‘What I am going to say is not a dogma of faith but my own personal view: I like to think of hell as empty; I hope it is’ (quoted in King Reference King2024). These words of Pope Francis on an Italian television programme might seem incompatible with the Catholic Church’s teaching on the reality of hell, but in fact the catechism of the Catholic Church supports Pope Francis. For while it ‘affirms the existence of hell and its eternity’, it also says, ‘The Church prays that no one should be lost… If it is true that no one can save himself, it is also true that God “desires all men to be saved” (1 Timothy 2:4), and that for him “all things are possible” (Matthew 19:26)’ (quoted in King Reference King2024). We might call such a position on hell ‘hopeful universalism’, in contrast to ‘believing universalism’. While believing universalists believe that all created persons will be saved, hopeful universalists lack this belief but hope it is true. Why wouldn’t hopeful universalists bring their belief into line with their hope? Perhaps because they want to be faithful to the Church’s teaching. Yet something keeps them hoping that what they don’t believe is in fact true – maybe an honest acknowledgement that none of us are deserving of salvation and a kind of solidarity with other human beings, or love for particular people they do not wish to be eternally damned. While I am not a hopeful universalist myself (I am a believing universalist), and so I can only guess at their motivations, I wish to defend this position – which I find reasonable – from a recent attack. Michael Rea has argued against hopeful universalism, maintaining that hopeful universalists should become believing universalists. More particularly, Rea focuses on those he describes as having ‘considered unconditional hope that universalism is true together with an absence of belief that it is true’ (Reference Rea2022, 266), where ‘considered hope’ is that which is reflectively embraced, and universalism is the view that the salvation of all humans is ‘guaranteed by God rather than being contingent on its just happening to turn out that all humans do whatever it is (if anything) that human beings must freely do in order to be saved’ (Reference Rea2022, 267). While I wish to defend the hopeful universalist, I will limit myself to considering those who (perhaps like Pope Francis) lack belief in universalism, rather than those who have a belief that universalism is false.

Rea’s argument against hopeful universalism begins with an analysis of hope as follows: ‘To hope for a state of affairs to obtain is, in part, to desire that it obtain’ (Reference Rea2022, 268). So, the hopeful universalist desires that universalism is true. But, Rea notes, one who worships God will ‘desire to… see God’s will done on earth’ (Reference Rea2022, 270) and have a second-order desire to prioritise the satisfaction of this desire. And the hopeful universalist, recognising that universalism might be false, believes that universalism might not be the will of God. So then, Rea concludes, the hopeful universalist must have a first-order desire (that universalism is true) that is inconsistent with a second-order desire ‘to prioritize the desire to see God’s will be done, whatever it might be’ – including the falsity of universalism (Reference Rea2022, 272).

Besides this way of putting the argument, Rea suggests another, which also begins with the analysis of hope as involving desire. Rea then notes that desire often presupposes the goodness of what is desired. While this might not always be true, he reasons that it is in the case of ‘considered theistic desires’, or ‘the considered desires of someone who worships God, where God is understood to have the traditional theistic attributes of omnipotence, omniscience, and perfect goodness’ (Reference Rea2022, 269). Rea writes, ‘The premise I want to defend… is that rational considered theistic desires always presuppose the goodness of what is desired’ since the worshipper will only rationally endorse desires for things she takes to be good (Reference Rea2022, 269). This means the worshipper who hopes for universalism presupposes the goodness of universalism. But God’s will, the worshipper must believe, is all-things-considered good. And so one who does not believe universalism to be true must not presuppose that universalism is all-things-considered good. But then we have a contradiction, between presupposing the goodness of universalism (the object of one’s rational considered theistic desire), and not presupposing its goodness (because one lacks belief that it is God’s will).

My response to both versions of Rea’s argument depends on the distinction between God’s antecedent and consequent will. Some understand this distinction to depend on the involvement of human volition. Eleonore Stump, for instance, writes, ‘God’s antecedent will is what God would have willed if everything in the world had been up to God alone. God’s consequent will is what God actually does will, given what God’s creatures will’ (Reference Stump2022, 4). I want to draw the distinction differently (and in a way that I think follows AquinasFootnote ¹). God’s consequent will I understand similarly to Stump – as what God actually does will. But God’s antecedent will I understand to be not a matter of God’s counterfactual willing, but of God’s valuing or judging good. What God antecedently wills is what God considers intrinsically good, whereas what God consequently wills is what God considers all-things-considered good. On Stump’s way of drawing the distinction, God’s antecedent and consequent wills only come apart because free creatures have a way of messing things up: God antecedently wants a world without any evil but consequently wills a world with much suffering because of the sinful choices of humanity. But on my way of drawing the distinction, God’s antecedent and consequent wills may come apart over things other than the results of free creaturely choices. For instance, God might antecedently will that no creature suffer but consequently will a world like ours with much animal suffering for some greater good (natural laws, or evolution, or who knows what) that has nothing to do with (and, indeed, occurred long before the occurrence of) any creaturely free choice.

Now, when we desire, or pray for, or attempt to align our will with, God’s will, there is a question of whether we mean God’s antecedent or consequent will. Different traditions may disagree about this question. Consider the following passage in Jewish law:

Noam Oren has argued that the ‘vanity’ of such prayers (for events that are already determined, but we do not know how) is not a matter of metaphysical impossibility, but ethical impropriety, as a violation of the Mishna principle that ‘One recites a blessing over the good just as one recites a blessing over the bad’. Following Moshe Halbertal, Oren reasons that ‘one should refrain from offering a vain prayer, not because it is ineffective, but because it reflects poor religious character. A truly religious person is meant to accept God’s actions in the world, and by praying for a miracle or for the past to be changed, one is, in fact, rebelling against God at a moment when submission is called for’ (Reference Orenforthcoming). This suggests that at least when it comes to petitionary prayer regarding events that are determined, we should aim to align our will with God’s consequent will (or at least refrain from praying otherwise). If one takes much, if not all, of the natural world to be determined, the consequences of following this principle are significant. If a hurricane is predicted to hit one’s island community tomorrow, and one takes hurricanes to be deterministic (if chaotic, and therefore unpredictable) natural systems, then one ought not to pray for one’s home to be spared.

Of course, it is an additional step to conclude that one ought not to hope for this – perhaps in the Jewish tradition it is permissible to privately hope for something that it is not permissible to request of God; Oren does not comment on this question, since he is focused on the ethics of requests. Other traditions, such as at least some forms of Buddhism, go further to discourage their adherents even from desiring what may turn out not to be the case. But I want to argue that one ought not to hope for or try to align one’s will with God’s consequent will. Eleonore Stump has likened those who try to align their will with God’s consequent will to the Whatever Faction for Mao Tse-tung, a group that was ‘committed to maintaining as true, and compulsory for all Chinese to believe, anything Mao said, whatever it was. In trying to desire whatever happens as good because God wills it, a person is as it were trying to be part of the Whatever Faction for God’ (Reference Stump, Rea and Pojman2008, 345). The problem with this, Stump reasons, is that it is ‘to accept as intrinsically good even those things which God wills as good only secundum quid, that is, as the best available in the circumstances. But God does not will as intrinsically good everything that he wills; what he wills in his consequent will, what is the best available in the circumstances, might be only the lesser of evils, not intrinsically good’ (Reference Stump, Rea and Pojman2008, 346).

But why shouldn’t our acceptance of God’s consequent will be a form of submission and thus true worship, as the Jewish tradition suggests? Stump offers one answer to this question in her discussion of what she calls the ‘desires of the heart’. These are desires that are not essential to our flourishing, but that a person cares deeply about and would be heartbroken to lose – such as a beloved family member. While some in the Christian tradition (those who Stump dubs as defenders of a ‘stern-minded attitude’) have been ‘unwilling to accord any value to the desires of the heart and, at worst, eager to extirpate the desires themselves’ (Reference Stump, Rea and Pojman2008, 342), Stump argues that love of neighbour is incompatible with this attitude, since such love requires desiring union with one’s beloved, which one cannot do if one loses one’s beloved.

While Stump has in mind cases in which God’s consequent will involves the loss of someone or something a person deeply cares about, I would argue that even in some cases where an individual’s desires of the heart are not at stake one should not attempt to align one’s will with God’s consequent will. After all, since God’s consequent will amounts to whatever actually occurs, it follows that God consequently wills wars, famines, plagues, genocide, and all manner of grave natural and moral evil. But there are at least two reasons we should not hope for such evil, even if we think it is very likely to occur, and so, to be God’s consequent will, and even if we believe that whatever God (consequently) wills must be for the greater good. First, we are unlikely to be able to see God’s reasons for allowing it – what greater good will come out of it in the end – and so hoping for God’s consequent will would amount to desiring the evil without desiring any particular good. And second, even if we have some inkling of a greater good, we ought not to desire such great evil that good may come. (We may be permitted to desire some minor evil for the sake of a great good, such as pain for the sake of healing – but at least some evils are too great for it to be permissible to desire them.) It is the prerogative of God alone (if it is even the prerogative of God) to desire and will such trade-offs.

What does this all have to do with Rea’s argument? In each version of the argument, we can distinguish one premise that is true only of God’s antecedent will, and one that is true only of God’s consequent will. Consider again the first version (Argument #1):

1. 1. The hopeful universalist hopes that universalism is true but believes that universalism might not be the will of God.

2. 2. Hope for a state of affairs involves a desire for that state of affairs to obtain.

3. 3. The hopeful universalist also desires that God’s will is done on earth and has a second-order desire to prioritise the satisfaction of this desire.

4. 4. So, the hopeful universalist must have a first-order desire (for universalism to be true) that is inconsistent with a second-order desire (to prioritise the desire for God’s will, which might involve the falsity of universalism).

And here is the second (Argument #2):

1. 1. The hopeful universalist hopes that universalism is true but believes that universalism might not be the will of God.

2. 2. Hope for a state of affairs involves a desire for that state of affairs to obtain.

3. 3. The hopeful universalist’s desires presuppose the goodness of what is desired.

4. 4. The hopeful universalist must conceive of God’s will as good.

5. 5. So then the hopeful universalist both presupposes the goodness of universalism (as the object of her desire) and does not presuppose its goodness (because she does not believe it to be God’s will).

My contention is that there is an equivocation in each version of the argument. In both versions, the first premise should be understood to mean that universalism is not the consequent will of God. For while God desires that all be saved (1 Tim. 2:3–4), allowing some to be eternally damned (or annihilated) may realise some of God’s other desires and thus be, all-things-considered, best. Just so, in Argument 2, the fourth premise should be understood to mean that God’s consequent will is all-things-considered good. But the third premise of Argument 1 should not be understood to mean that the hopeful universalist desires that God’s consequent will is done on earth; and likewise in Argument 2, the third premise should not be understood to mean that one’s desires presuppose the all-things-considered goodness of what is desired – for the reasons given above. Since God’s consequent will or what is all-things-considered best includes events that are intrinsically deeply evil, it cannot be that we are expected to desire these things. There may be some cases where God expects us to hope and pray for the all-things-considered good even if this means sacrificing some intrinsic good for ourselves or others that we care about; but there are some evils that are bad enough that we should not be expected to hope or pray for them, even if we have faith that through them God will bring about some greater good.

Now I propose that at least from the perspective of the hopeful universalist, the damnation of some fellow humans is such an evil – one that might be all-things-considered best but not one that we can be expected to hope or pray for because it is so evil, intrinsically speaking – indeed, the worst possible thing that could happen to a person, paling in comparison to any evils of this life. In response, Rea might ask the following: if the hopeful universalist believes that damnation is so evil, why would she not be a believing universalist as well? Why maintain the possibility that a perfectly good, perfectly loving God would not guarantee the salvation of all people? My answer draws a parallel with answers to the problem of evil. We might ask the theist why she would hope for no more wars, famines, plagues, or genocide, but not believe that God will prevent all such evils in the future.Footnote ² The answer is obvious: we know full well that these evils will continue to occur. But, Rea might press, we have good evidence that the occurrence of such evils is compatible with God’s perfect goodness (at least, if we take ourselves to have good evidence of the existence of God!), since we witness these evils on a daily basis – but we do not have comparable evidence that the damnation of some human beings is compatible with God’s perfect goodness. So, the situations are not parallel, and our intuition that damnation is the worst possible evil that could occur to someone should push us from being hopeful universalists to being believing universalists.

But hopeful universalists may have strong reasons to resist this move, depending on what reasons they have for refraining from embracing universalism in the first place. Suppose they take themselves to have as strong reasons to believe in the authority of the Catholic Church as they do to believe in the existence of God. Then, since the Catholic Church’s teaching is opposed to (believing) universalism, it would seem the cases of hell and this worldly evil are indeed parallel, and they should refrain from believing in universalism just as they refrain from embracing atheism. But just as it is permissible – and may even be morally required – of someone to continue hoping (against hope) and praying for an end to horrendous evil, so it is permissible and may even be morally required of a person in this epistemic position to continue hoping (against hope) and praying that God guarantees the salvation of all people.

After concluding his arguments that hopeful universalists’ hope is incompatible with their lack of belief in universalism, Rea notes that they thus face a choice: give up the hope, or start believing what they hope to be true. He then draws an analogy to the situation faced by a biblically faithful person confronting passages that seem to teach that God commanded Israel to commit genocide against the Canaanites: one should obviously believe – and with confidence – that God did not in fact command such an atrocity. And Rea has no time for appeals to ‘piety and humility’ in the face of difficult scriptural passages; he writes, ‘how can it be consistent with piety or humility to leave unchallenged the hermeneutical principles that led one to interpret scripture as teaching a doctrine that, according to one’s best (and presumably scripturally shaped) conceptions of love and goodness, is inconsistent with the perfect love and goodness of God?’ (Reference Rea2022, 273).

Rea does not flesh out the analogy, and what it means for the case of belief in universalism, but his reasoning suggests that the hopeful universalist should likewise reason from the obvious goodness of universalism together with ‘the perfect love and goodness of God’ to the conclusion that it is true. He also suspects that what is getting in the way of the hopeful universalists becoming believing universalists is itself a lack of humility or piety – because they hold ‘hermeneutical principles doggedly fixed’ no matter where these principles lead them (Reference Rea2022, 273). But I wonder if this misrepresents the hopeful universalists’ epistemic position. As suggested in the introduction to this paper, I hypothesise that their refraining from belief in universalism may stem not from biblical interpretation alone but from faithfulness to their church’s teaching. And trust in the institution of the Church – and in the Holy Spirit’s work in leading the Church to truth – does seem a sign of piety and humility. Furthermore, hopeful universalists do not seem to find it obvious that the perfect love and goodness of God conflicts with the possibility of eternal damnation for some created persons.

In the second half of his essay, Rea considers a number of objections to his argument against hopeful universalism. A consideration of one of his replies will serve to clarify some points of disagreement between us. Rea writes,

Rea illustrates this point with an example: if you purchase a lottery ticket that you know has a low chance of winning, you should believe that God will likely not bring about or allow your winning, which means God is unlikely to have an all-things-considered desire for your winning. But then your hope that you win is in conflict with your second-order desire to prioritise desiring God’s will. So, according to this logic, you should not hope that you win the lottery. Rea reasons, ‘An interesting further consequence, if this is right, is that I should not pray to win – and, generalizing the argument now, I should not pray for anything that I do not think is likely to happen. After all, it would seem to make no sense to pray for things one does not hope will occur’ (Reference Rea2022, 276). Indeed, Rea points out, ‘it looks as if my argument will also show that worshippers of God cannot rationally hope for anything they think might not happen, regardless of what they think of its likelihood’ (Reference Rea2022, 276). This would be a problematic implication, indeed. But Rea thinks it does not actually follow from his argument that one should not hope or pray for anything she thinks might not happen, for universalism is a special case, and disanalogous from cases like the lottery ticket. The key difference lies in the fact that the sort of universalism in question ‘depends solely on God’s will’:

So on the one hand, Rea reasons, universalism (on his characterisation) implies that God guarantees the salvation of all, and on the other hand, it’s implausible to suppose that God guarantees the outcome of every lottery. ‘However’, Rea notes, ‘if a worshipper of God finds herself unconditionally hoping that God will guarantee a particular outcome – as would be the case if one were hoping for God to “rig” a lottery in one’s favour… then I do think there is tension with her second-order desire to prioritize God’s will’ (Reference Rea2022, 277).

In response, we might begin by asking why hopeful universalists don’t believe that God guarantees the salvation of all. The answer might be that, like in the case of the lottery, they think God leaves the outcome ‘to chance, the operation of natural laws, human free choices, or some combination of these’. Indeed, it is a common anti-universalist line of reasoning that our eternal fate is in our hands and that it seems unlikely, or at least possible, that not all created persons will make the choices necessary for their salvation. It is a matter of debate among those who reject universalism whether God could guarantee the salvation of all people, or whether such a guarantee is incompatible with some other good that God desires (such as the free response to an offer of salvation). So one possible reply to Rea is to reject his characterisation of the hopeful universalist’s position. For those who do not believe it is possible for God to guarantee the salvation of all people, their hope might be slightly weaker: that all people are in fact saved – that they make the free choices necessary for their salvation – on parallel with the hope that one does in fact win the lottery. This does not mean that on this picture God stands idly by, waiting to see what his creatures will do; hopeful universalists might imagine God to be actively working to bring all people to salvation. But while they might hope that God is clever enough to realise this aim, they would lack belief that God can guarantee its realisation, perhaps because God lacks middle knowledge. Rea might charge that I have changed the terms of the debate, and I guess I have. But if most people who say they hope that universalism is true but do not believe it to be true simply hope that God does in fact save all people, then it seems more important to consider the rationality of this position than some hypothetical one. I am not sure about the actual epistemic position of most hopeful universalists, but my hunch is that someone focused on the question of universal salvation may hope that this outcome occurs and not care as much whether the outcome, if indeed it does occur, was guaranteed or not.

In any case, this is just one possible reply to Rea’s argument from (dis)analogy with the lottery case: perhaps the cases are analogous after all. The other possible reply is to reject the claim that the position of one who hopes that God will guarantee universal salvation while not believing that God will is similarly inappropriate to the position of one who hopes that God will guarantee that she wins the lottery while not believing that God will do so. If there is something wrong with the latter case (and I’m not sure there is), it’s not that one lacks a belief in an outcome that one hopes God will guarantee. Consider again the case of some significant evil that might occur in the near future – say, that one’s child might be swept away in a flood. One who hopes that this evil not occur may not merely be hoping that the natural course of things will pan out in one’s favour; one may be hoping for God to step in and guarantee the child’s safety. And this seems an entirely unproblematic hope, even if one lacks belief that God will do so (perhaps because one has powerful evidence that it would take a near miracle to save one’s child at this point). The difference between this case and the lottery one (if there is a difference) has to do with the goodness of the object of hope. Given some background assumptions, it might not be a particularly good state of affairs for one to win the lottery – or at least, one’s not winning the lottery is not a significant evil. But it is surely a significant evil for one’s child to die in a flood, just as it is a significant evil for a person to be eternally damned.Footnote ³

In responding to another objection to his argument – that hopeful universalism might be an instance of ‘(reasonably) resisting certain kinds of personal transformation’, Rea writes that the hopeful universalist ‘had better believe that God is perfectly good and that she will, therefore, be all-things-considered better off if she brings her preferences into line with God’s’ (280). If what I have argued in this paper is correct, then ‘God’s preferences’ here is ambiguous: it may refer to God’s all-things-considered preferences, that is, God’s consequent will, or God’s valuing of intrinsically good states of affairs, that is, God’s antecedent will. If this phrase refers to God’s antecedent will, then it is true that one will be better off bringing her own preferences into line with God’s, but she may do this while resisting believing universalism. But if the phrase refers to God’s consequent will, then it is false that one will be better off if she brings her preferences into line with God’s; indeed, she will be less herself, less human, being forced to give up the desires of her heart.