Chapter 13

Does the Argument from Evil Prove Too Much?

McClymond’s Sharpest Parry, and Where It Misses

A. The Argument on the Table

Thomas Talbott says a perfectly loving God could not permit any creature to be lost forever. McClymond answers: by that logic, a perfectly loving God could not permit the Holocaust, either. And yet here we are.¹

It is a sharp parry. For a moment it stops you. If the argument is simply “a loving God would not allow dreadful things,” then the world we actually live in refutes it before we ever reach the question of hell. Children die in earthquakes. Whole peoples are swallowed by genocide. Cancer hollows a mother of three while her family watches. A loving God, on any honest account, has already permitted horrors so appalling they break the mind. So how can the universalist turn around and say, “But He would never allow eternal loss”?

McClymond presses exactly this point. He lays out Talbott’s famous trilemma, then revises it to show where the logic leads.

I want to grant the seriousness of this. McClymond is not dodging the argument from divine love. He is pressing it harder than Talbott himself, pushing the reasoning past hell and into the present world. If love cannot tolerate eternal loss, he asks, how does it tolerate the appalling losses of this life?³

That is a real question, not a trick. Any honest universalist has to answer it. I believe it has a clear answer. But the answer depends on a distinction McClymond’s revision quietly erases.

B. Why It Has Force

The force of McClymond’s move lies in its form. Arguments of the type “a loving God would not permit X” really do carry a burden. They must explain why God permits the dreadful things He plainly does permit. If they cannot, they overreach.⁴

This is the lesson of the problem of evil in philosophy. Since at least the third century before Christ, thinkers have pressed the question: if God is both loving and all-powerful, why does evil exist at all? The ancient version, preserved in Lactantius’s De Ira Dei and attributed to Epicurus, puts the challenge bluntly: God either wants to prevent evil and cannot, or can and does not want to, or neither, or both. If both, why is there evil?⁵

Philosophers of religion have spent centuries on this. The free-will defense, developed most rigorously by Alvin Plantinga, argues that a world with genuinely free creatures is worth the risk of the evil those creatures may freely choose. Even the atheist philosopher J. L. Mackie conceded that Plantinga’s defense succeeds against the logical problem: there is no strict logical contradiction between God’s existence and the existence of evil.⁶ The debate then moved from the logical problem of evil (is God’s existence logically incompatible with evil?) to the evidential problem (does the amount and kind of evil make God’s existence unlikely?).⁷ And as Hronich shows, the problem of hell raises the stakes further: if God is omnipotent, omnibenevolent, and omniscient, why does He permit not just temporary suffering but eternal torment or permanent extinction?⁴⁴

McClymond’s charge is that Talbott’s argument against eternal hell is really the problem of evil dressed up in universalist clothes. And the charge works against a certain kind of reasoning. If someone simply argues, “A loving God would never allow anything terrible,” then yes, that argument proves too much. It rules out not only hell but also the world as we know it.

Notice how tightly the parallel is drawn. McClymond takes Talbott’s three propositions about God’s redemptive will, God’s power, and the eternal loss of sinners. Then he rewrites them as propositions about God’s will against evil, God’s power to prevent evil, and the existence of evil. The revision is elegant. It makes the universalist’s argument look identical in form to the ancient Epicurean challenge. And if the forms are identical, then anyone who presses Talbott’s argument is standing on the same ground as the atheist who says no loving God could exist at all.³⁹

That is a serious charge. The universalist does not want to be an accidental Epicurean. I take it seriously, and I think any universalist who shrugs it off has not understood what McClymond is doing.

The question is whether Talbott’s argument really takes the form McClymond’s revision gives it. I do not think it does. Watch the move that makes the difference.

C. The Response

1. The Relevant Difference Is Redemptive Horizon

Here is where McClymond’s revision goes wrong. He treats “evil God permits” as one undivided class. The Holocaust and eternal hell sit side by side as instances of the same thing: dreadful outcomes God allows. If His love tolerates the one, it might tolerate the other.⁸

But there is a difference between them, and it is not small. The Holocaust was a horror beyond words. Yet even the Holocaust had an end. Its survivors lived to see justice, or at least to see the regime that murdered their families destroyed. Their suffering, real and monstrous as it was, was not the final word about them. It could be redeemed. Not cheaply, not easily, but in principle: God could and one day will wipe away those tears, raise the dead, heal the broken, and bring even the perpetrators to repentance and judgment.⁹

Eternal, unredeemable loss is a different category entirely. If a soul is lost forever, that suffering is not terrible-but-temporary. It is not dreadful-but-redeemable. It is infinite. It has no horizon. God will never wipe those tears, because the tears never stop. The person is not healed but tormented without end, or annihilated without hope.¹⁰

That is the distinction McClymond’s revised propositions erase. He swaps Talbott’s specific claim about unredeemable, eternal loss for a general claim about any evil at all. Then he shows the general claim is too broad. Of course it is. But that was not Talbott’s claim.¹¹

Think of it this way. A good surgeon permits a painful operation because it leads to healing. The pain is terrible, but the healing redeems it. A surgeon who cuts open a patient, inflicts agony, and then walks away forever, leaving the wound open for eternity, is not practicing medicine. He is practicing cruelty. The difference is not the severity of the pain. It is whether the pain leads somewhere.¹²

Or think of a parent. A loving parent allows a child to suffer the consequences of a bad decision, because the suffering teaches. The child touches the stove, and the burn heals. The parent does not prevent the burn, because preventing it would mean the child never learns. But the same parent would never lock the child in a room with the hot stove forever. The willingness to let suffering teach is not the willingness to let suffering destroy. Any parent knows the difference. I believe God does too.

Temporary evil, however severe, can be outweighed. Paul himself makes this case: “For this light momentary affliction is preparing for us an eternal weight of glory beyond all comparison” (2 Corinthians 4:17).¹³ Paul is not minimizing the affliction. He is saying it leads somewhere. It is not the final chapter. Eternal loss, by definition, has no further chapter. It is the end of the story, and the end is permanent agony or permanent destruction.

The point is not that God enjoys suffering or designs it for fun. The point is that suffering within a story that ends in healing is a different moral category from suffering that is the ending. McClymond’s revision treats them as the same. That is where his argument breaks.

2. The Free-Will Theodicy Already Accepts Temporary Evil

I want to be clear about what the universalist does and does not deny. The universalist does not deny that God permits horrors. We live in a world full of them. The universalist affirms what the free-will defense affirms: God has made creatures with genuine freedom, and genuine freedom includes the real possibility of terrible choices. Free agents have used that freedom to commit unspeakable acts.¹⁴

What the universalist denies is that those horrors are the last word. God permits the terrible chapter. He does not permit a terrible ending. He allows His creatures to wander, to rebel, to wound one another and themselves, for a time. But He does not abandon them in that rebellion forever. As I argued in Chapter 11, the universalist’s theodicy is a free-will theodicy with a restorative ending: evil enters through freedom, and grace meets it there and does not quit until the evil is consumed and the person is healed.¹⁵

McClymond’s revision collapses this distinction. He reasons: if Talbott cannot explain temporary horrors, he cannot rule out eternal ones. But Talbott can explain temporary horrors. Temporary horrors are the price of real freedom in a world where God’s redemptive plan has not yet reached its end. They are not the destination. They are the road. And the road has a destination.

Here is another way to see it. Imagine two stories. In the first story, a father lets his son wander into a dangerous city, knowing the son may suffer but trusting that the suffering will teach him and that the father will find him and bring him home. In the second story, the father lets his son wander into a dangerous city and then locks the gate behind him forever. The first story is painful but purposeful. The second is abandonment. The universalist says God is telling the first story, not the second.

Paul says it plainly: “For I consider that the sufferings of this present time are not worth comparing with the glory that is to be revealed to us” (Romans 8:18).¹⁶ He does not say the sufferings are light. He says they are not the last measure on the scale. Glory outweighs them. That is the logic of a redeemable evil. And if it is true for the Christian who suffers now, the universalist asks why it would not be true for every person God has made.

3. It Is the Eternity and Finality, Not the Severity, That the Argument Targets

This is the point I most want to make clearly. Talbott’s argument does not say, “A loving God would not allow anything very bad.” It says, “A loving God would not allow anything endlessly, irrevocably bad for a creature He loves.”¹⁷

Notice what the argument targets. Not severity. Not duration measured in years or centuries. It targets finality. It targets the claim that God would write off a creature forever, closing every door, ending every offer, and consigning that creature to either unending torment or permanent extinction with no further word of grace.

That is why the argument does not prove too much. It proves exactly what it claims: that endless, unredeemed loss is incompatible with the love of a God who is perfect in knowledge, perfect in power, and perfect in goodness. It does not claim that every instance of severe suffering is incompatible with that love. It claims that an infinite instance with no redemptive horizon is.¹⁸

There is a simple test for whether an argument “proves too much.” An argument proves too much when its conclusion is broader than its premises can support. Does Talbott’s argument yield the conclusion that no evil could exist in a world governed by a loving God? Only if you strip the argument of its key qualification: the word “eternal.” Once you put the qualification back, the conclusion is narrower and more defensible: not “God would never permit evil” but “God would never permit evil that is endless and unredeemable.”¹⁹

4. Reframe as the Argument from Love, Not from Evil

There is another way to see why McClymond’s comparison to Epicurus does not hold. Epicurus’s ancient problem of evil is a negative argument. It starts with the badness of what exists and reasons backward: if things are this bad, no good God can be running the show.²⁰

Talbott’s argument works differently. It is grounded not mainly in the badness of an outcome but in the positive character of God. It starts with who God is: “God is love” (1 John 4:8).²¹ A God who is love in His very nature wills the good of every person He has made. And a God who is both loving and all-powerful has the resources to pursue that good until it is realized. The argument is not, “Evil is so bad that God cannot exist.” It is, “God’s love is so deep that no one He loves will be lost to it forever.”²²

That is a different logical shape. Epicurus starts with the problem and denies the God. Talbott starts with the God and trusts the resolution. The traditional theist already takes this second step when she says, “I do not know why God permits this suffering, but I trust that He has a good purpose for it.” The universalist takes the same step one degree further: “I trust that His purpose will not stop short of the last lost soul.”²³

McClymond’s reply assumes the two arguments share a structure. They share a surface resemblance, but their foundations face in opposite directions. One reasons from evil to atheism. The other reasons from divine love to universal hope. Calling the second one “the problem of evil in disguise” misses the way it works.²⁴

Consider the traditional believer who stands at the graveside of a child and says, “I do not understand why God allowed this. But I trust His love, and I believe He will make it right.” That person is not reasoning like Epicurus. She is not concluding that God does not exist because the world is cruel. She is starting from trust in God’s character and reasoning toward a resolution she cannot yet see. The universalist does the same thing, only further: “I trust that the resolution God brings will leave no one behind.” Both positions rest on the same foundation. Neither one is the Epicurean move turned sideways.

5. Concede the Limit

I want to be honest about what the argument from love does not do. It does not prove, in the strict logical sense, that every creature will certainly be saved. I have said this before in this book, and I will say it again: human freedom is real, and a genuine “yes” requires a possible “no.” The bare logical possibility that some creature might refuse God forever cannot be ruled out by force.²⁵

What the argument does show is that eternal loss is deeply improbable given the character of God. A God of perfect love, perfect knowledge, and infinite patience has resources that no earthly lover or counselor possesses. He meets every soul in His own unveiled presence and knows exactly what each person would freely choose in any situation. Kirk MacGregor, in his analysis of Molinism and eventual universalism, reaches a conclusion that strengthens this point: given middle knowledge, neither eternal loss nor universal salvation can be ruled out on philosophical grounds alone. But if God knows what every creature would freely choose in any possible set of circumstances, He possesses exactly the resources needed to bring about a world where all are freely saved.⁴³ The analogy between human love and divine love breaks before the reality. We cannot persuade a stubborn friend to change. But we are not God. We do not know the one word, the one moment, the one unveiling of glory that would open that friend’s heart.²⁶

God does. And He has forever. That is not coercion. It is the most patient, most personal, most perfectly fitted appeal that any being could receive. Is it possible that someone resists even that? In the abstract, yes. Is it probable? I do not think so. Is it certain? I would not dare to say. I hold the hope with conviction, not with certainty. And that modesty, I think, is enough to answer the charge that the argument overreaches.²⁷

D. Engagement with Specific Scholars

Let me bring the voices together.

McClymond is at his sharpest here. His revision of Talbott’s propositions is clever and well-constructed. He takes Talbott’s narrowly focused trilemma and broadens it into the classic Epicurean problem. That is a real philosophical move, not a rhetorical trick, and it deserves a careful answer rather than a quick dismissal.²⁸

Talbott himself is aware of the relationship between his argument and the problem of evil. His case in The Inescapable Love of God does not rest on a naive “love would never allow pain” premise. He argues specifically about the eternal loss of a creature loved by God. His three propositions are focused on the irreconcilability of some sinners, not on the existence of suffering as such. McClymond acknowledges that Talbott frames the paradox around “the specific evil of sinners who remain forever unreconciled to God,” but his revision strips away that specificity to show the broader implication.²⁹ The question is whether the revision is fair or whether it changes the argument into something Talbott did not claim.

I believe it changes the argument. And I am not alone. Andrew Hronich, in his careful case for universal restoration in Once Loved, Always Loved, makes a similar point: the distinction between temporary and eternal evil is not a convenient escape hatch but a genuine category difference. Hronich observes that Jesus himself in Luke 12:47–48 “affirms the notion of differentiation in punishment, but not of an eternal punishment. Some receive more blows than others, but none receive an endless beating.”³⁰ A God who permits suffering on the way to healing is doing something categorically different from a God who consigns a creature to suffering with no healing, ever. The first is a doctor. The second is, at best, a warden, and at worst something no Christian should want to say about God.

The tradition itself has recognized something like this distinction. Irenaeus of Lyon, one of the earliest and most orthodox of the church fathers, taught that God permits suffering as part of the growth of the soul toward maturity. His “soul-making” theodicy is built on the assumption that the suffering serves a purpose and leads somewhere.³¹ John Hick, the most famous modern defender of this approach, was a universalist. That was no accident. Hick argued that the soul-making theodicy rises and falls on universalism. As Hronich records Hick’s reasoning: “The sufferings of the damned in hell, since they are interminable, can never lead to any constructive end beyond themselves and are thus the very type of ultimately wasted and pointless anguish.”⁴⁵ Robin Parry, in the third edition of The Evangelical Universalist, presses the same point. Parry tells the story of a mother whose thirteen-year-old son dies in a car accident. The mother’s suffering leads her to Christ. But what about the boy? If his death served her salvation but not his own, then God used the child as a mere instrument, not as a person. Universalism removes the problem: God can use the boy’s death for his own good too, beyond the grave.⁴⁶

Parry’s broader point against McClymond is devastating. In his extended reply to The Devil’s Redemption, Parry argues that McClymond’s claim that traditionalism handles evil better than universalism gets things exactly backward. Eternal hell eternalizes evil. It means that countless creatures fall forever short of their created purpose, that sin is never fully destroyed, and that God is never truly “all in all” but only “all in some.”⁴⁷ That is not a solution to the problem of evil. It is the problem of evil made permanent.

R. Zachary Manis, in Sinners in the Presence of a Loving God, traces how Talbott himself views divine judgment. For Talbott, God’s retribution is not the opposite of His redemptive work but one of its instruments. God does not punish to “settle the score.” He punishes because He will not shield the unrepentant from the natural results of their sin, “because doing so would be at cross purposes to His goal of bringing about their restoration.”⁴⁸ The punishment serves the restoration. The fire burns to heal, not to destroy. That picture of judgment fits the category distinction this chapter draws. A God who punishes to restore is telling a different story from a God who punishes forever with no restoration in view.

Marilyn McCord Adams, a philosopher who spent decades on the problem of “horrendous evils,” argued that certain evils are so severe they can only be redeemed by a divine good that defeats them from within.⁴⁰ Adams was not a universalist in the strict sense. But her logic points the same direction: if God is the kind of being who can defeat horrendous evil, then the question becomes whether He would stop short of defeating it in every person.

David Bentley Hart presses this point with a force that deserves hearing. In That All Shall Be Saved, Hart argues that if “justice” means anything at all, it cannot mean the eternal torment of finite creatures. If “love” means anything at all, it cannot mean that God abandons creatures made in His image to endless suffering. If “goodness” means anything at all, it cannot mean that the story ends with a God who is “all in all” while some of His children are forever outside His reach.⁴² Hart’s argument here is not sentimental. He is pressing the meanings of the very words we use when we speak of God. The “proves too much” charge says Talbott’s argument proves that no loving God exists. Hart’s reply is that a God who permits unredeemable loss is already a God who fails the definition of love.

On the other side, I want to be fair to McClymond’s own reasoning. He is not merely scoring a debating point. His concern is that Talbott’s philosophical method smuggles in an assumption about God’s nature that does more work than Talbott acknowledges. If you start with a certain picture of divine love, McClymond says, you can derive almost anything. The universalist derives universal salvation. But the same method, pressed further, threatens to dissolve the reality of evil entirely.⁴¹ That is a genuine worry, and it is one reason the boundary matters so much (see Chapter 12). The biblical universalist does not dissolve the reality of evil. He insists on it. Evil is so real it takes the cross and the purifying fire to destroy it. What the universalist refuses is the claim that evil wins in even one soul.

E. Counter-Objections

“Redeemable Is Doing All the Work and Just Assumes Universalism”

This is a fair pushback. Am I not arguing in a circle? I say eternal loss is different from temporary suffering because temporary suffering can be redeemed. But “redeemed” means healed and restored, which is what universalism claims. So my defense of the argument is the argument itself.

Not quite. The claim that temporary suffering can be redeemed does not assume universalism. It assumes only what the free-will defense already assumes: that God has a good purpose for the evils He permits, and that His purpose leads to a resolution. Every orthodox Christian, whether traditionalist, conditionalist, or universalist, believes that the suffering of the redeemed is temporary and redeemable. The only question is the scope. The universalist says the scope of redemption matches the scope of the atonement: it reaches everyone. That is a further claim, but the category distinction between redeemable and unredeemable evil does not depend on it. It depends on the widely shared conviction that God does not permit suffering without purpose.³²

“Some Temporary Evils Are So Severe They Are Effectively Unredeemable”

Can any amount of future glory truly “outweigh” the Holocaust? Can a child’s tortured death really be redeemed by anything that comes after? I feel the force of this objection. It is the hardest corner of the problem of evil, and I will not pretend to dissolve it.³³

Three things need saying. First, the Christian faith has always believed that God can redeem even the worst. Paul, who knew Roman torture, insisted that the coming glory outweighs present suffering (Romans 8:18). John envisions a day when God “will wipe away every tear from their eyes, and death shall be no more” (Revelation 21:4).³⁴ The hope is not that suffering is trivial but that God is greater than it.

Second, notice that this objection cuts against every theodicy, not just the universalist one. If some evils are effectively unredeemable, then the traditionalist faces the same problem the universalist does, and worse: the traditionalist must explain why God permitted suffering that even He will never heal, not only in this life but forever. The universalist at least says: God will heal it. The healing may take eternity, but it comes.

Third, I think the objection mistakes the nature of redemption. To redeem suffering is not to undo it, as if it never happened. It is to bring the sufferer through it into a wholeness that absorbs and defeats the wound. A mother who lost a child and was later reunited does not forget the grief. The grief becomes part of a story that ends in joy, and the joy is deeper for having passed through the grief. That is not a tidy resolution. It is the best hope we have in a world where evil is real.

The answer is the same line I have drawn throughout: the difference between terrible-but-temporary and terrible-and-final. Love can permit pain on the way to healing. Love cannot abandon a person to pain that never ends and never heals, because “love never ends” (1 Corinthians 13:8).³⁵ A love that quits is not the love Scripture describes. And a love that says, “I have done all I can; the rest is up to you, forever,” is not the love of a God who has infinite patience, infinite knowledge, and infinite time.

F. Conclusion

McClymond’s “proves too much” rejoinder is the sharpest parry in his engagement with Talbott. It deserves the respect of a careful answer, and I hope I have given it one.

The argument from divine love does not prove too much. It proves exactly what it claims. It says that a God of perfect love can permit terrible things on the way to healing, but cannot abandon any creature to suffering that is endless and unredeemable. It says the difference between redeemable horror and unredeemable loss is not a matter of degree but a matter of kind. And it says that the love of God, the power of God, and the patience of God together make universal restoration not a sentimental wish but a reasonable, biblical hope.³⁶

McClymond wants to push us all the way to Epicurus. But the road forks. Epicurus starts with the badness of the world and concludes that no loving God exists. The universalist starts with the love of the God who does exist and concludes that His love will not rest until every creature is home. One is a problem. The other is a hope. And the hope does not collapse into the problem, because it is grounded in a Person, not in an argument about evil.³⁷

I think McClymond is right about one thing: any universalism that rests on sentimentalism, on the vague feeling that a nice God would never let bad things happen, really does prove too much. That version of universalism cannot explain this world. It cannot account for a single cancer ward or a single war. And if it cannot explain this world, it cannot claim to know how the next one ends. That is not the universalism I hold, and it is not the one this book defends.

The universalism this book holds is built on something harder. It is built on the conviction that God’s love is not a sentiment but a fire, that His judgment is not cruelty but surgery, and that the cross is the proof that He will go to any length to bring His children home. The difference between redeemable suffering and unredeemable loss is the difference between a God who finishes what He starts and a God who gives up. Scripture paints a God who finishes.

Paul put it best. I will give him the last word. “The last enemy to be destroyed is death” (1 Corinthians 15:26). Not managed. Not quarantined. Destroyed. And when death is gone, what remains? “God may be all in all” (1 Corinthians 15:28). That is not sentimentalism. That is the scriptural promise. And a promise from that God is worth more than all our arguments put together.³⁸

We will meet the no-difference worry next (Chapter 14), and the question of whether freedom can resist even optimal grace in Chapter 26. For now, the argument from love stands. It does not prove too much. It proves enough.