Chapter 2
What Kind of Case This Book Makes, and Why the Gap Matters
Here are two sentences. Both are true. Only one is a proof.
"I am certain that two plus two is four."
"I am certain my closest friend will not betray me."
The first is a matter of logic. No new information can overturn it. It holds by the bare structure of numbers and addition. The second is a matter of trust, built on years of evidence: a character observed, a pattern tested, a love proved again and again. It is not a theorem. It could, in some bare logical sense, turn out wrong. But it would shock me to my bones if it did, because everything I know about that person points the same way. Both are real certainties. But they are different kinds of certainty.1
This chapter asks a simple question: which kind of certainty does the universalist hope belong to? Is the claim that God will save everyone a logical proof, as sure as arithmetic? Or is it a deeply grounded conviction, built on everything we know about God, love, justice, freedom, and grace?
The answer matters more than it looks. It shapes how the whole book works. It decides what I can honestly promise and what I cannot. And it draws a line between this book and the strongest universalist argument in print: the argument that salvation is not just likely but necessary, that a rational mind cannot finally refuse God. That argument belongs to David Bentley Hart, and I admire it. But I will not follow him there. This chapter explains why, and what I hold instead.2
A book that tries to prove too much loses the reader's trust. If I claimed that universal salvation is as certain as a math problem, you would rightly ask: then why does anyone disagree? What about human freedom? What about the people who seem to want nothing to do with God? A proof that waves away those questions is not honest. It has traded intellectual humility for a clean ending.
But a book that claims too little loses something else. If I said the hope of universal salvation is just a pleasant guess, a nice thought with no real weight behind it, then why write thirty chapters? Why spend years building a case? A bare wish is not worth a book.
I want to land in the right place between those two. The case I will build across this book is not a deductive proof. It is not a single argument that locks the answer into place the way a syllogism does. It is a cumulative case: many independent lines of reasoning, each one pointing in the same direction, each one adding weight, until the combined force makes the universalist conclusion far more reasonable than any rival. Think of it like a rope. No single strand carries the full load. But twenty strands, braided together, hold.3
This is not a lesser kind of argument. It is how most of our deepest convictions actually work. You believe your friend is loyal not because of one single proof, but because of a thousand small pieces of evidence. A jury reaches a verdict not from one witness, but from many, taken together. A scientist trusts a theory not because one experiment settles it, but because experiment after experiment, in different labs, points the same way. The cumulative case is how we reason about the things that matter most.4
Plain Words
Three terms will come up often in this chapter. A deductive proof is an argument where the conclusion follows with iron certainty from the premises, like a math proof. A cumulative case is an argument where many separate reasons, taken together, make the conclusion far more reasonable than the alternatives. Metaphysical necessity means something could not possibly be otherwise; it is logically impossible for it to fail. This book argues for conviction, not necessity.
And here is why it matters for this book in particular. The critics have a favorite charge. Michael McClymond argues that universalism, across its history, tends to eclipse the very grace it claims to magnify. What seems to be "all grace" turns out, on inspection, to be "no grace." The soul saves itself by nature. Or God saves by necessity. Or the sinner expiates his own way through purgatorial suffering. McClymond presses this in the introduction to The Devil's Redemption and builds the cumulative case in his final chapter. James De Young sharpens the point from a different angle. He charges that universalism is "more deterministic and coercive than Calvinism," because it insists that everyone must repent, before death or after it, leaving no room for a genuine refusal.5 The best way to answer that charge is to show, from the start, that this book makes no such claim. The case is strong. The conviction is real. But it is held with both hands on human freedom, and that means it is held as conviction, not as proof. I have not proved that God will save everyone the way you prove a theorem. I have shown that it is, by far, the most reasonable thing to expect from the God revealed in Christ.
Start with a distinction. There are three grades of confidence a universalist could hold, and they are very different from one another.
The first is bare hope. On this view, we may hope that God will save everyone, but we have no strong reason to think He will. It is a wish, not a settled position. The great Catholic theologian Hans Urs von Balthasar is sometimes read this way: we may "dare to hope" that hell is empty, but we cannot go further than that.6 I respect this posture. But I think it underestimates the evidence. The character of God, the structure of His love, the logic of His justice, and the resources of His grace all point in the same direction. A bare hope does not do justice to how strong that convergence is.
The second is overwhelming conviction. On this view, the evidence for universal salvation is so strong, so many-sided, and so deeply rooted in the character of God that the conclusion is by far the most reasonable one available. It is not a proof. It is the kind of certainty you have about a trusted friend: tested, grounded, supported from every angle. But it keeps a small, honest gap open. Because human freedom is real, the bare logical possibility of a final refusal cannot be sealed shut by argument alone. Robin Parry, one of the most careful evangelical universalists, puts himself in this camp. He calls himself a "confident universalist": not merely hoping that all will be saved, but affirming it with real confidence, while stopping short of logical necessity.36 This is where I stand.7
The third is metaphysical necessity. On this view, universal salvation is not just very likely. It is logically unavoidable. A rational will, fully seeing God, cannot refuse Him. There is no possible world in which any creature is lost forever. This is the position of David Bentley Hart, and it is the strongest philosophical argument for universalism in print.8
These three are not points on a spectrum. They are different kinds of claim. The move from conviction to necessity is not just a step up in confidence. It is a change in the kind of argument you are making, and it comes with a price I am not willing to pay. The next few sections explain what that price is.
Some readers will hear "conviction, not proof" and think I am hedging. Let me push back on that.
A cumulative case is not a weak case. It is a different kind of case. Think of it this way. If I want to prove that a triangle's angles sum to 180 degrees, I need one argument, and that argument is decisive. Done. But if I want to know whether a defendant is guilty, no single piece of evidence settles the question. I weigh the testimony, the forensic evidence, the motive, the alibi, the character witnesses, the timeline. Each one adds weight. Together, they can make the case overwhelming, even beyond reasonable doubt, without any one of them being a proof on its own.9
The universalist case works the same way. Here are the strands this book will braid together. God's character is perfect love (Chapter 3). That love wills every person's good (Chapter 4). The logic of divine love presses toward saving all (Chapters 5 and 7). Justice rightly understood is restorative, not retributive in the endless sense (Chapters 8 and 9). The philosophical problem of hell is severe (Chapter 10). The free-will defense of hell, examined closely, does not hold (Chapter 11). God's knowledge of what every person would freely do, in any situation, gives Him the tools to save without forcing (Chapter 14 and Chapter 15). Death does not lock the door on grace (Chapter 19). Eternity gives the offer of love all the time it needs (Chapter 21). No single strand is a proof. Together, they form a rope that holds.
Eric Reitan and John Kronen, in God's Final Victory, call this the comparative method. The idea is simple: set two whole packages of belief side by side and ask which one a perfectly good and wise God would choose. On one side, place the package that includes endless conscious torment. On the other, place the package that includes the final salvation of all. Then ask: which package fits better with what we know about God's love, power, and wisdom? Reitan and Kronen argue that the comparison, taken as a whole, favors the universalist package.10
The Shape of This Book's Case
This book does not rest on a single knockout argument. It builds a cumulative case, where many independent lines of reasoning all point the same way. Each chapter adds a strand to the rope. Together, they make the universalist conclusion far more reasonable than the alternatives, without claiming it as a logical proof. The full weighing comes in Chapter 29.
Reitan's later work, Troubled Paradise, sharpens the same idea. Every theological position, Reitan argues, carries costs: modifications to our broader network of reasonable beliefs that a given position forces us to make. Calvinism costs you God's universal love. Arminianism costs you God's ability to finish what He starts. Eternal torment costs you proportionality, the goodness of God's purposes, and the final peace of heaven. The reasonable choice is the position with the fewest and least severe costs, even when no position is cost-free. He reaches universalism not as a necessity, but as the "least costly" escape from the problem of heavenly grief. On the question of freedom, Reitan goes further still. Even on a radical libertarian view, God can ensure the salvation of all by keeping creatures indefinitely free to repent. Across infinite opportunity, the probability of eternal refusal approaches zero.11 That is very close to the method of this book.
Now I need to do something that may seem strange. I need to state, at full strength, the best universalist argument in print, and then explain why I will not follow it.
David Bentley Hart argues that universal salvation is not merely likely. It is necessary. His reasoning runs roughly like this. The will is always a movement toward the good. We desire things because they look good to us, and we turn away from things because they look bad. If a person fully and clearly saw God as He really is, with no ignorance, no deception, and no corruption of desire, that person would have no reason left to refuse Him. Every motive for refusal would disappear. And if there is no reason to refuse, then refusal is not a real possibility. It is not that God forces anyone. It is that a truly free will, seeing the good clearly, cannot choose against it. Therefore, in the end, every rational creature will freely choose God. That is what freedom means when it is not distorted by darkness.12
Hart on Freedom and the Good
"Given the dynamism of human nature, given its primordial longing for the Good, given the inherent emptiness of evil, given the finitude of evil's satisfactions and configurations and resources, no rational nature could freely persist forever in its apostasy from the Good. There is no power in that nature or in evil equal to such an act."
David Bentley Hart, That All Shall Be Saved, Meditation FourHart presses the point further. He argues that Christ did not sin, not because He lacked freedom, but because He was free of all the things that make us sin: ignorance, delusion, and corruption of the will. In Hart's words, Christ alone was "fully free," precisely because He could not deviate from the good.13 If that is what true freedom looks like, then the freedom to sin is not a feature but a defect. It is something we should be healed of, not something we should cling to. And the healing is not coercion. It is the removal of a disease.
I need to be plain about what I think of this argument. It is brilliant. It is deeply serious. It draws on a philosophical tradition that runs from Plato through Gregory of Nyssa to Thomas Aquinas. Hart himself is fully orthodox: Trinitarian, creedal, and committed to the Incarnation, the atonement, and the resurrection. He is not a counterfeit universalist by any measure. And his argument has real force. I feel its pull every time I read it.14
But I cannot follow him. Here is why.
If the will, when it sees the good clearly, cannot refuse it, then the "yes" to God is not a free choice in the sense that it could have gone the other way. It is a necessary response to what the will sees. And if the final "yes" is necessary, then two things follow that trouble me deeply.
First, if a perfectly clear-sighted will cannot refuse God, then the fall should not have happened. Adam and Eve saw God more clearly than we do. If clear sight removes the power to refuse, how did they refuse?15 Hart has an answer (the fall reflects a will that was not yet fully clear-sighted), but it pushes the problem back without solving it. At some point, a creature saw enough to know better and chose anyway. That is what libertarian freedom means.
Second, and more deeply: if the saved cannot do otherwise, then the love they offer God is not the kind of love that could have been withheld. And love that cannot be withheld is not love in the fullest sense. It is a beautiful response, yes. But the beauty of love, the thing that makes it love and not just a pleasant reaction, is that it is given when it could have been kept. A "yes" that was never in danger of being a "no" is a different thing from a "yes" that was freely chosen with the "no" still on the table.16
A Parting Between Allies
Hart and I want the same ending. We both believe God will save everyone. But we get there by different roads. Hart says the saved cannot do otherwise; their "yes" is necessary. I say the saved can do otherwise; their "yes" is free. The difference matters. His road is shorter and cleaner. Mine is longer and harder. But mine keeps something his does not: the real, two-way power of the creature to say no. And keeping that power is what makes the hope honest.
I will develop this argument more fully in Chapter 13, where the philosophy of freedom gets its full treatment. For now, the point is this: the price of Hart's necessity is libertarian freedom, and I will not pay it. That means my conclusion is conviction, not proof. And that is not a weakness. It is the cost of keeping freedom real.
I want to be plain about the place where my case is not airtight. I owe you that honesty, and I think the argument is stronger, not weaker, for admitting it.
Here is the open question. If human freedom is real, if a "yes" to God must be a genuine choice and not a determined response, then the bare logical possibility of a final "no" cannot be sealed shut by argument alone. A genuine "yes" implies a possible "no." That is what makes it genuine. And if a "no" is possible, then I cannot prove that every creature will say "yes." Not in the strict sense.17
This is the one genuinely open point in the book. I face it head-on in Chapter 17, where I ask: could anyone refuse God forever? And I weigh it with everything else in Chapter 29. But let me preview my answer here.
A possible "no" is not a probable "no." The fact that something could happen does not mean it is likely to happen. I could, in some bare logical sense, decide tomorrow that I hate my own children. Nothing in the laws of logic prevents it. But everything I know about love, about who I am, and about who they are makes it so unlikely as to be unthinkable. It is not impossible. It is just overwhelmingly unlikely.18
The universalist case is like that. A creature could, in some bare abstract sense, refuse God forever. But consider what that would require. It would mean a person, fully healed of every wound, stripped of every deception, standing in the unveiled presence of infinite love, with all the time eternity allows, choosing misery over joy. Not once, but endlessly. Not from ignorance, but in full sight. Not from habit, but against every reason the soul can recognize. The rest of this book argues that such a refusal, while not strictly impossible, is so unlikely that we are right to hold the opposite as a firm conviction. Everything God is and everything God can do makes a final "no" vanishingly improbable.19
As Thomas Talbott has argued, the things that drive us to refuse God are ignorance, self-deception, and bondage to disordered desire. Remove those, and the motive for refusal disappears. Not because freedom disappears, but because the darkness that misdirects freedom is gone.20 God does not override the will. He clears the eyes. And a will that finally sees, this book contends, will freely choose the good.
Scripture hints at the same thing from a different angle. James tells us that "even the demons believe, and shudder" (James 2:19, ESV). Knowledge alone is not enough. The demons know who God is, and they do not love Him. Seeing God does not force love. But it removes the ignorance that makes refusal seem reasonable. What remains, once the eyes are clear, is a choice. And this book argues that a healed, clear-sighted soul will make it freely.21
Because I keep freedom real, I must also keep a fallback real. If the one open question ever cut the other way, if some creature did hold out against God forever, in defiance of every appeal, what then?
My fallback is conditional immortality (sometimes called annihilationism). On this view, the finally unrepentant are not tormented forever. They simply cease to exist. God does not sustain a soul in endless agony. He lets it go.22
I do not expect this fallback to be needed. Everything I know about God, love, freedom, and the resources of grace tells me He will bring everyone home. But if I am wrong, the floor is conditional immortality, not eternal conscious torment. I hold this firmly: nothing in the character of God could support the endless, purposeless suffering of any creature. If universal salvation is the ceiling, conditional immortality is the floor. Endless torment is not an option I will ever land on.23
Reitan and Kronen make a similar move in God's Final Victory. After building their case, they consider annihilationism and what they call "soteriological agnosticism," the view that we cannot know who, if anyone, will be finally lost. On annihilationism, they press an elegant point. If their earlier arguments show that God can save without violating freedom, then He need not annihilate the unrepentant to spare them suffering. Why not? Because "finite suffering followed by infinite bliss has infinite positive value." Depriving a creature of eternal joy with God, merely to spare it temporary pain, is no act of mercy. Even so, they hold that a reader who is not fully persuaded by the universalist case should at least reject eternal torment. Conditional immortality or soteriological agnosticism is a defensible floor, even if the universalist ceiling does not hold.24
What Is Conditional Immortality?
Conditional immortality (also called annihilationism) is the view that the finally unrepentant do not live forever in torment. Instead, they are allowed to pass out of existence. They are not sustained in suffering. Immortality, on this view, is a gift given to the redeemed, not a property forced on everyone. This book argues for universal salvation. But if that hope were ever to fail, the fallback is conditional immortality, never eternal conscious torment.
Why conditional immortality and not eternal torment? Because the moral argument against endless punishment is independent of the universalist case. Even if I could not show that God will save everyone, I could still show that a God of perfect love and perfect justice would not sustain a creature in pointless agony forever. The proportionality problem (Chapters 8 and 9) does its work whether or not the universalist case succeeds. Endless torment fails on its own terms.25
Let me place this book among the voices it draws on, so you can see where the lines are.
David Bentley Hart, as I have said, holds that universal salvation is necessary. His argument runs through a particular view of freedom: the will is an appetite for the good, and a will that fully sees the good cannot turn away from it. Hart draws on Gregory of Nyssa, who argued that evil is finite and privative, that human nature moves irresistibly toward God, and that no rational creature could finally persist in apostasy from the Good.26 I take Gregory's insight seriously. But I read it differently. I think it tells us what is overwhelmingly likely, not what is metaphysically certain. The dynamism of human nature does press toward God. Evil does run out of fuel. But the press is not a compulsion, and the running-out is not a proof. It is a reason for profound confidence. That is enough.
Thomas Talbott, in The Inescapable Love of God, builds the most important philosophical structure this book uses: the inconsistent triad (developed in Chapter 5). Talbott himself leans toward a necessary universalism, holding that once all ignorance and deception are removed, refusal loses its footing entirely. I borrow his structure and most of his reasoning, but I stop one step short. I keep the gap that freedom requires.27
Jerry Walls, writing as a non-universalist who takes the hope seriously, presses the hardest objection to Talbott's position. If God brings ever-increasing misery on those who rebel until they repent, Walls argues, then at some point the "pressure" breaks the creature's freedom rather than preserving it. Walls compares it to a medieval practice of pressing the accused under iron weights until they agreed to plead. "There is a limit on how much 'pressure' our freedom can bear," he writes. If Talbott is right that no illusion can endure forever under God's loving pressure, then God effectively forces the knowledge that makes refusal impossible. I think Walls overstates the case. God's persuasion need not be coercive. But the worry is real, and it is one more reason this book keeps freedom at the center. The answer is not to weaken the case for universal salvation, but to show that God can win every heart without breaking any will. That argument comes in Chapter 15.37
Eric Reitan and John Kronen, in God's Final Victory, model the method I follow most closely. They build what they call a "comparative defence": set the doctrine of hell beside the doctrine of universal salvation and ask which one a perfectly good and wise God would choose, given core Christian principles. They are plain about the kind of conclusion this yields. It is not a deduction from axioms. It is a judgment about relative plausibility: which whole package of belief sits better with what we already hold about God? They hold their conclusion as reasonable and well-grounded, not as a proof. Their humility about the argument's shape matches mine.28
Andrew Hronich, in Once Loved Always Loved, sharpens the freedom question in a way that matters here. Hronich agrees with Talbott that a rejection of God cannot be both free and fully informed, but he frames this in terms of what he calls "rational freedom": true freedom requires true knowledge and true sanity of mind. Without them, one is not wholly free. The creature who thinks resistance to God is profitable is operating under delusions, and God can remove those without violating freedom. Hronich's contribution is the precision: the removal of ignorance is not coercion. It is the restoration of the conditions that make genuine choice possible.38
Timothy Stratton, in Human Freedom, Divine Knowledge, and Mere Molinism, gives the account of libertarian freedom this book relies on. On Stratton's view, a free choice is one where the agent, not some prior cause, is the source of the action. The agent can think, evaluate, and choose among alternatives without being causally forced to one outcome. God can know what a free creature would do in any situation (that is middle knowledge, treated fully in Chapter 14) without that knowledge making the choice unfree. Knowledge is not causation. Certainty is not necessity.29 Stratton is not a universalist. But the account of freedom he defends is the engine I use to argue that God can save without forcing. I borrow his machinery and turn it toward the universalist conclusion.
The critics, Michael McClymond and James De Young, would say this in-between position is unstable. McClymond argues in his concluding appraisal that a one-word explanation for universalism is not "love" but "metaphysics," because universalist theology consistently rests on assumptions about God and the world that are not drawn from Scripture. De Young presses a related point. Universalism "decrees that all must repent and come to faith, either before dying or after dying," which, he argues, amounts to God overriding the will under the guise of love.30 This chapter is part of the answer to that charge. The universalism I defend does not slide into necessity. It keeps libertarian freedom. It keeps the one open question. It keeps the honest admission that the case is conviction, not proof. If anything, it is more committed to freedom than most traditionalist positions, because it refuses to let God's purposes fail even while insisting that God never overrides the will.
Someone will say: "A merely probable hope is cold comfort. If you cannot promise me that God will save everyone, what good is it? The grieving parent, the worried spouse, the person who has lost someone to unbelief: they want a guarantee, not an odds ratio."
Common Objection: "Probability Is Not Enough"
If universal salvation is only overwhelmingly likely, not certain, doesn't that leave the grieving without real comfort? Doesn't it mean that some people might still be lost forever?
I hear this, and I take it seriously. But I think the objection misunderstands what conviction is. A well-grounded conviction is not cold. It is the warmest kind of confidence a free creature can hold. I am convinced that my friend will not betray me, and that conviction gives me peace, even though I cannot produce a mathematical proof. The conviction rests on everything I know about my friend's character. In the same way, the universalist conviction rests on everything we know about God's character: His love, His power, His wisdom, His patience, and His refusal to give up on anyone.31
The alternative is to overclaim. And overclaiming, in the end, is a lie. If I told the grieving parent, "It is logically impossible for your child to be lost," I would be saying something I cannot back up. And when the parent pressed me, and I could not sustain the claim, the comfort would collapse. A conviction built on the character of God will not collapse. It is built on a foundation that cannot shift, even if the logical form is not a proof. Paul puts it this way: "Work out your own salvation with fear and trembling, for it is God who works in you, both to will and to work for his good pleasure" (Philippians 2:12–13, ESV). Grace and freedom work together. God acts; the person responds. That is the shape of the hope.32
A sharper objection runs like this: "If you grant that a final 'no' is logically possible, haven't you already conceded the traditionalist's point? The traditionalist says some people will refuse God forever. You admit it is possible. So you agree with the traditionalist in principle and just disagree about the numbers."
No. There is a large difference between possible and probable. It is possible, in the barest logical sense, that every coin I flip for the rest of my life will land on heads. Nothing in the laws of logic forbids it. But no reasonable person expects it to happen. The question is not whether a final "no" is possible, but whether it is likely, given everything God is and everything God can do. The rest of this book argues that it is not.33
The traditionalist does not merely say a final "no" is possible. He says it is actual, and for most of the human race. That is a claim about how the world will actually end. This book will challenge it on every front: from the character of God, from the nature of love, from the logic of justice, from the resources of grace and knowledge, and from the unlimited time of eternity. Granting a bare possibility is not granting the traditionalist's conclusion. It is being honest about the shape of the argument.34
Which Universalism Is This?
This chapter draws a line against necessitarian universalism: the claim that salvation is logically unavoidable and that the rational will cannot refuse God. Hart is the most brilliant voice in that camp, and he is fully orthodox. He is a source this book uses with gratitude. But this book parts from him on freedom. It keeps libertarian free will. It holds the conclusion as conviction, not necessity. And it refuses to treat a "yes" that could never have been a "no" as genuine love. The hope defended here is freely chosen and honestly held, not forced by logic.
Let me gather the threads.
This book does not claim to prove that all will be saved. It claims something humbler and, I think, more honest: that universal salvation is by far the most reasonable thing to expect from the God revealed in Jesus Christ. The case is cumulative, not deductive. It is built from many independent lines of reasoning, each one adding weight, and the full accounting comes in Chapter 29.
I have parted company with Hart, not because his argument is weak, but because its strength comes at a cost I am not willing to pay: the reality of libertarian freedom. I keep freedom, and I accept the consequence: my conclusion is conviction, not proof. There is one genuinely open question, and I face it rather than papering over it (Chapter 17).
If that one open question ever cut the other way, my fallback is conditional immortality, never eternal conscious torment. The moral case against endless punishment stands on its own, whether or not the universalist case succeeds.
And now the argument begins. The next chapter starts where every sound argument about God must start: with who God is. If we get His character right, everything downstream follows. If we get it wrong, nothing else can save the case. So we begin with God.35
↑ 1. The distinction between logical certainty and personal certainty (or what epistemologists call “propositional knowledge” versus “confidence grounded in evidence”) is standard in epistemology. For a clear introduction, see Alvin Plantinga, Warranted Christian Belief (New York: Oxford University Press, 2000), ch. 1. The analogy of the trusted friend is mine, but the underlying point—that rational conviction need not take the form of deductive proof—is broadly shared across the analytic tradition.
↑ 2. David Bentley Hart, That All Shall Be Saved: Heaven, Hell, and Universal Salvation (New Haven: Yale University Press, 2019). Hart’s argument for the metaphysical necessity of universal salvation runs through his four meditations, but the philosophical core is in Meditation Four, where he draws on Gregory of Nyssa’s metaphysics of the will and the finitude of evil.
↑ 3. The “rope” metaphor for cumulative-case reasoning is common in philosophy of religion. See Richard Swinburne, The Existence of God, 2nd ed. (Oxford: Clarendon, 2004), ch. 1, where Swinburne defends the cumulative approach to theistic arguments. C. S. Lewis used a similar image of converging clues in Mere Christianity (London: Geoffrey Bles, 1952).
↑ 4. The jury analogy captures the legal standard of “beyond reasonable doubt,” which is cumulative rather than deductive. For the philosophical application, see Basil Mitchell, The Justification of Religious Belief (London: Macmillan, 1973), who developed the cumulative-case method for theological claims.
↑ 5. Michael J. McClymond, The Devil’s Redemption: A New History and Interpretation of Christian Universalism, 2 vols. (Grand Rapids: Baker Academic, 2018). McClymond’s charge that universalism tends toward determinism and the “eclipse of grace” is stated most directly in §0.10, “A Theological Irony: Universalism’s Eclipse of Grace” (V 1, p 24): “The effort to extend grace to all has repeatedly ended up compromising or even eliminating the notion of grace.” McClymond identifies three ways grace is eclipsed: salvation occurs “according to nature, not according to grace” (the spark-of-the-divine teaching); it becomes metaphysically necessary (as in Sophiology or Barth); or the sinner expiates his own way clean. The charge is developed further in ch. 12, “The Eclipse of Grace: An Appraisal of Christian Universalism” (V 2, pp 999ff.), where he argues that universalist theology rests “on assumptions about the nature of God, the nature of humanity, or the nature of good and evil that are neither taught in Scripture nor embodied in the larger Christian tradition” (V 2, p 1000). See Chapter 22 for the full reply. James B. De Young, Exposing Universalism: A Comprehensive Guide to the Faulty Appeals Made by Universal Salvationists (Eugene, OR: Wipf & Stock, 2018), presses a related charge. De Young argues that “by insisting that all those in hell must repent, UR is more deterministic and coercive than Calvinism” (Preface, point 17; cf. pp. 66, 217–19, 238–39). His Part 2, “The Appeal to Emotion and Reason,” includes an excursus titled “Even Reason Demands That Hell Must Be Everlasting,” in which he contends that universalism’s corrective view of hell would actually give the devil reason to hold out: “If the fires of hell are only therapeutic, and limited in their length, would not the unrighteous and the devil in particular only have to wait long enough and the fires would cease?” See Chapter 26 for the full reply.
↑ 6. Hans Urs von Balthasar, Dare We Hope “That All Men Be Saved”?, trans. David Kipp and Lothar Krauth (San Francisco: Ignatius Press, 1988). Balthasar’s position is more complex than “bare hope”; he is sometimes read as closer to conviction than he admits. But his formal posture is that we may hope, not that we may assert.
↑ 7. On the distinction between “hope” and “conviction” as theological postures, see Eric Reitan and John Kronen, God’s Final Victory: A Comparative Philosophical Case for Universalism (New York: Continuum, 2011), §1.2, pp. 4–11. Their method is cumulative and comparative, not deductive: they aim to show that “no form of DH is as plausible as some form of DU” (p. 2), which is a conclusion about relative plausibility, not about logical necessity. Their “comparative defence” (the term is theirs) parallels this book’s method of asking which whole package of belief a perfectly good and wise God would choose. See also Reitan, Troubled Paradise, ch. 1, where the cost-comparison framework is laid out in plain terms.
↑ 8. Hart, That All Shall Be Saved, Meditation Four. Hart argues that “any truly free condition could be nothing other than the love of God.” His argument draws on a classical metaphysics of the will as an appetite for the Good, inherited from Plato through Gregory of Nyssa.
↑ 9. For the distinction between deductive and cumulative reasoning in philosophy of religion, see Swinburne, The Existence of God, ch. 1; and Mitchell, The Justification of Religious Belief, chs. 1–2.
↑ 10. Reitan and Kronen, God’s Final Victory, §1.2 “Overview of Argument and Methods,” pp. 4–11. They describe their approach as a “comparative defence”: “our aim is to show that, granted certain core Christian principles, no form of DH [the doctrine of hell] is as plausible as some form of DU [the doctrine of universal salvation]” (p. 2). Their argument is “conservative in its foundations (if not in its conclusions) and dialectical in its approach” (p. 2), relying on premises that mainstream Christians already accept. The comparative method is the template this book follows: set the major eschatological positions side by side and ask which sits best with what a perfectly good God would choose.
↑ 11. Eric Reitan, Troubled Paradise: Limited Salvation and the Problem of Heavenly Grief (Eugene, OR: Cascade, 2024), ch. 1, “Solutions, Theodicies, and Defenses.” Reitan introduces a cost-comparison framework: rather than a binary distinction between “theodicy” and “defense,” he envisions “a continuum” of reconciling stories of greater or lesser plausibility, where a “cost” is “a modification to our broader network of reasonable beliefs that the solution forces us to make.” The reasonable choice is the least costly option (ch. 1). In ch. 8, Reitan applies this method to the universalist question and concludes that even on a radical libertarian view of freedom, God can “guarantee—mathematically if not metaphysically—the salvation of all” by preserving creatures indefinitely free to repent, since across infinite opportunity the probability of eternal refusal approaches zero (the infinite-opportunity argument, developed with John Kronen in God’s Final Victory, ch. 8).
↑ 12. Hart, That All Shall Be Saved, Meditation Four. Hart’s argument is rooted in the classical view, shared by Plato, Aristotle, and Aquinas, that the will is an appetite for the good. On this view, we always will what appears good to us; we err only when ignorance or corruption causes us to misidentify the good. Perfect sight of God, Hart argues, removes every such distortion.
↑ 13. Hart, That All Shall Be Saved, Meditation Four: “What distinguished Christ in this regard from the rest of humanity, if Christological orthodoxy is to be believed, is not that he lacked a kind of freedom that all others possess, but that he was not subject to the kinds of extrinsic constraints upon his freedom (ignorance, delusion, corruption of the will, and so forth) that enslave the rest of the race.”
↑ 14. It is important to emphasize that this book treats Hart as a source, not a counterfeit. He is fully orthodox: Trinitarian, creedal, committed to the deity of Christ, the atonement, and the resurrection. See Appendix B for the full boundary doctrine and the Hart distinction.
↑ 15. This objection to necessitarian universalism is a standard one. If a clear-sighted rational will cannot refuse the Good, then the fall of Adam and Eve becomes deeply puzzling, since they enjoyed an immediate relationship with God. Hart addresses this by arguing that the pre-fallen state was not yet the fullness of clear-sighted freedom, but this appears to concede that degrees of sight allow degrees of refusal, which is the libertarian’s point. See also the discussion in Chapter 13.
↑ 16. The point here is not that love requires the exercise of refusal, but that it requires the possibility of refusal. A love that is metaphysically impossible to withhold is a qualitatively different thing from a love that is freely given. This is the core of the libertarian case for the value of freedom, developed at length by Alvin Plantinga in God, Freedom, and Evil (Grand Rapids: Eerdmans, 1974).
↑ 17. This admission is developed fully in Chapter 17 and weighed against all the other evidence in Chapter 29. The point is faced rather than papered over. It is the book’s single most honest moment.
↑ 18. The distinction between logical possibility and reasonable expectation is a staple of probability theory and epistemology. A logically possible event with an extraordinarily low probability does not warrant planning for, believing in, or treating as though it will happen. See Swinburne, The Existence of God, ch. 1, on the role of probability in rational belief.
↑ 19. This is the core of the infinite-opportunity argument, developed by Reitan and Kronen in God’s Final Victory, ch. 8, “Freely Chosen Universal Salvation,” pp. 152–77, especially §8.3 “Libertarian Freedom and the Infinite Opportunity Argument” (pp. 160–66). See also Reitan, Troubled Paradise, ch. 8, where the argument is restated accessibly: “so long as God ensures that the unregenerate remain free to change their minds indefinitely, the salvation of all is mathematically certain even if we grant the radical libertarian view of freedom” (ch. 8, “Escaping the Problem”). Reitan uses the analogy of rolling dice: the probability of never rolling a six approaches zero as the number of rolls approaches infinity. Freedom is not randomness, but if the choice to accept grace is genuinely possible at every opportunity, then across infinite opportunity the probability of eternal refusal approaches zero. Parry makes the ironic corollary: “The only way for God to keep the damned continually choosing to resist him forever is by shielding them from the real consequences of their actions, thus denying them the possibility of making a fully informed decision. But by doing this, God is arguably interfering with their freedom!” (The Evangelical Universalist, 3rd ed., ch. 1). In other words, the traditionalist who insists on eternal refusal may be the one who undermines freedom, not the universalist.
↑ 20. Thomas Talbott, The Inescapable Love of God, 2nd ed. (Eugene, OR: Cascade, 2014), chs. 11–13. Talbott’s key formulation: “As long as any ignorance, or deception, or bondage to desire remains, it is open to God to transform a sinner without interfering with human freedom; but once all ignorance and deception and bondage to desire is removed, so that a person is truly ‘free’ to choose, there can no longer be any motive for choosing eternal misery for oneself.” The crux of his argument is that once all such distortions are removed, there remains no motive for choosing misery. See R. Zachary Manis, Sinners in the Presence of a Loving God: An Essay on the Problem of Hell (Oxford: Oxford University Press, 2023), ch. 7, for a careful philosophical engagement with Talbott’s argument, and Chapter 17 for this book’s development of the question.
↑ 21. James 2:19 (ESV): “You believe that God is one; you do well. Even the demons believe—and shudder.” The verse shows that intellectual knowledge of God does not automatically produce love. This is important: the universalist here does not claim that seeing God forces love. The claim is that seeing God removes the ignorance that makes refusal seem rational, and that a will freed from that ignorance will, in the end, freely choose love. Talbott sharpens this into a dilemma for the traditionalist in The Inescapable Love of God, 2nd ed., ch. 11: “Either I am fully informed concerning who God is and the consequences of rejecting him, or I am not. If I am not fully informed, then I am in no position to reject the true God; and if I am fully informed, then (as Craig himself insists) I am incapable of rejecting the true God freely.” Craig concedes that, for some people, “the degree of revelation that would have to be imparted to them in order to secure their salvation would have to be so stunning that their freedom to disobey would be effectively removed.” This book’s answer avoids the dilemma by way of middle knowledge and optimal grace (see Chapter 14 and Chapter 15): God need not overwhelm the will with irresistible evidence but can instead compose the sequence of appeals that a free creature would, in the end, freely accept. See also Chapter 17.
↑ 22. For a clear defense of conditional immortality, see Edward William Fudge, The Fire That Consumes: A Biblical and Historical Study of the Doctrine of Final Punishment, 3rd ed. (Eugene, OR: Cascade, 2011). Fudge argues that immortality is a gift to the redeemed, not a natural property of all souls, and that the finally unrepentant simply cease to exist.
↑ 23. George Sidney Hurd, Extermination or Restoration? A Restorationist’s Response to Annihilationism (self-published, 2018), engages the conditional-immortality position from a universalist perspective. Hurd argues that while annihilationism is preferable to eternal torment, the full biblical picture points to restoration rather than extinction.
↑ 24. Reitan and Kronen, God’s Final Victory, §9.2 “Annihilationism and Soteriological Agnosticism,” pp. 179–81. They consider whether the doctrine of annihilation (DA) or the doctrine of soteriological agnosticism (DSA) offers a stronger alternative than the doctrine of hell. On annihilationism, they argue that if their earlier chapters succeed in showing that “creaturely freedom is no impediment to God saving all from eternal alienation, then they also succeed in showing that God need not annihilate them.” Furthermore, “finite suffering followed by infinite bliss has infinite positive value (the suffering is swamped by the good that follows). It is no act of mercy to deprive a creature of eternal joy in union with God for the sake of avoiding merely finite suffering” (p. 180). Their overall position is that even a reader who is not fully persuaded by the universalist case should at minimum reject eternal torment in favor of conditional immortality or soteriological agnosticism.
↑ 25. The moral case against eternal conscious torment is developed in Chapter 8 (what punishment is for), Chapter 9 (the proportionality problem), and Chapter 10 (the philosophical problem of hell). The key premise is that a finite life of finite sins cannot earn an infinite, unending punishment, and that the “infinite offense against an infinite God” premise does not carry the weight placed on it.
↑ 26. Hart, That All Shall Be Saved, Meditation Four. Hart draws extensively on Gregory of Nyssa, On the Making of Humanity and On the Soul and the Resurrection. Gregory’s argument that evil is inherently finite and that the rational will’s dynamism must in the end outstrip it is one of the most important philosophical arguments for universalism in the patristic tradition.
↑ 27. Talbott, The Inescapable Love of God, 2nd ed., chs. 11–13. Talbott builds his case through the inconsistent triad (ch. 4) and then presses it in his freedom chapters: once all ignorance and deception are removed, he argues, refusal loses its footing. His position is complex: he calls himself “an unrepentant libertarian” about the emergence of free moral agents, but he also identifies an insight in compatibilism “according to which moral freedom requires that we learn from our mistakes in a rational way that shapes our moral character and makes certain decisions psychologically impossible” (summarized in Robin Parry [Gregory MacDonald], The Evangelical Universalist, 3rd ed. [Eugene, OR: Cascade, 2025], ch. 1). This book follows Talbott’s structure (the inconsistent triad, the inescapable love) but stops short of his necessitarian conclusion. See Chapter 5.
↑ 28. Reitan and Kronen, God’s Final Victory, §1.2, pp. 4–11, and §9.1 “Summary of Our Argument,” pp. 178–79. In §9.1, they summarize their comparative defence by recapping the two versions of universalism they find most plausible: DU1 (God saves all by employing efficacious grace on a Thomistic view of freedom) and DU2 (God saves all by preserving creatures indefinitely free to choose communion with God on a Molinist view). In either case, “some version of DU is more plausible than any version of DH.” The humility of the conclusion mirrors this book’s: it is about relative plausibility, not deductive certainty.
↑ 29. Timothy Stratton, Human Freedom, Divine Knowledge, and Mere Molinism (Eugene, OR: Wipf & Stock, 2020). Stratton defines libertarian freedom as follows: “An agent’s choice, action, evaluation, or judgment that is not causally determined by something or someone else. The opportunity to exercise an ability to choose among a range of alternative options in a situation where antecedent conditions are insufficient to causally determine the outcome.” Stratton distinguishes sharply between certainty (an epistemic property) and necessity (a causal or metaphysical property): God’s knowing what a creature will do does not cause the creature to do it.
↑ 30. McClymond, The Devil’s Redemption, §12.1 “The Cumulative Argument: A Survey of Preceding Chapters” (V 2, pp 999ff.) and §0.10 (V 1, p 24); De Young, Exposing Universalism, Part 2 “The Appeal to Emotion and Reason” (chs. 4–5) and the section “Universalism More Deterministic and Coercive Than Calvinism” (pp. 217–19). De Young writes that universalism “decrees . . . that all must repent and come to faith, either before dying or after dying. God’s love triumphs over his justice; his love demands this (which is also strange, for how can love be demanding?)” (p. 218). Both critics argue that universalism inevitably slides toward determinism. This book’s answer begins here and is developed fully in the objection chapters (Chapter 22–Chapter 28).
↑ 31. The conviction that God will save all rests on the convergence of the arguments in Parts II through V. No single strand is a proof; together they constitute the cumulative case weighed in Chapter 29.
↑ 32. Philippians 2:12–13 (ESV). The passage beautifully captures the interplay of divine grace and human response that defines this book’s view: God works in us “both to will and to work,” and yet we are told to “work out” our salvation. The two are not in competition; they are complementary.
↑ 33. The distinction between logical possibility and reasonable expectation is not an evasion. It is the standard way that rational agents reason about future events under uncertainty. A logically possible outcome with vanishingly small probability does not warrant treating it as the expected outcome. See the discussion of the infinite-opportunity argument in Chapter 17.
↑ 34. The traditionalist claims not merely that a final “no” is possible but that it is actual, and, on the doctrine of eternal conscious torment, that it is actual for a significant (and in some formulations, the majority) portion of the human race. That is a claim about the actual outcome of history, not about bare logical possibility, and it is that claim this book contests.
↑ 35. Chapter 3 develops the character of God and the unity of the divine attributes. Every argument in this book flows from the claim that God’s love, justice, holiness, and power are not rivals but one.
↑ 36. Robin Parry [Gregory MacDonald], The Evangelical Universalist, 3rd ed. (Eugene, OR: Cascade, 2025), introduction. Parry writes: “In the original edition of this book I described myself as a ‘hopeful dogmatic universalist.’ That language has some potential to confuse, so I now speak of myself as a ‘confident universalist.’” He distinguishes this from “hopeful universalism” (the view that Scripture gives grounds for real hope but no certainty) and from necessitarian universalism (the view that salvation cannot fail as a matter of logical necessity). His posture is very close to this book’s: “I believe we can affirm the restoration of all things with confidence.” At the same time, Parry insists on humility: “All theological systems need to be offered with a degree of humility, and one that departs significantly from the mainstream Christian tradition calls for even more.”
↑ 37. Jerry L. Walls, “A Philosophical Critique of Talbott’s Universalism,” in Robin Parry and Christopher Partridge, eds., Universal Salvation? The Current Debate (Grand Rapids: Eerdmans, 2003); see also Walls, Hell: The Logic of Damnation (Notre Dame: University of Notre Dame Press, 1992), ch. 5. Walls argues that Talbott’s claim that “no illusion can endure forever” entails God’s imposing clear knowledge on the sinner through ever-increasing misery: “If God causes those persons who continue to rebel against him to grow ever more miserable and tormented, then it seems that God is imposing on those persons the clear knowledge that he is the source of happiness, and sin the cause of misery.” Walls concludes: “If God allows us to retain libertarian freedom, some illusions may endure forever.” Walls is not a universalist; this book agrees with his insistence on libertarian freedom but disagrees with his conclusion. See Chapter 11 and Chapter 15.
↑ 38. Andrew Hronich, Once Loved Always Loved: The Logic of Apokatastasis (self-published, 2023). Hronich draws on both Talbott and Reitan to argue that “rational freedom stipulates that true freedom necessitates true knowledge and true sanity of mind.” On this view, “a decision to reject God forever cannot be both free and fully informed.” The creature who thinks resistance to God is profitable “is operating under delusions and deceptions that arise from some prior delusion or deception.” Hronich’s development of the “rational freedom” concept is one of the strongest recent statements of this position. See also Reitan and Kronen, God’s Final Victory, ch. 7, on the incoherence of a free, fully informed rejection. See Chapter 17 for this book’s development of the argument.