctrine of creation, stated as a distinction. God plus the world is not greater than God alone. The whole preceding argument has been an attempt to show what that distinction means and why it is coherent. But the distinction is given in the confession, not deduced from a theory of divine choosing. Which is why the might-not-have-been cannot be dismissed as an idle modal flourish without dismissing the Christian distinction along with it. To deny that the world might not have been is to deny that God would have been God without it, and that denial the Church has never made.

The idle-contingency objection

The strongest challenge to this comes from neither of the accounts above. It comes from a friend who has pushed me harder on it than anyone, and it deserves a direct answer rather than a deflection. It runs like this.

Grant that God’s determination to create is not a choice among options. Grant that it is eternal, sufficient, non-elective, and that nothing further is needed for the world to exist. Then what work is “contingent” still doing? The Son depends on the Father and is not thereby contingent. If an eternal, sufficient, non-elective determination stands behind the world, why is the world any more contingent than the Son? You have called its contingency real, but perhaps it is idle — true when the creature is considered in the abstract, empty once you notice that a sufficient eternal act infallibly produces it.

This is the right question, and I have no answer that makes it go away quietly.

The disanalogy with the Son is the terminus. The Father’s determination terminates in one who is God — which is why its sufficiency and the Son’s unfailingness are the very same fact. There is no distance between the expressing and the expressed in which contingency could arise, because the Son is not given being; he is the divine being, eternally, of the same substance. In contrast, the creative determination terminates in what is not God — in being that is received, held, participated. And here the sufficiency of the act does not swallow the contingency of the effect, because the effect’s contingency is not a shortfall in the cause. It is the mode of the effect. The creature is given being, and givenness is the very mode of contingency. So the parallel fails at exactly the point where it was supposed to bite.

The creature’s contingency, on this account, is not a speculation about what God might have done. It is a truth about what the creature is: a thing whose being is on loan, which therefore stands at every moment of its existence in the condition of not having had to be. That is as real as the creature itself, and nothing about the reliability of the giving reaches it. You cannot argue a gift into an entitlement by pointing out how surely it was given.^[21]

I will not pretend this closes every question. Whether received being carries genuine and not merely notional contingency is the joint on which the whole argument turns, and it is the place I would most want to be pressed. But I do not think it is idle. If it holds, the world’s might-not-have-been is not a courtesy title but a real fact about the kind of thing a creature is.

What follows for the child

Now the payoff, and it comes in two movements, because the horror has two ends: a beginning and a finish. It must answer to nothing at both.

At its beginning: the child’s suffering was not necessary. Not the world in which it happened, which need not have been at all; not the order in which it occurred, which was owed to no one and to nothing; not the child herself, whom God gave freely and need not have given. There is no arrangement of things anywhere that required her agony. It fills no slot.^[22] It was not the price of anything, nor the condition of anything, nor the dark against which some light was to be seen. It simply happened, in a world that did not have to exist, to a creature who did not have to be made.

And it should not have happened.

Before arguing why this is available to me and not, so far as I can see, to either of the accounts above, I need to meet the sharpest form of the objection, because without it the argument looks easier than it is. The most precise version of the opposed position does not say that the world’s existence was absolutely necessary — necessary in every possible world, as though God had no choice whatever. It says something more careful: that given the end for which God creates, the permission of evil is conditionally necessary to securing that end. Grant that creation is contingent, the objection runs, and any necessities within it are conditional. The only necessity we need attach to God’s permission of evil is this: should God create for the deifying union of all rational creatures, then the permission of evil is necessary to reaching that end. Call it conditional necessity, and it sounds modest enough.

But conditional necessity is sufficient for Ivan’s complaint, and that is the thing to see. He does not need the horror to be necessary in every possible world. He needs it to have been necessary in this one. In the world God actually made, the one in which the child was in the outhouse. And on the conditional-necessity account, it was. Given that God created for this end, the permission of her suffering was necessary to it. The conditional qualifier restates the felix culpa more precisely; it does not dissolve it. “It had to be, in the world that was made” is enough to return the ticket.

So the question becomes acute, and it is the right question: did the end require the permission? Could God have secured universal deification(the final union of all rational creatures with himself) without the permission of evil as its condition?

I believe he could, and the reason is already in the argument. The deiform end is secured not by the conditions under which the creature exists but by the nature God gives it. A rational spirit is necessarily ordered to deifying union — neither I nor Hart nor Wood would deny this; it is the heartbeat of universalist metaphysics. That necessity is in the creature’s essence: given that God creates a rational spirit, its end is as certain as the square circle. But that necessity says nothing about whether evil must be permitted on the way. Evil is not what makes the creature deiform; the creature’s nature does that. And this comports with the dominant tradition’s Christology, which holds the Incarnation itself contingent on the fall: had there been no sin, there would have been no Incarnation, but no less deification either, since the creature was always destined for deifying union by its nature. The permission of evil is the condition of redemption, not the condition of deification. Redemption presupposes what it redeems; deification is wider than redemption. What evil does is make the path take a certain shape, one that requires redemption rather than simple completion. But the end itself was never waiting for the evil as its necessary occasion. Grace meets genuine waste and brings the creature to its end; it does not need the waste in order to do so.

The difference is the difference between two parents. A parent who says “my child needed to suffer in order to become who she is” has made the suffering necessary to the flourishing. A parent who says “I gave my child life, and suffering came, and grace met it” has not. The first is theodicy in its mildest form. The second is grace operating on waste — which is an entirely different, and I would say more glorious, claim.

And consider why this is available to me and not, so far as I can see, to either of the accounts above. If the world is the radiance of the divine nature, then the order in which the horror occurred is the necessary shining of God, and the horror is a shadow within that: a defect in what God unavoidably is. And if created being is the Son’s own self-utterance, then the horror falls within the Word’s own outward life. Either way there is a tether, and the horror is attached to something that had to be. On the account I am defending, the tether is cut at the root: the creature is not God, its existence is not entailed by God’s being God, and so the horror that befell it is entailed by nothing whatsoever.

At its finish: the horror is over, and over absolutely.

This second half needs saying because a suffering that was unnecessary at its origin could still be retained forever at its end, and then it would not answer to nothing after all, it would answer to whatever it is retained for. The worry was put to me precisely: if all created history is eternally present to God, then the child’s agony never becomes absent from the divine present. It is eternally contextualized by glory, but never ontologically left behind. And a horror eternally retained under glory is a horror that has been kept.

I share the worry entirely; it is my own objection to providence-that-incorporates, in a different key. And the answer lies in a distinction that is easy to miss.

That God knows a suffering is not the same as God sustaining someone in it. Aquinas is right that God conserves creatures in being by his knowing and willing — nothing exists except as held in that act. But what matters is not whether God knows-sustains the creature; it is in what condition he sustains them. To know a wound that has been healed is not to keep on wounding. God sustains the healed one, not the wound. The scar is known; the cutting is over. What would make evil ongoing is divine upholding of a creature in the condition of suffering — some perdition still maintained in being, some creature still held in that loss.

And that is precisely what universal restoration removes. If there is no one finally lost, then there is no one being sustained in loss: no subject persists in the horror, no perdition is upheld, and the surd, which never had substance of its own and now has no creature persisting in it, simply is not, finally, and without residue. It is known as having been and having been borne, not retained as occurring.

This is where an eternal hell makes the problem unanswerable. On that view evil is maintained forever, and not merely remembered: God himself holds the damned in being within their loss, world without end. That is evil never ontologically past in the strongest imaginable sense, not a scar but a wound perpetually reopened by the one who could close it. The worry that evil is never finally over is not a difficulty for universalism but a description of infernalism.^[23]

But universalism alone will not finish the job, and one piece of the argument remains. Suppose evil was a defect within an order that had to be, a shadow of the necessary radiance, a moment in the Word’s necessary outward life. Then even with every creature restored, the horror was still a feature of what could not have been otherwise. You can annihilate the effects; you cannot make it the case that the necessary order was not the sort of order in which this happened. Something remains. No person in torment, certainly, but a fact about the structure of things, and a fact that had to be.

Only if the world need not have been does the horror become fully nothing. Then there is no necessary order for it to have been a feature of. It was an event in a world that was gift, befalling a creature that was gift, and when the creature is healed and the surd dissolved, nothing whatever is left, not a residue, not a structural scar, not a note in a score that had to be written, because there was no score that had to be written.

So the two halves hold each other up: because the world might not have been, the horror answers to nothing at its origin, and because all shall be restored, it answers to nothing at its end. Neither alone would do, since a necessary world with a universal restoration gives you a horror that is over but had to be, while a contingent world with an eternal hell gives you a horror that need not have been but is never over. Only the two together give what Ivan is owed: it did not have to be, and it is finished.

But does this not make God worse?

Another objection will occur to any serious reader, and it is an objection I cannot sidestep.

It runs like this. You have made things worse, not better, for on your account God freely gave a world he need not have given, knowing what would be in it, and the suffering in it serves nothing at all, not even a greater good. God permitted pointless horror for no reason whatever. At least the theodicist’s God had a purpose; at least the God who permits evil for the sake of free union is doing something, whereas your God permits waste. What kind of love is that? Have you not simply replaced the God who needs evil with a God who tolerates the intolerable for no reason, and is that not exactly the God to whom the ticket should be returned?

What I believe, and will state without hedging, is this: a God who had a reason for the child’s agony is more terrible than a God who had none.

Consider what having a reason would mean. It would mean the suffering was for something, that it purchased a good, enabled a union, made possible a glory, served a deification. It would mean that when God looks at the child in the ward, he sees, among other things, what her suffering is accomplishing. And that is the God Ivan hands back his ticket to. Not the God who is powerless. Not the God who is absent. The God who is using her. Every theodicy, however refined, ends by saying: it was not for nothing. And Ivan’s answer, which I take to be the beginning of Christian wisdom rather than its enemy, is that a good purchased with that coin is not worth having, and that a God who spends children in that currency is not worth worshiping.

So between a God who permits an unnecessary horror and a God who assigns it a use, I do not think the second is the kinder; I think the second is the accusation itself. The whole force of the desolating claim (i.e. that it answers to nothing) is that it is the only claim under which God is not the one who spent her.

There is a more sophisticated form of the objection, and it deserves its own answer. It runs: you have set up a false scale. The question is not whether the horror answers to something — of course it does, in the sense that a wise God only ventures a risk whose end outweighs it. The question is whether the end is incomparably worth the risk. And it is. Compare the actual horrors of all history to the infinite, ecstatic, ever-increasing beatitude of final union with God — there is no comparison. The math is not close. So the God who ventures the world, knowing what will be in it, and redeems every loss infinitely, is not a monster but the most loving being conceivable. Do the math.

The incomparable worth of the end is not in dispute. What is in dispute is this: can a horror be incomparably outweighed and still answer to nothing? I do not think so. When the horror enters a ledger — even a ledger in which one side is infinite and the other finite — it has answered to something. The child’s suffering has been weighed. It came out on the lighter side, infinitely lighter, but it was on the scale. And a suffering that has been weighed against the worth of the end is a suffering that served the calculation, however briefly and however decisively it lost. Ivan does not return his ticket because the math comes out wrong. He returns it because he refuses to do the math at all. Not because the end is unworthy — he grants the end — but because any end that required the child’s admission to the ledger has already done the thing he refuses: it has made her suffering answerable to something other than her.

The only account under which the horror enters no ledger is one in which no ledger had to be opened. Where the world need not have been, no risk had to be ventured, no cost had to be weighed, and the horror answers — genuinely, ontologically — to nothing at all.

And “answers to nothing” is not “God is indifferent.” This is the distinction on which everything hangs. To say the suffering serves no purpose is not to say God does not care about it. It is to say God’s love for her is not conditioned by what her suffering accomplishes. He does not love her because her agony contributes; he loves her though it contributes nothing at all. That is a greater love and not a lesser one, since a love that requires the suffering to be productive is a love with an interest in the suffering.

One further clarification, because without it a familiar thought will quietly undo the argument. Good things have come from suffering — beauty, courage, patience, mercy, art that could not have been made by someone who had never known devastation. All of this is real, and I do not want to deny a syllable of it. But there are two entirely different claims folded into that observation, and they must be kept apart.

The first: good things follow from suffering. Sequitur. They come after it, shaped by its presence — which is to say, redemption meets them there. This is simply true and no one disputes it.

The second: good things follow because of suffering, in the sense that the suffering was their condition. Suffering was the occasion they needed, the crucible that produced them, the thing without which they would not have been. Propter. On this account, redemption was not meeting the suffering; it was waiting for it. This is the felix culpa in its mildest form, and it is the claim that the horror answered to something after all, namely the goods it made possible.

These are not the same claim, and only the first is warranted. Consider a master artist working with damaged material — wood that has been split, stone with a fault running through it. She produces something beautiful from it, and the fault is visible in the work. But we do not conclude that the damage was necessary to the beauty, or that she needed the material damaged. We conclude that she is extraordinary: that she can work with anything. The beauty testifies to the artist’s power, not to the damage’s productivity. Had the material been undamaged, a different beauty would have existed. Nothing would have been missing.

So it is with grace. The beauty and virtue that emerge from suffering testify to grace’s inexhaustible creativity — its refusal to be finally defeated by anything, its power to work with whatever it finds. They do not testify to the suffering’s productive role. Grace is never at a loss for material, because grace is infinite; it could have worked with anything, and would have. The particular beauty that emerged from particular suffering is precious and real. It is not evidence that the suffering was needed. It is evidence that grace is never defeated.

The good that comes from evil testifies to grace’s power over waste. It does not testify to the waste’s indispensability. Hold those apart and you can affirm everything (the paintings, the poems, the modes of virtue that suffering occasions) without giving the horror a purpose it never had. The suffering still answers to nothing. Grace answers to everything.

Nor is this a God who stands off from the waste and permits it from a distance. The whole burden of “Returning the Ticket” was that the God revealed in Christ enters suffering instead of justifying it. Here the same thing must be said again and more strongly, because the homoousion that holds the creature apart from God is what makes the Incarnation astonishing rather than obvious. The Son whose procession terminates in God, who need never have touched the not-God at all, takes the not-God as his own: the participated, defectible, dying flesh of the creature, and bears in it an affliction he did not author, did not require, and did not need.

And when he bears it, he does not explain it: on the cross he tells us nothing about what the suffering is for. He cries out and asks why he has been abandoned. That cry is not a lapse in the story but the truest thing in it: a suffering that answers to nothing, uttered by God.

I do not think there is any consolation in this beyond that one, and I will not manufacture more. The child’s suffering was not necessary. It was not useful. It is not now being kept. It should not have been, it need not have been, it is finished, and God did not stand outside it.

That is not a theodicy. It is what is left when theodicy has been refused twice over, and it is enough, I think, though it answers Ivan nothing. It stands where he stands, and points to one who stands there too, with his own body in it.

But the wounded God does not stay only wounded, and here the last thing must be said, because it is what keeps solidarity from being merely company in the dark. On the third day the tomb is empty, and what the empty tomb declares is not that the suffering was worth it after all. It declares the opposite. Christ is risen from the dead, trampling down death by death. Notice the grammar of that ancient acclamation, which the Church has sung for longer than it has argued anything. Death is not enlisted. It is not befriended, or repurposed, or shown in hindsight to have been a hard necessity serving a good end. Death is trampled by death — used against itself, made to consume itself, and in consuming itself shown to have had no substance of its own to begin with. The resurrection is not evil’s promotion into meaning. It is evil’s exposure: the revelation that the thing which seemed to hold the child in its grip had only the borrowed being it had commandeered, and could not hold even that against the fullness it tried to extinguish.

This is why “answers to nothing” is not, in the end, a claim I have to win by argument. It is the verdict Pascha renders. I have been reasoning forward all along (from what proceeds in the generation of the Son, from what a creature is, from the homoousion) and never upward from a weighing of the child’s suffering to a conclusion about God; and this is where that discipline arrives. I do not first establish that the horror was useless and then look to the cross for confirmation. I look to the cross, and to the empty tomb beyond it, and read off from the event what evil must have been, given what was done to it on that first Easter Sunday. A God who had needed the horror — who had spent it, banked it, made it the coin of some purchase — could not trample it. He could only cash it. The triumph is real only because evil was never an ally. Evil was never necessary, never useful, never an acceptable cost of anything. That the tomb is empty means the account was never kept.

So the child’s suffering was not necessary. Her suffering is not now being kept, and it did not have to be, and it is finished. The God who tells us so does not tell us from the far side of the interval. He tells us from inside the not-God, from within the dying flesh he took as his own and carried through death into life, having used it for nothing, having needed it for nothing, and having, in rising, left nothing of it behind. That is not a theodicy. It is the refusal of theodicy carried past the cross to the morning after it, and finding there not an argument, and not now only a wounded God, but a risen one, with the marks still in his hands.

---

This essay leaves one question deliberately open: whether creaturely contingency, grounded in the not-God terminus of the creative act, is genuine and not merely notional once the reliability of the giving is granted. It is the joint on which the whole argument turns, and it deserves a fuller treatment than a theodicy essay can carry. A follow-up will take up that question directly, working through the metaphysics of participated being and the analogical structure of creaturely existence in more detail than is possible here. If the argument of this piece interests you, that is where it goes deepest.

---

Enjoyed this essay? Please subscribe to support my work.

^[1]The full case is in “Returning the Ticket,” to which this essay is a companion rather than a sequel. The long version is now freely available to all readers. It answers the justification-complaint and this essay the necessity-complaint; neither justifies, and both are refusals. They are meant to be read as two halves of a single argument, and each is incomplete without the other.

^[2]David Bentley Hart, The Doors of the Sea: Where Was God in the Tsunami? (Eerdmans, 2005), ch. 2, “Divine Victory.” Hart’s formulations there are unambiguous: being is “the first good, the first gift of God’s gratuitous love,” and though God “had no need of us, still he loved us when we were not.” I quote them at length because they are routinely forgotten by readers (myself among them) who have absorbed the later Hart and read the earlier through it.

^[3]Doors of the Sea, ch. 2. The full clause runs that “as God did not will the fall, and yet always wills all things toward himself, the entire history of sin and death is in an ultimate sense a pure contingency, one that is not as such desired by God.” Note the scope: the history of sin and death is the pure contingency. This is not the same as saying the world’s existence is contingent, and the difference is the whole subject of this essay.

^[4]Doors of the Sea, ch. 2, refusing the God “who has need of suffering and death, who ‘chooses’ them as the instruments of a perfect and universal design that — but for them — could not be realized.” I know of no sharper refusal of the felix culpa in modern theology, and nothing in what follows is meant to obscure that Hart made it.

^[5]David Bentley Hart, “The Devil’s March: Creatio ex nihilo, the Problem of Evil, and a Few Dostoyevskian Meditations,” in Creation ex nihilo: Origins, Development, Contemporary Challenges, ed. Gary A. Anderson and Markus Bockmuehl (University of Notre Dame Press, 2017), 302; and The Light of Tabor: Toward a Monistic Christology (University of Notre Dame Press, 2025), 78, where creation from nothingness is called “the most radical and rigorous formulation of the metaphysics of emanation imaginable.” The 2017 statement matters more than the 2025 one for my purposes, since it comes in an essay devoted to creation and the problem of evil: this is not a late speculative turn but a settled position, held for at least a decade in precisely the context where its consequences for evil would be felt. In fairness, Hart’s emanation is a purified emanation — he argues that creation from nothing corrects the emanationist tradition by removing any principle exterior to the divine source, and would say his version is therefore more compatible with divine freedom, not less. Nothing I argue rests on the word “emanation” by itself.

^[6]Light of Tabor, 46: creation is “the total donation, out of the infinite plenitude of divine Being, of both essence and existence to creatures that are absolutely contingent on that infinite gift.” That “absolutely contingent” is doing dependence-work, not modal work (i.e. the creature holds nothing of its own) which is precisely why it can sit in the same paragraph as the emanation language without contradiction.

^[7]“The Devil’s March,” 300. The sentence is worth having whole: “while one must avoid the pathetic anthropomorphism of imagining God’s decision to create as an arbitrary choice made after deliberation among options, one must still affirm that it is free, that creation can add nothing to God, that God’s being is not dependent on the world’s, and that the only necessity in the divine act of creation is the impossibility of any hindrance upon God’s expression of his goodness.” I take this seriously as evidence that Hart wants what I want; my argument is that unimpededness will not do the work that might-not-have-been does.

^[8]“The Devil’s March,” 300–301. The claims are that God’s “moral ‘venture’ in creating is infinite”; that “as God’s act of creation is free, constrained by neither necessity nor ignorance, all contingent ends are intentionally enfolded within his decision”; and, on the following page, that “as creation is an eternal act, its term is the divine nature.” No one of these by itself settles the question of creation’s modality. The three together, joined to the emanation identification, make it very hard to hear “free” as “might not have been.”

^[9]“The Devil’s March,” 314. The passage warrants reading in full, since Hart is there dismantling the free-will defense of hell by an argument that generalizes further than he takes it: “what is hazarded has already been surrendered, entirely, no matter how the dice fall; the aleatory venture may be indeterminate in terms of God’s intention, but the wager is itself an irrevocable intentional decision, wherein every possible cost has already been accepted.” That principle, applied to the question of creation’s modality rather than to the question of hell, is the engine of this essay.

^[10]“The Devil’s March,” 317: “eternity reduces the price of evil to absolutely nothing. For if anything were to be eternally lost — the least little thing — then the goodness of creation could never be more in the end than a purely conditional goodness, a mere relative evaluation, rather than an essential truth. And then neither could God be the Good as such.” I accept this entirely; my argument is that it secures the finish and leaves the origin unaddressed.

^[11]“The Devil’s March,” 300. The sentence gives with one hand and takes with the other, and I do not think Hart is being careless. I think the pattern is structural, and that a theology which locates evil’s contingency in the fall rather than in the world’s existence will find itself needing the occasions.

^[12]Doors of the Sea, ch. 2: “the union of free spiritual creatures with the God of love is a thing so wonderful that the power of creation to enslave itself to death must be permitted by God.” The structure is precise: privation denies evil substance; permission-for-the-sake-of grants it purpose.

^[13]Jordan Daniel Wood, The Whole Mystery of Christ: Creation as Incarnation in Maximus Confessor (University of Notre Dame Press, 2022), and his subsequent “Reply to My Critics,” where he is emphatic that the identity of Word and world is hypostatic rather than natural, rejects the assumption “that creation is not really God unless it is naturally God,” and disclaims any reading on which union means “the annihilation of either.” Critics who charge him with a natural collapse are attacking a position he has explicitly disowned, and I do not make that charge here.

^[14]Wood, The Whole Mystery of Christ, ch. 4, “The Whole: Creation as Christ.” The two formulations are that “it is our personal freedom that renders the created world a fallen world,” and that particular persons “freely prefer fantasy to the Word.” Evil, for Wood, originates in creaturely freedom. He does not necessitate the fall, and it would be a caricature to say he does.

^[15]Wood, The Whole Mystery of Christ, ch. 4, drawing on Gregory of Nyssa: the final restoration “will make the memory of evil things utterly vanish,” the false world will “dissipate into the nothing it is,” and evil is a “positive nothingness” not reckoned to exist. This strand is very nearly the position I defend, and my quarrel is not with it but with what accompanies it.

^[16]Wood, The Whole Mystery of Christ, ch. 4: Christ “endowed the very false ‘principles’ our sin falsely incarnate, namely the ‘law of death,’ with the deeper principle of providence, the complete deification of even this universe and of the ‘me’ I make in vain.” The phrases are his and I have not softened them, because everything in my objection turns on the word endowed.

^[17]This rejoinder is not one Wood makes in so many words, but it is the strongest reply available to him and I have tried to state it at full strength rather than knock down a weaker one. A related fairness point: at The Whole Mystery of Christ, ch. 4, Wood explicitly entertains the very distinction I am pressing — that “there are but two ways to exist: as God or as not-God” — and deliberately moves past it, holding that Christo-logic transcends the essence-logic of God and not-God. He has not overlooked the wall; he has decided to go through it. My argument is a reason for stopping where he goes on, and it rests on the judgment that Chalcedon’s hypostatic union presupposes the distinction of natures rather than transcending it: without two genuinely distinct natures there is nothing for one person to unite, and Chalcedon’s four negations exist to keep the natures from meeting as natures.

^[18]The distinction between formative suffering and privative evil tracks the classical privatio boni — evil as the absence of a due good, not a positive substance. Aquinas distinguishes within this: some pains belong to the natural order of a creature coming to be, others are pure privation, the corruption of what should be. The scraped knee is not a privation; the cancer is. I am using the distinction in that classical sense, though a fuller treatment would need to address cases where the line is harder to draw — whether, for instance, the suffering that produces a certain kind of courage is formative or privative depends on whether the suffering was ever the creature’s due. What I am confident of is that the cases Ivan presses (the child in the outhouse, the serf boy hunted by hounds) are unambiguously privative: nothing about them is any creature’s formation, and no good the child could be said to gain flows from them. It is of those cases, and the class they represent, that the essay speaks.

^[19]The background here is Thomist (Summa Theologiae I.19.3, that God wills things other than himself but not necessarily) though I am putting it to a use Aquinas does not, and departing from the standard account at one point. The usual formulation locates creation’s contingency in the divine will’s relation to its object. I am locating it in the object’s own mode of being, because I think the will-relation formulation invites exactly the deliberative picture that has to be refused.

^[20]Robert Sokolowski, The God of Faith and Reason: Foundations of Christian Theology (University of Notre Dame Press, 1982), esp. ch. 3. The formulation — that God “could be all that he is” even if the world had never been, so that the world exists not out of any necessity but as sheer gift — is what Sokolowski calls “the Christian distinction” between God and the world. I take it to be dogmatic rather than speculative: it articulates what creatio ex nihilo already confesses, and the metaphysical argument of the preceding section is best read as an unfolding of it rather than an independent proof. It is worth adding that this places the might-not-have-been on the same footing as the homoousion and the universal restoration on which the rest of the argument depends — each is a datum of the Christian confession that theology labors to understand, not a conclusion reasoned to from neutral ground.

^[21]The burden, it seems to me, lies the other way: it is the objector who must explain how the reliability of a giving converts a gift into an entitlement, and I do not see how the conversion could be made. A gift whose giving is eternal and unfailing is not thereby a payment owed.

^[22]A necessary clarification, since I hold a claim that sounds like the one I am rejecting. With Hart I affirm that “God could not create a rational will not oriented toward deifying union with himself, any more than he could create a square circle” (You Are Gods [University of Notre Dame Press, 2022], 12; the same argument and image recur at Light of Tabor, 41 and 48). This is the heartbeat of universalist metaphysics and I would not surrender it. But its necessity is conditional: given that God creates rational spirits, they are necessarily deiform, and therefore necessarily arrive. It says nothing about whether the antecedent holds — whether God creates at all. The claim governs the arrow, not whether the arrow was fired. Everything in this essay depends on keeping those two apart: the end of creation is necessary, the existence of creation is gift.

^[23]It is worth recording that Hart reaches the same conclusion by a different road, and reaches it first. His argument is that if anything at all were eternally lost, creation’s goodness could be only conditional and God could not be the Good as such (”The Devil’s March,” 317). Mine is that an eternally sustained perdition would mean evil eternally upheld rather than merely remembered. The arguments are complementary, and on this question there is no daylight between us.