Chapter 22

The Free “Yes”

Optimal Grace, Middle Knowledge, and the One Open Question

The Question on the Table

A friend of mine, a pastor of thirty years, told me the kindest version of the hardest objection I have ever heard. He said it over coffee, not in a debate. “I want this to be true,” he said. “But if God guarantees the ending, are we still free? And if we are free, how can He guarantee the ending?”

That is the question this chapter must answer. It is one word long: freedom.

We have spent twenty-one chapters building a case. God is love (Chapter 2). Christ died for all (Chapters 8–12). God desires all people to be saved (Chapters 13–17). And God’s purpose cannot be defeated (Chapters 18–21). But the whole case leans on one hinge. If freedom is real, can the outcome be certain?¹

The objection deserves its full weight. It has come from some of the finest minds in the tradition. And its motive is noble: it wants to protect the reality of love. Love that is forced is not love. A “yes” that could not have been “no” is not a real “yes.” If universal salvation means that God drags everyone to Himself whether they want to come or not, then the tradition is right to refuse it. I would refuse it too.²

But what if that is not what it means? What if the God who made you knows you better than you know yourself, and knows exactly what it would take for you to say yes freely? What if grace does not override the will but heals it? What if time is not the enemy of freedom but its friend?

This chapter answers in three layers. First, grace: God’s appeal is not generic but optimal, fitted to each heart by a God who knows each person fully. Second, freedom itself: free choices run on reasons, sight, and desire, and grace that heals blindness does not override the will but sets it free. Third, honesty: I will face the one open question this book cannot close, name the fallback, and tell you why I hold my conviction anyway.

Why It Matters

The freedom question is not a side issue. It is the best objection the tradition has. It is better than the judgment texts (those are honestly engaged in Appendix D). It is sharper than the “all kinds of people” reading of 1 Timothy 2:4 (answered in Chapter 13). It strikes at the root of the hope itself, because it asks whether what I am claiming is even logically possible.³

Here is the sharp version of the charge. If God ensures that every person eventually says yes, then what we are describing is irresistible grace applied to everyone. That is Calvinist soteriology taken universal. The Calvinist at least limits the scope. The universalist keeps the mechanism and removes the limit. But the mechanism is the problem. Irresistible grace, by definition, overrides the person’s own choice.⁴

I want to honor what is right in this worry. The tradition is guarding something precious: the reality of the person before God. A love that bulldozes consent is not love. A salvation that bypasses the will is not salvation. If I cannot answer the freedom objection without giving up freedom, then I cannot answer it at all, and everything from Chapter 2 forward collapses.

So let me say at the outset what I am not arguing. I am not arguing that God forces anyone. I am not arguing that the final “yes” is anything less than each person’s own, given freely, from the heart. And I am not arguing that the outcome is guaranteed the way a mathematical theorem is guaranteed. I am arguing something more careful than that, and more honest.

The stakes are personal. If God’s saving will is real (1 Timothy 2:4), and if God’s purpose cannot be thwarted (Isaiah 46:10), then either freedom is an obstacle God cannot get past, or freedom is something God knows how to work with. The first option makes God weaker than the problem. The second option makes grace wiser than we imagined. I think Scripture points to the second.⁵

The Response

1. The Objection at Full Strength

Let me state the case against my position as strongly as I can. A thoughtful Calvinist will put it like this: if universal salvation is certain, then the choice is not genuinely free. A thoughtful Arminian will put it like this: if the choice is genuinely free, then universal salvation is not certain. Both agree on the premise: certainty and libertarian freedom cannot coexist.⁶

The tradition is not being stubborn here. It is being consistent. If love requires that the beloved can say no, and if some people say no, then love must accept the loss. Jerry Walls, one of the most careful thinkers on this question, argues that God gives every person an optimal measure of grace, and yet holds that final loss remains possible. A decisive rejection of God is, in his view, a settled response, one made with true understanding of what one is rejecting. On this view, the door to heaven is locked from the inside, and God will not break it down.⁷

I respect that position deeply. Walls is not my opponent; he is an ally on nine-tenths of the ground. His doctrine of optimal grace is one of the most important building blocks of this chapter. Where I part from him is on whether a refusal, given enough light and enough time, can actually remain settled forever. But I want to earn that disagreement, not assume it.

That is a fair challenge. I will answer it in the moves that follow, but I want the reader to feel its weight first. The tradition’s motive here is to protect the reality of love. That motive deserves honor. The question is whether the tradition has drawn the right conclusion from it.⁸

2. The Optimal-Grace Move

The first layer of the answer is grace. Not generic grace. Not a one-size-fits-all offer tossed into the universe and left to land where it may. But grace that is, in Walls’s own words, “deeply personal and individual.”⁹

“Optimal grace is not an abstract, impersonal concept. To the contrary, it is a deeply personal notion. It means that God knows each of us as individuals and understands how best to speak to us and reveal himself to us so that we will return his love.”

— Jerry L. Walls, Purgatory: The Logic of Total Transformation, ch. 5

Walls contrasts this with a merely “cursory” offer of salvation. God does not extend grace “merely in a cursory fashion so that everyone has some minimal chance, in order to establish God’s justice in damning them if they reject it.” The offer is not a formality. It is a courtship. God knows what would reach each heart, and He brings that appeal with patience, skill, and love.¹⁰

James Beilby makes the contrast even sharper. He notes that some theologians speak of God giving “sufficient grace,” meaning grace that is enough to justify God’s condemnation if it is refused. Beilby asks the obvious question: “Sufficient for what?” If the grace is calibrated to justify a verdict rather than to win a heart, then it is not really grace at all. It is a legal formality.¹¹

This distinction matters enormously for the freedom question. Think of it this way. If God gives every person optimal grace, then the person who says “no” is not saying no to a distant, unclear, half-heard offer. That person is saying no to the best possible appeal a perfectly loving God could make, fitted to that particular heart. And if that person is given not just one chance but every chance that divine patience can provide, then the question is no longer whether God is trying hard enough. The question is whether any refusal can really survive that kind of love, given enough time. I will come back to that question in move five.¹²

Notice what optimal grace does not do. It does not bypass the will. It speaks to the will. It appeals to the person’s own reasons, desires, and capacity to see the good. God knows how best to draw a positive response from each person, without destroying or overriding their freedom.¹³ The person’s “yes” remains the person’s own. The difference is that it is a “yes” given in response to the clearest, most loving, most patient appeal imaginable (see also Chapter 16).

3. The Middle-Knowledge Move

How does God know what would reach each heart? This is where middle knowledge comes in. Middle knowledge is God’s knowledge of what each person would freely choose in any situation God might place them in. It is not knowledge of what God forces them to do. It is knowledge of what they would do freely.¹⁴

The idea comes from the sixteenth-century Jesuit theologian Luis de Molina. Theologians sometimes describe three “moments” in God’s knowledge (not moments in time, but logical stages). First, God knows everything that could happen: all possibilities (natural knowledge). Second, God knows what each free creature would do in any possible set of circumstances (middle knowledge). Third, God knows what will happen, because He has chosen which circumstances to create (free knowledge). Could, would, will.¹⁵

Kenneth Keathley has given one of the clearest accounts of this framework. He sums it up plainly: from the infinite set of possible worlds that could happen, there is an infinite subset of feasible worlds which would accomplish God’s purposes. God freely chooses one of the feasible worlds, and He perfectly knows what will happen. On his account, God’s knowledge of what each person would freely do allows God to arrange the world in a way that respects freedom while achieving His purposes. In the Molinist model, God controls all things, yet humans possess real freedom.¹⁶

Timothy Stratton has pressed the case for libertarian freedom further. His “FreeThinking Argument” contends that if our thoughts are wholly determined by prior causes, then we have no reason to trust our own reasoning. If something outside you always determines exactly what you think and believe, then you are not really reasoning at all. You are a passenger, not a pilot. The very act of rational thought, Stratton argues, requires that we possess freedom that is not determined by forces outside ourselves.¹⁷

What does this mean for universal salvation? If God knows what each person would freely do in any situation, then God can order circumstances, timing, and encounter so that each person’s free “yes” is actually reached. Not forced. Reached. The difference is the difference between a parent who locks a child in a room and a parent who knows the child well enough to set the table with the child’s favorite meal. The child still chooses to sit down. The parent simply knew what would work.¹⁸

Kirk MacGregor has taken this connection between Molinism and universal salvation further than most. In his chapter on the relationship between Molinism and eventual universalism, MacGregor defends three claims. First, Molinism makes the traditional view (that some are lost forever) logically possible. Second, Molinism also makes eventual universalism logically possible. Third, and remarkably, if libertarian freedom is true, then Molinism is the only way God could guarantee universal salvation. Neither simple foreknowledge nor open theism provides the resources to get there.¹⁹

MacGregor engages R. Zachary Manis’s argument that there may be no “feasible universalist world,” no possible arrangement in which all free creatures are saved. MacGregor replies that if God knows in His middle knowledge that there are feasible worlds where all eventually embrace God, then an all-loving God would create one. The probability that such a world exists is, in MacGregor’s judgment, beyond human assessment and therefore cannot be called improbable. His conclusion: on Molinist grounds, eventual universalism is logically possible and fitting for a God of love.²⁰

I should be clear about two things. First, MacGregor is not himself a dogmatic universalist. He argues for the possibility and fittingness of eventual universalism on Molinist grounds. He would very much like to be wrong about his own disbelief in it. And he notes that the Molinist philosopher Alvin Plantinga has mused that perhaps “people get second chances, third chances after death, fourth chances, nth chances” and that “maybe it will be that in the long run nobody finally turns God down.”²¹ Second, middle knowledge is not the foundation of my hope. The foundation is the character of God and the witness of Scripture, built across Parts One through Three of this book. Middle knowledge is a frame that shows how the hope can be both real and non-coercive. A reader who rejects Molinism can still hold the thesis of this chapter on other grounds. But for those who find the framework compelling, it removes one of the strongest roadblocks: the claim that guaranteed salvation requires guaranteed coercion.²²

4. The Freedom-Analysis Move

Now we must go deeper. What is a free choice? This is where the freedom objection starts to lose its footing, because the objection assumes a picture of freedom that does not survive close inspection.

A free choice is not a random event. It is not a coin flip. When I choose something freely, I choose it for reasons. I see something as good, or as desirable, or as the best available option, and I move toward it. My reasons may be confused. My sight may be clouded. My desires may be distorted by habit or fear or lies I have believed about myself. But there is always a why. A choice with no reason at all is not an exercise of freedom; it is a spasm.²³

Andrew Hronich presses this point with great care. He argues that human beings are naturally ordered toward union with God. God is the end (the Greek telos, “purpose”) toward which our deepest longings point. When we choose against God, we do not choose against our own nature on purpose and in full sight. We choose against God because something is blocking our vision: ignorance, deception, or bondage to desire.²⁴

These are what Hronich, drawing on the work of John Kronen and Eric Reitan, calls “salvation inhibitors.” They are the obstacles that block a free “yes” to God. They are not freedom itself. They are freedom’s enemies. When grace removes ignorance, it does not take something away from the person. It gives something back. When grace breaks the power of addiction to a false good, it does not reduce freedom. It restores it.²⁵

Kronen and Reitan develop this further. They argue that true rational freedom with respect to the choice of salvation is possible only for those who actually see God clearly. Before that moment, choices about God are always made under some degree of ignorance, and ignorance limits freedom. If this is right, then God can bring all of a person’s motives into alignment with the choice of loving union, and when He does, the resulting choice is both inevitable and free. It is free in the way that the saints in heaven are free: not because they could sin but choose not to, but because their healed character no longer has any motive for sin.²⁶

Hronich draws a conclusion that deserves careful thought: to have rejected God is to never have known Him, for if one truly knew God, one would see Him as the end that satisfies all the creature’s deepest longings.²⁷ He does not deny that a creature can temporarily reject God. He denies that a creature can do so finally, with full knowledge and full freedom, because such a choice would require reasons that the truth dissolves.

Consider a biblical example. James 2:19 tells us that the demons “believe and shudder.” They see God clearly. They know He is real. Their knowledge does not make them love God. Sight alone is not coercion. But sight does remove one of the great inhibitors: ignorance. And when ignorance is removed, the reasons for refusal begin to starve.²⁸

5. The Starvation-of-Refusal Move

Here is the move that ties the argument together. I want to ask a simple question: can a “no” last forever?

Not a “no” for now. We all say no to God sometimes. That is how sin works. The question is whether a “no” can be sustained without end, against the full light of truth, against the full appeal of love, against the full weight of reality, for all eternity.

Every actual refusal runs on some reason. Fear. Misunderstanding. Wounded pride. Love of a rival good. Hatred born of pain. These are the fuels of a “no.” And every one of them is finite. Fear dissolves in the presence of perfect love (1 John 4:18). Misunderstanding breaks under truth. Pride collapses when it sees the God who washed feet. The rival good loses its shine when the real good stands in front of you. These are not infinite resources. They run out.²⁹

Thomas Talbott has pressed this argument from several angles. He argues that no illusion can endure forever, because the more one rebels against God, the more miserable the rebellion becomes, and the more incentive one has to repent. The connection between sin and misery is not a punishment imposed from outside; it is the nature of the thing. Sin is the attempt to find happiness apart from the source of happiness. The attempt cannot work, and given enough time, the person will see that it cannot work.³⁰

Robin Parry puts the point sharply. If God fully reveals Himself and clarifies the real consequences of refusal, then continuing to resist God becomes increasingly difficult. And here is the twist: the only way for God to keep someone in permanent rebellion would be to shield them from the truth, to keep them deceived. But that would be God interfering with their freedom, not protecting it. As Parry argues, the traditional view that insists on permanent rebellion must explain how a person can keep rejecting God forever without God Himself blocking the person’s access to the full truth.³¹

Walls raises a serious counter to this. He argues that self-deception is a real power, and that the choice for evil can become a settled disposition in which a person perverts his moral freedom so thoroughly that he no longer desires the good. On this view, some illusions might endure forever, because the person has chosen to look away from the light and can keep choosing to look away.³²

I take that worry seriously. But I think it underestimates two things. First, it underestimates the God of Luke 15, who does not wait for the lost sheep to find its own way back but goes after it “until he finds it” (Chapter 4). Second, it underestimates the scope of time. Scripture never sets God’s patience a clock. “With the Lord one day is as a thousand years, and a thousand years as one day” (2 Peter 3:8; see Chapter 14). The God who is Himself temporal, as R. T. Mullins has argued, does not experience eternity as a timeless now but as real duration in which real engagement with creatures is possible.³³ That means the courtship is not notional. It is real. And it has all the time love needs.

The starvation argument does not say that refusal is impossible at any given moment. It says that a refusal sustained forever requires infinite fuel, and the fuels of refusal are finite. Time is on grace’s side. The light keeps shining. The love keeps coming. And the reasons to say “no” keep shrinking.³⁴

6. The Honest-Limit Move

Now I must say the hardest thing in this book.

I cannot prove that every person will say yes. The bare logical possibility that some creature could refuse God forever cannot be ruled out by force of argument. A free “yes” requires a possible “no.” And a possible “no” means that the “yes” is never a mathematical certainty. If I claimed otherwise, I would be giving up the very freedom I just spent five moves defending.³⁵

This is the book’s one open question. I have named it before (Chapter 1), and I will name it again in the conclusion (Chapter 24). I do not pretend it is not there.

But I hold my conviction anyway, and here is why. The probability that any refusal is final is, on the account I have given, vanishingly small. Every fuel of refusal is finite. God’s love is infinite. God’s patience is infinite. God’s knowledge of each heart is perfect. And Scripture does not merely say that God hopes for the salvation of all; it says that God desires it (1 Timothy 2:4), that His purpose will stand (Isaiah 46:10), and that at the end “God may be all in all” (1 Corinthians 15:28; Chapter 19). A conviction grounded in these texts and in the character of God can be full-hearted without being a logical proof.³⁶

And if the open question were ever to cut the other way? If some creature, against all reason and all love, refuses forever? Then I hold the fallback I have named throughout this book: conditional immortality. The finally unwilling, if such there be, are not tormented forever. They cease. They are not preserved in agony for all eternity as a monument to divine wrath. God does not do that. The fallback is annihilation, never eternal conscious torment (see Appendix D).³⁷

One more distinction. David Bentley Hart arrives at the same destination I do, the salvation of all, but by a different road. Hart argues that universal salvation is not merely overwhelmingly probable but metaphysically necessary. On his view, a truly free choice is, in the end, only one: the choice of God, because God is the only object that can satisfy the rational will. To choose against God with full sight would not be freedom but a kind of madness, and madness is not a choice at all.³⁸

I am grateful for Hart’s exegetical and historical work. His reading of the Greek “all” and “world” texts is among the finest available, and this book has drawn on it freely (Chapter 9; Chapter 19). But I part company with him on this one point. I keep libertarian free will. I hold that the abstract possibility of final refusal is real, even if the probability is negligible. My hope is conviction, not theorem. Freedom stays real. And that is exactly what makes the “yes” beautiful.³⁹

Engagement with Specific Scholars

The lines of this debate are drawn between two groups, and I want the reader to see exactly where each one stands.

On the side of the argument I am making, the key voices belong to Thomas Talbott, Andrew Hronich, Robin Parry, and (with the qualification just given) David Bentley Hart. Talbott’s argument that clear sight of God’s goodness makes a final rejection unintelligible has been the most debated version of the case. Robin Parry has defended it with an important twist: God’s self-revelation does not force the will; it simply clarifies the situation in a way that the will changes. And as Parry notes, such clarification is completely different from forcing someone to act against his or her will.⁴⁰ Hronich adds the mechanism: efficacious grace removes the salvation inhibitors, and the resulting choice, though inevitable, is free because all the person’s motives now point in the same direction (see Chapter 21 for the formal argument).⁴¹

On the middle-knowledge side, Keathley, Stratton, and MacGregor provide the philosophical machinery. Keathley and Stratton are not themselves universalists; they defend Molinism as a framework for understanding God’s sovereignty and human freedom. MacGregor goes further, arguing that eventual universalism is logically possible and fitting on Molinist premises, and that if libertarian freedom is true, Molinism is the only model that could deliver it. He concludes that the matter “can only be decided on scriptural grounds; philosophical considerations do not apply.” I use their work with gratitude and without pretending that they endorse my conclusion.⁴²

Kronen and Reitan give the universalist case its most careful philosophical form. In God’s Final Victory, they argue that God can bestow efficacious grace by removing all salvation inhibitors. Far from interfering with rational freedom, such grace is its fulfillment. Their work on the Argument from Efficacious Grace is developed in Chapter 21 and forms the philosophical backbone of the freedom analysis I have given here.⁴³

On the other side, the strongest voices are Walls, William Lane Craig, and Manis. Walls argues that optimal grace is real, but that a decisive rejection under optimal grace is also possible. Craig has pressed the point that sinful irrationality can resist even the strongest evidence. Manis argues that even on Molinist premises, there may be no feasible world in which all free creatures are saved.⁴⁴

I think Walls is right about optimal grace and wrong about the durability of refusal. I think Craig is right that sin produces extraordinary irrationality, but I think he underestimates the God who meets irrationality with patience and wisdom (Chapter 14). And I think Manis’s argument, as MacGregor has shown, does not survive scrutiny once we allow for the possibility that God’s engagement with each creature extends beyond the boundaries of earthly life.⁴⁵

Counter-Objections

“A grace that always wins in the end is irresistible in slow motion.”

This is the strongest remaining objection, and it deserves a clear answer. The charge is that if grace always wins, then the outcome is determined, and calling it “free” is just marketing.

Here is the difference. Irresistible grace, as the Calvinist tradition defines it, bypasses the will. God changes the person from the inside without the person’s consent. Optimal grace does not bypass the will. It persuades the will, through the person’s own healed reasons and genuine consent, at whatever pace that consent requires.⁴⁶

The test is simple. Is the final “yes” the person’s own? On the account I have given, the answer is an unqualified yes. The person says yes because the person has come to see the truth, to feel the love, and to want the good. No one has overridden anything. God has removed the obstacles, healed the sight, and then waited for the person’s own heart to respond. That is not irresistibility. That is courtship.⁴⁷

“Hart is more honest. If refusal is really possible, your ‘all will be saved’ overreaches.”

This objection says that if I admit the logical possibility of final refusal, I have no right to speak with confidence about universal salvation. Either it is necessary (Hart) or it is uncertain (and I should stop calling it a conviction).

I understand the force of this. But a conviction grounded in God’s character and competence can be full-hearted without being a modal proof. We hold many convictions this way. I am convinced that the sun will rise tomorrow, though I cannot prove it will. I am convinced that my wife loves me, though I cannot construct a syllogism that rules out every alternative. Scripture’s own oaths and purpose clauses (Chapter 13; Chapter 18) license confidence exactly this strong.⁴⁸

“Then preach the fallback and drop the confidence.”

The final pushback says: if conditional immortality is the honest fallback, then be consistent and preach annihilationism rather than universalism.

No. The fallback exists for intellectual honesty, not for hedging the hope. The texts that sing of universal restoration are not singing about annihilation. “Every knee shall bow, and every tongue confess that Jesus Christ is Lord” (Philippians 2:10–11; Chapter 19). “As in Adam all die, so also in Christ shall all be made alive” (1 Corinthians 15:22). “That God may be all in all” (1 Corinthians 15:28). These are not texts about the disappearance of the lost. They are texts about the victory of love. I believe what they say. I face the one open question. And I preach the hope, not the fallback.⁴⁹

Conclusion and Pastoral Application

Let me pull the threads together.

The freedom objection asks whether God can guarantee the ending without destroying the choice. The answer is that God does not guarantee it by force. He pursues it by love. His grace is optimal: fitted to each person, patient, wise, unhurried. His knowledge is perfect: He knows what each person would freely do in any circumstance, and He arranges the encounter accordingly. Freedom itself, rightly understood, is not the enemy of the hope but its condition. A free choice runs on reasons, sight, and desire; and grace that heals the reasons, clears the sight, and liberates the desire does not override the will. It sets the will free.

Can a “no” last forever? Every refusal runs on finite fuel: fear, ignorance, pride, deception. And every fuel runs out under infinite light, infinite patience, and infinite love. Time is on grace’s side. Scripture never sets God’s patience a clock.

Can I prove it? No. Freedom is real, and a real “yes” requires a possible “no.” This is the one open question, and I face it without flinching. But I hold my conviction, because the God of this book, the God whose character is love (Chapter 2), whose cross covers all (Chapter 10), whose will reaches every heart (Chapter 16), whose purpose cannot be stopped (Chapter 18), does not lose the last argument with any heart He made.⁵⁰

And this matters for how we live. It matters for how we pray, because we can pray with confidence, not just with fingers crossed. It matters for how we grieve, because we do not grieve as those who have no hope (1 Thessalonians 4:13). It matters for how we evangelize, because we are heralds of a victory, not salesmen working against a deadline. And it matters for how we worship, because the God we worship is not a God who loses anyone but a God who wins every heart.

Paul wrote to the Philippians: “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). Both halves of that sentence are true. We work. God works in us. The “yes” is ours. And the power behind the “yes” is His.

That is the free “yes.” Not forced. Not fake. Not a theorem. A heart set free by love, saying the only word a healed heart can say.

Love never fails.⁵¹