Hypothetical universalism (HU) is one of those positions that sounds like it’s giving everybody what they want.

• You want to say “Christ died for all” in some real sense? HU says yes.
• You want to keep penal substitution and not turn the cross into mere moral influence? HU says yes.
• You want to keep election and not collapse into universalism? HU says yes.

So you get this neat little package:

1. Christ died for all people in some real, meaningful sense.
2. Christ’s death is penal and substitutionary.
3. Only the elect are finally saved; the atonement is applied only through faith.

And at first glance, it looks like the problem has been solved. We get to preserve the ability to indiscriminately preach, “Christ died for you,” while also retaining penal substitution and election. But what happens if we stop assuming those claims actually fit together and ask whether HU can really make them all cash out consistently?

Once we look more closely at the claims of HU alongside penal substitution, however, a serious problem begins to emerge: for a large class of people—the non-elect—the atonement becomes functionally fictitious, little more than an empty promise.

That’s the initial vacuity problem.

It isn’t a complicated philosophical trick. It’s just the result of asking basic questions like:

• What does “for” mean?
• What does an atonement actually do?
• What is the cross accomplishing prior to faith?
• What is the difference between “a payment” and “a coupon”?

Before we get into the theology, it helps to name the kind of problem we’re dealing with. It’s not exotic. It’s actually very common in modern discourse.

Matt Walsh’s What Is a Woman? became popular for one simple reason: he asked a basic definitional question that everyone assumed should be easy to answer, yet many of the responses he received were vacuous.

The most famous example is the circular answer:

“A woman is anyone who identifies as a woman.”

That answer sounds like it is saying something, especially when it is delivered confidently and backed by powerful social forces. But once you slow down and actually examine it, the answer has no informative content. It never tells you what a woman is. It defines “woman” by appealing to “woman” itself, so the definition simply assumes the very thing it is supposed to explain. In philosophical terms, the explanandum, the thing that needs to be explained, is simply smuggled into the explanans, the thing offered as the explanation. Rather than giving us something more basic in virtue of which we can understand what a woman is, the answer sends us right back to the original term. The explanation therefore moves in a circle:

• “Woman” is defined by “woman.”
• The definition depends on the term being already understood.
• So the statement doesn’t give you a criterion. it just gives you a tautology.

That is what I mean by vacuity. The language has the form of an explanation and sounds as though real content is being communicated, but once you slow down and ask what the words actually mean, there is nothing there to cash out. It is like asking someone what a dog is and being told, “A dog is a dog,” or “anything that identifies as a dog.” Those answers use the word you are asking to have defined, but they never give you the content of the definition. They sound like answers grammatically, but they do not actually tell you what a dog is.

And I think that is exactly the problem that begins to appear in hypothetical universalism. The problem is that the system allows you to retain certain familiar phrases like “Christ died for all,” “the legal barrier has been removed,” and “there is an atonement for everyone,” while making it increasingly difficult to say what those claims concretely amount to for the non-elect.

So the question I want to keep pressing is basically the theological equivalent of Walsh’s question:

What is an atonement?

And specifically:

• What does it mean to say Christ died “for” someone in a penal-substitutionary sense?
• What does that “for” do in the courtroom of heaven?
• What actually changes for that person as a result?

Because if the answer becomes something like:

“Christ atoned for them in the sense that He made it possible for them to be atoned for…”

Then we have fallen into the same basic pattern: the word is still being used, but its content has quietly disappeared. We retain the vocabulary while losing anything definite that the vocabulary is supposed to communicate.

Now, to be fair, not every statement that appears circular or repetitive is vacuous. Sometimes two expressions pick out the same reality while presenting it under different descriptions, and discovering their identity actually tells us something.

“Superman is Clark Kent” is a good example. Strictly speaking, it is not a tautology. It is an informative identity statement. “Superman” and “Clark Kent” initially present what appear to be two different referents under two different descriptions. Learning that Superman is Clark Kent collapses that distinction. You have learned something you did not previously know: the hero described one way and the reporter described another way are numerically the same person.

So the problem is not simply that the same reality appears on both sides of a statement. The question is whether the explanans actually adds anything to our understanding of the explanandum. “Superman is Clark Kent” does, because it identifies two previously distinct descriptions as referring to the same individual. “A dog is a dog,” by contrast, gives us nothing. It does not identify two independently understood descriptions, supply distinguishing properties, or tell us what makes something a dog. It simply hands the explanandum back to us as though it were an explanation.

That kind of statement can still be informative because it answers a genuine question:

Do these two different ways of referring to something actually pick out the same thing?

So I am not claiming that anything that sounds circular is automatically meaningless or vacuous. Sometimes the informational content lies precisely in establishing an identity between two things that were previously understood under different descriptions. What makes the statement informative is not mere repetition, but the discovery that two apparently distinct referents are in fact one and the same.

But that is not what is happening with a statement like:

“A woman is anyone who identifies as a woman.”

That statement does not bring together two independently understood ways of referring to the same reality, as “Superman is Clark Kent” does. There are not two descriptions whose identity is being disclosed. Instead, the thing that needs to be defined is simply placed back inside its own definition. To know what it means to “identify as a woman,” you already need to know what a woman is. But that is precisely what the definition was supposed to tell us.

So the definition moves in a circle. “Woman” is explained by “identifies as a woman,” while “identifies as a woman” can only be understood if we already know what “woman” means. The explanandum has simply been imported into the explanans.

And that is why the problem is not merely that the statement is circular. It is that the circle never gives us any independent content. There is no criterion telling us what makes someone a woman, no boundary distinguishing women from non-women, and no independently understood reality to which the term refers. The label ultimately explains itself by means of the same label.

That is the kind of vacuity I am flagging when I press hypothetical universalism. I am not arguing that every statement involving identity, repetition, or apparently circular language is meaningless. The question is whether the language can ultimately be cashed out in terms that tell us something substantive. HU can retain the form of substantive claims such as “Christ died for all,” “the legal barrier was removed,” or “atonement was made for everyone.” But if, when pressed, we cannot say what Christ actually accomplished for the non-elect, what legal liability was actually removed from them, or what this supposed atonement concretely does for them, then merely continuing to use the word atonement does not solve the problem.ing to supply the forensic content that would make those claims determinate in God’s courtroom.

1. What Does It Mean to Say “Christ Died for You”?

Everything turns on the meaning of that little word for.

People use it as though its meaning were obvious: “Christ died for you.” But the moment you say that you affirm penal substitution, you have already given that word a very specific kind of content.

Within a penal substitutionary framework, “Christ died for X” cannot merely mean:

“Christ did something general that might eventually become relevant to X.”

It means something concrete and forensic. Christ takes upon himself the guilt of X’s sins. He bears, in X’s place, the penalty those sins deserve. And because that penalty has been borne, God’s judicial claim against X on account of those sins has been satisfied.

That is what substitution means. A penal substitute does not merely create the possibility that the condemned person might someday escape punishment. He stands in the condemned person’s place and bears the penalty that otherwise belonged to him. The whole point of substitution is that the penal work is actually done for someone.

So if Christ truly dies for X in that penal sense, certain consequences follow. The law cannot continue to stand over X demanding satisfaction for the very sins Christ has already satisfied. God cannot justly exact the same penalty from Christ as X’s substitute and then exact it again from X. If the punishment has really been borne in full, then with respect to those sins there is no remaining penal debt to collect.

That is what gives penal substitution its actual content. It is not merely that Christ has done something beneficial, opened a possibility, or created an offer that may later become effective. PSA says that something objective has already happened in the courtroom of God. Someone’s guilt has been imputed to Christ, someone’s penalty has been borne in his place, and justice has been satisfied with respect to those sins. The atonement is therefore not waiting for faith to become substitutionary. Faith does not turn a general provision into a penal payment. The substitution has already taken place for the people whose guilt Christ bears and whose punishment he endures.

This is why the meaning of the word for is so important in any discussion of hypothetical universalism. If Christ dies for the non-elect in exactly this penal and substitutionary sense, then HU inherits the obvious problem of explaining how those same people can still be condemned for sins whose penalty has already been fully borne by their substitute. But if “Christ died for them” means something weaker than that, then simply repeating the word for does not preserve penal substitution. The word remains, but the penal-substitutionary content has been removed.

You can say “Christ died for all” as often as you like. The real question is: What does “for” actually mean?

2. Hypothetical Universalism’s Core Claim

Now plug HU’s universal claim into that structure.

HU wants to say that Christ died for every individual in a genuinely penal and substitutionary sense. At the same time, it also says that many of those same individuals die in unbelief and remain under God’s wrath forever.

Call one such person P.

HU therefore wants to affirm both of the following:

1. Christ died for P, where for retains its penal-substitutionary content.
2. P nevertheless remains liable for those sins and finally suffers their penalty himself.

At that point, the question is not especially complicated. What happened to P’s guilt?

If Christ truly bore P’s guilt and endured the penalty due to P’s sins, then justice has been satisfied with respect to those sins. But if P remains judicially liable for those same sins and God still exacts their penalty from him, then it is difficult to see in what meaningful sense Christ actually satisfied that liability on P’s behalf.

That is the pressure point. You cannot simply rise above it by repeating, “sufficient for all, efficient for the elect.” That may be a useful formula, but a formula is not yet an explanation. The underlying question remains untouched:

Did Christ actually satisfy divine justice for P’s sins or did he not?

If the answer is yes, then we need an account of how P can remain liable for a debt that has already been satisfied by his substitute. If the answer is no, then we need an account of what it means to say that Christ died for P in the same penal-substitutionary sense in which he died for the elect.

And once that question is pressed, HU is driven toward a three-way fork.

2.1. The Three-Way Fork

Once the question is put this way, there seem to be only three basic options.

1. Universalism
   Christ actually bears the full penalty due to every individual’s sins. Divine justice is therefore satisfied for every individual with respect to those sins. If there is no remaining penal liability, then no one can finally be condemned for them. The result is universal salvation.
2. Double punishment
   Christ actually bears the full penalty due to every individual’s sins, but some of those same individuals are nevertheless condemned and bear the penalty themselves. In that case, the same penal liability is satisfied in Christ and then exacted again from the sinner. God effectively demands payment twice for the same debt.
3. A weakened sense of “for”
   To avoid both universalism and double punishment, HU must qualify what it means when it says Christ died for the non-elect. Christ’s death cannot actually discharge their penal liability in the same sense that it discharges the liability of the elect. Instead, it does something more general: it provides a possible satisfaction, removes some conditional obstacle, or makes salvation available if some further condition is met.

And it is at this third option that the problem of vacuity begins to appear.

The point is not that HU verbally denies penal substitution. Quite the opposite. It continues to use all of the familiar language: Christ “died for” everyone, bore sins, made atonement, and removed the legal barrier. The problem is that when we ask what those claims actually amount to for the non-elect, the penal content begins to disappear.

HU does not have to deny PSA in words. The problem is that it risks draining PSA of its content precisely where its universal claim is supposed to matter.

If Christ did not actually discharge P’s penal liability, then in what distinctly penal-substitutionary sense did Christ die for P? That is the question the third horn still has to answer.

---

3. Where the Vacuity Lies

The easiest way to test HU’s universal claim is to ask what remains true when we turn from the elect to someone who is finally lost.

HU says that Christ “died for the non-elect.” But what does that actually amount to in their case?

Their guilt is never removed. Their legal standing before God is never changed. They remain under divine wrath, die in their sins, and finally bear the penalty themselves.

So what, exactly, did Christ’s “dying for them” accomplish for them in the courtroom of God?

Did it discharge their guilt? No.

Did it satisfy their penal liability in such a way that they are no longer answerable for those sins? No.

Did it secure their forgiveness? No.

Did it obtain any saving forensic benefit that ever actually belongs to them? No.

So what are we left with?

For the non-elect, the universal atonement begins to look entirely hypothetical. Christ’s death could have saved them had they believed, but in their actual case it never results in the removal of guilt, the satisfaction of their legal liability, reconciliation with God, or any changed judicial standing whatsoever.

And that is the initial vacuity problem.

HU wants to say that there is something objectively true of every person because Christ died “for” every person. But when we ask what is objectively different about the non-elect because Christ supposedly died for them, it becomes increasingly difficult to identify anything distinctly penal or substitutionary at all. They are guilty before the cross, guilty after the cross, under wrath before the cross, under wrath after the cross, liable to punishment before the cross, and liable to punishment after the cross.

What, then, has the universal for you actually added?

The irony is that the only part of the atonement that clearly has substantive forensic content is the part that actually removes guilt, satisfies liability, secures forgiveness, and results in salvation. And that is precisely the part HU ultimately restricts to the elect.

So in practice, the atonement is definite where it actually accomplishes something. The universal element survives largely at the level of language: Christ “died for” everyone, even though for the non-elect that death never actually alters their guilt, wrath, liability, or standing before God.

---

4. The “Legal Barrier Removed for Everyone” Reply

At At this stage, someone will usually respond:

“The atonement is not empty for the non-elect. Christ’s death removed the legal barrier between God and every sinner. The only thing preventing the non-elect from being saved is unbelief. So the cross really does accomplish something for everyone: it universally removes the legal barrier.”

That sounds like an answer until we ask the question doing all the work:

What exactly is the legal barrier?

If you affirm penal substitution, “legal barrier” cannot remain an undefined theological placeholder. It has to cash out in forensic terms.

4.1. What Is the Legal Barrier?

Within PSA, the barrier between the sinner and God is not some mysterious object standing between them. It is the sinner’s actual judicial condition before God: guilt, liability to punishment, and the law’s righteous claim to condemn.

That means removing the legal barrier has a fairly definite meaning. If Christ removes it for P, then Christ has dealt with the thing that made P legally condemnable in the first place. P’s guilt has been borne, his penal liability has been answered, and divine justice no longer has an unsatisfied claim against him for those sins.

Now universalize the claim:

“Christ removed the legal barrier for every individual.”

If that means what PSA ordinarily means by satisfying penal liability, then the problem immediately returns. If the legal claim against every sinner has actually been satisfied, on what basis do some of those same sinners remain legally condemned?

You cannot have it both ways. Either the barrier has actually been removed, in which case the relevant liability no longer remains, or the liability still remains, in which case the barrier was never actually removed.

At that point, “Christ removed the legal barrier for everyone” often turns out to mean something much weaker:

Christ made it possible for everyone’s legal barrier to be removed.

But notice the change. We began with an accomplished universal removal and ended with a universally available possibility of removal. Those are not the same claim.

And that is precisely the problem I have been pressing. The substantive language remains, but when we ask what it means in the case of the non-elect, the accomplishment retreats into a hypothetical.

4.2. Unbelief Does Not Solve the Problem

The next response is usually:

“The legal barrier has been removed, but the non-elect are condemned because they do not believe.”

But unbelief is not some morally neutral switch that merely prevents an otherwise completed transaction from taking effect. Unbelief is itself sin. It is part of the sinner’s guilt before God.

That creates another dilemma.

If Christ’s universal satisfaction includes the sin of unbelief, then unbelief cannot simply be introduced afterward as though it were an untouched penal liability capable of grounding final condemnation. The very sin being appealed to as the reason for condemnation would itself fall within the satisfaction supposedly made for that person.

But if Christ did not bear the non-elect person’s unbelief, then the decisive legal liability that finally condemns him was never removed in the first place.

So saying “the only thing left is unbelief” does not escape the problem. It simply relocates it. We still have to ask whether the guilt of unbelief was borne by Christ for that person.

If yes, why does that guilt remain condemnable?

If no, in what sense was his legal barrier comprehensively removed?

4.3. Moving the Removal to Faith Gives Up the Universal Claim

At this point, the position can shift again:

“Christ’s death secured the basis upon which God may forgive anyone, but guilt and condemnation are actually removed only when a person believes.”

That is a much clearer claim. But now notice what has happened.

The actual forensic removal is no longer universal.

The cross provides the ground or possibility of forgiveness universally, while the removal of guilt, change of legal status, and release from condemnation occur only in the case of believers.

And believers are the elect.

So the thing with genuine penal content ends up being particular after all.

For the non-elect, Christ’s death never removes their guilt, never changes their judicial standing, never releases them from condemnation, and never discharges the liability under which they finally perish. What is universal is not the actual removal of the legal barrier, but the possibility that it could have been removed under another condition.

That may be a coherent way to speak about a universal provision. But it is not the same thing as saying that Christ actually removed the legal barrier for every individual.

And once again, the universal claim begins to look verbal rather than forensic. The language says the barrier was removed for everyone. The actual legal reality says it was removed only for those who believe.

---

5. What Is the Atonement Before Faith?

HeThere is another way to expose the same problem, and it brings us directly back to the analogy with What Is a Woman?

The force of the question “What is a woman?” is that it does not allow someone to hide behind the continued use of the word woman. You have to tell us what the thing actually is. In the same way, hypothetical universalism can repeatedly tell us that Christ made “atonement” for everyone, but eventually we have to ask the more basic question:

What is an atonement?

What is this thing that Christ supposedly accomplished for every individual?

Everybody can agree to a rough formula:

atonement + faith = salvation

But that formula can conceal the very question we need answered:

What is the atonement, in itself, before anyone believes?

What does the atonement actually contribute to the equation? What has Christ objectively accomplished such that, when faith enters the picture, salvation follows?

On a robust penal substitutionary view, the answer is straightforward. An atonement is an objective satisfaction of divine justice. Christ bears the guilt and penalty of those for whom he dies. The satisfaction does not come into existence when someone believes. Faith does not complete the punishment, add something missing to Christ’s satisfaction, or turn a non-atonement into an atonement. Faith receives Christ and the benefits of a work that is already finished.

Faith does not make the atonement become an atonement.

But once HU universalizes the atonement, the same definitional problem we saw earlier begins to emerge. It is easy to say that there is an “atonement for everyone.” The question is what the word atonement means when applied to someone who is finally lost.

Suppose the atonement is “for everyone,” yet considered apart from faith it does not actually discharge any particular person’s guilt, change any particular person’s legal standing, or settle whether any particular person will finally bear his own penalty.

Then what exactly is this universal atonement?

Take our non-elect person P again. Faith never comes in his case. Christ’s atonement for P therefore cannot mean that Christ definitively satisfied P’s penal liability, because P remains liable and eventually suffers punishment himself. It cannot mean that P’s debt was cancelled, because the debt is still charged against him. It cannot mean that P’s case was settled in the courtroom of God, because his case ends in condemnation.

So what does atonement mean here?

This is where the parallel with “What is a woman?” becomes important. Simply repeating the word does not answer the question. Saying “Christ made an atonement for P” is not informative if, when asked what an atonement for P actually consists in, we are simply told that Christ did something that would have atoned for P if P had believed.

We have to know what the thing itself is.

Otherwise we end up with something like:

Christ performed a general act that could become savingly relevant to P if P believed, but which, considered by itself, does not settle P’s guilt, liability, or standing before God.

But now compare that with the original PSA claim.

The formula was:

atonement + faith = salvation

It now begins to look more like:

a universally available possibility of satisfaction + faith = actual satisfaction and salvation

What exactly is the first term contributing?

Faith is not meritorious. It does not bear wrath. It does not satisfy justice. It does not supply some missing portion of the penalty. So if Christ’s death, considered apart from faith, has not actually satisfied P’s penal liability, what happens when faith is added that suddenly makes the satisfaction real?

Where did the penal work occur?

If Christ already satisfied divine justice for P, then we are back to asking how P can later be punished for a liability Christ already satisfied.

If Christ did not actually satisfy divine justice for P unless and until P believes, then the universal atonement is not itself a completed penal satisfaction for P. What exists universally is at most a provision or possibility of satisfaction.

That is why the question should be pressed in exactly the same definitional way:

What is an atonement?

What is the thing to which faith is added?

What has Christ actually accomplished before faith ever enters the picture?

You cannot answer “What is an atonement?” merely by saying, “It is what Christ accomplished for everyone,” any more than you can answer “What is a woman?” by saying, “Anyone who identifies as a woman.” In both cases, the term under examination still needs content. The explanandum cannot simply be handed back to us inside the explanans.

If atonement means an actual satisfaction of divine justice, then we can ask whose liability has actually been satisfied. If it means something else under HU, then that something else needs to be defined. Merely retaining the word atonement does not establish that the thing being described still has the penal substitutionary content ordinarily associated with the word.

The Initial Vacuity Problem

The initial vacuity problem for hypothetical universalism therefore begins with an extremely simple question:

What is an atonement?

If an atonement is a real and objective satisfaction of divine justice in which Christ bears the guilt and penalty of those for whom he dies, then the same penal liability cannot both be satisfied by Christ and remain outstanding against the sinner.

HU cannot consistently say both:

Christ actually bore and satisfied the penal liability of every individual.

and:

Many of those same individuals remain liable for those very sins and finally suffer their punishment themselves.

If Christ really satisfied the liability of every individual, universalism threatens. If God nevertheless exacts that same liability from the sinner, double punishment threatens. If neither follows because Christ did not actually discharge the non-elect person’s liability, then we have to return to the definitional question: in what sense was there an atonement for him?

HU can preserve the sentence:

“Christ made atonement for every individual.”

But the sentence itself is not yet an explanation. We still have to ask what atonement picks out in the case of the non-elect.

His guilt remains. His liability remains. His condemnation remains. He remains under wrath and finally suffers the penalty himself.

If all that can be said is that Christ’s death could have resulted in the removal of those things had he believed, then what is universal is not an accomplished penal satisfaction for every individual. What is universal is a possibility of salvation grounded in Christ’s death.

That may be a substantive theological position, but calling that possibility an “atonement for P” does not by itself show that Christ actually atoned for P in the penal substitutionary sense.

This is why the definitional question matters so much. Just as the question “What is a woman?” forces us to move beyond the mere repetition of a label, the question “What is an atonement?” forces HU to tell us what its universal atonement actually consists in.

Once we ask where guilt is actually removed, where penal liability is actually satisfied, and where wrath is actually exhausted, HU gives those substantive benefits only to the elect.

Further Suggestion:

Christ Died for Sins, Not Categories

The Intrinsic Power of Christ’s Atonement: A Case for Particular Redemption