n this blog series between my friend Suan and I shall be arguing for an apparently audacious position not too different from that articulated in David Bentley Hart’s recent book That All Shall Be Saved (2019): that “if Christianity taken as a whole is indeed an entirely coherent and credible system of belief, then the universalist understanding of its message is the only one possible” (emphasis mine). The conditional formulation is significant because, like Hart, I do not primarily intend to engage in apologetics for the “Christian religion”—at least not the defensive kind of apologetics against “unbelievers’ objections to the faith,” even though I have little doubt that universalism helps remove certain intellectual and moral barriers to belief in the Christian God in particular. Instead, my goal is at once more modest and more ambitious: to argue that non-universalist understandings of Christian eschatology—wherein all of God’s creatures are not ultimately redeemed—make Christianity an ultimately irrational faith in the specific sense that the classical Christian understanding of God and its narrative of creation and redemption is called into radical question if it includes God imposing or permitting eternal misery on finite rational creatures.
With those preliminary comments regarding my scope and goal stated and clarified, let me now address Suan’s points against universalism. He presents three arguments, but, as we shall see, the first and second and second and third are interrelated: (1) J.L. Mackie’s problem of evil; (2) the so-called coopting problem; (3) classical theism is most compatible (emphasis mine) with the traditional view that not all will be saved.
First, Suan’s insistence that universalists must explain why God didn’t create Mackie’s world to begin with is fundamentally misguided for many reasons. The most obvious one is that it misunderstands and, in turn, distracts from the purpose of an argument for Christian universalism, which is not to solve the question of the traditional ‘problem of evil’ of why an omnipotent, omniscient and perfectly good God would allow any evils to occur, even if it might (as I think it does) significantly reduce the weight of what Marilyn Adams called “horrendous evils” (see below). Rather, it only intends to remove the problem of an eternal hell. Whether the notion of an eternal hell is, in fact, such a problem may seem to be debatable or even categorically denied by Thomists: for an eternal hell, as a Thomist would argue, is an instance of a moral evil that is fully self-imposed and so fully just/deserved within the logic of what allegedly ‘naturally’ follows when one chooses to reject God ad infinitum. I will question the significance of this natural/moral evils distinction when it is applied to the case of an eternal hell below; but for now it is important to isolate the specific case of moral evil that is an eternal hell and distinguish it is an independent problem from the broader ‘problem of evil’.
The justification for this can be readily seen if one reminds oneself here that, according to a Christian worldview, suffering is in some senses a ‘given’ (even if in a more primordial sense “unnatural”), because the fact that we inhabit a “fallen” world is a ‘given’ (the origin of its ‘fallenness’ will be dealt with later). However, the question of an eternal hell is not merely a question about how some finite ‘natural’ disaster such as an earthquake or ‘moral’ calamity such as the Holocaust might call into question—inductively—or show inconsistent—deductively—the existence of a good and loving God; it is a question of eschatological, and hence eternal, significance, from a human perspective and what it reveals about God’s nature. Moreover, it is one that I am convinced calls into radical question the internal coherence of the Christian tradition—whether its central hope is ultimately true or not (although I’m sure assuming it is)—for that tradition proclaims—at least in its dominant streams until early modern deviants in the wake of the Protestant Reformation like Calvinism and its seductively dualistic doctrine of praedestinatio duplex arose—not only that (1) God, being infinitely benevolent, desires for all human beings to be saved and to reach a knowledge of the truth, but also that (2) God, being infinitely free, will be finally victorious over all evil and death. As someone who is wholly convinced that if Christianity is true in any way a universalist understanding must be true too, I would claim, therefore, that (1) and (2) are in significant tension with each other on the traditional view of hell but not at all if we dispense with it. In other words, Christian universalism, like the traditional view of hell, already presumes that Christianity solves the problem of evil ultimately, but claims that as an eschatology, it, unlike the latter, has more explanatory power and scope in actually resolving the problem of evil eventually: here, God will be victorious over all evil not in his perfect supremacy over all by consigning or leaving his “enemies” in some state of eternal penal suffering, but in his being ultimately (as 1 Corinthians 15:28 proclaims concisely yet unequivocally) “all in all” by successfully ‘wooing’ all in intimate union with that which fulfills the deepest desires of our hearts (the absolute Good which, as the source and end of all relative goods, makes them seem ‘good’).
Once one clearly sees what specific problem universalism is aiming to resolve, complaining that it does not resolve or address the ‘problem of evil’ simpliciter becomes not just misguided in an innocent way but ridiculously so. The problem of an eternal hell has nothing at all to do with whether I can provide a ‘morally sufficient reason’ for God’s allowance of finite evils (a framing which not only I but Suan also would presumably reject in any case as a proponent of classical theism à la Brian Davies; see below on God not being a moral agent), but only whether the persistence of infinite evils in humans (such as it would obtain in an eternal hell) is internally consistent with (1) and (2) above: God being infinitely benevolent in desiring all to be saved and infinitely free in his eternal act of creatio ex nihilo.
It is precisely because (I agree with him that) God is not a moral agent like us—but rather the transcendent Good as such whom we can nonetheless immanently participate in and partially recognise via analogy—that Suan errs most egregiously to presume that I as a universalist affirm all six of Mackie’s propositions in his logical argument against God’s existence from evil. Here it appears that, anticipating that I will go on to argue that God will eventually eliminate all evil from his creation by redeeming all from their enslavement to sin, Suan imagines that I would affirm Mackie’s third premise—“a good being does all it can to eliminate evil.” But, in fact, I do not affirm Mackie’s third premise, nor (as might be expected at this stage) do I accept the inference that is supposed to follow from the fifth premise (“an omnipotent being can eliminate all evil”): that God, because he is also “a good being”, should eliminate all evil, immediately or at all. However, I could still argue, in denying these premises, that we have good reason to suppose that if the Christian revelation is in any way true, God will eventually eliminate all evil—not because he is a moral agent like us (albeit omnipotent etc.) and hence with negative obligations to eliminate evil if he is able to do so, but because such a conclusion would be the most consistent with what Christians have proclaimed about the nature of God classically (as Origen and Gregory of Nyssa argued with particular eloquence): that judgement and suffering (retributive or undeserved) always gives way to restoration and healing because of who God is, a God whose essence is an unrelenting, unceasing Love.
After his section on Mackie, Suan segways into his second argument by claiming that a series of universalist answers to the question of why we don’t have Mackie’s world to begin with—such as the one I just offered in the previous paragraph—can be co-opted to show (as he says in a question-begging parenthetical comment) “that the sort of world universalism pictures is incongruent with reality.” Suan’s first counter-objection (a)—that, if God is not obligated to create Mackie’s world, he could equally be not obligated to save all—fails for at least two reasons: (1) there are other (and arguably better) ways to show why God might not have created Mackie’s world to begin with apart from denying that God has obligations (see below on soul-building); (2) it illicitly conflates protology (first things) and eschatology (last things) by jumping from “we cannot fault God for permitting e.g. an isolated natural disaster or a Holocaust” to “we cannot fault God for permitting an eternal hell.” I would argue that we certainly not only can but should critique an image of God that permits an eternal hell—for an analogous reason to why we not only can but should fault an image of a parent that permits their child to throw herself into a fire that kills her out of respect for her “liberty.” If it is then objected that God appears to be a “failed parent” too in allowing horrendous evils, notice how this would be a classic case of the tu quoque (Latin for “you too”) fallacy: upon raising the specific eschatological issue of eternal hell from a parental analogy, it is fallacious to infer the falsity of my universalist conclusion by asserting that God appears to be (analogously) a “failed parent” in permitting various horrendous evils in this present fallen world.
The same issue arises when Suan makes the counter-objection (b) by asking: “why is God’s love ultimate rather than immediate?” This is a question that not only a universalist but every theist has to answer, including Suan himself. But by asking this question he’s failing to recognise that we’re not comparing possible world scenarios of how God could have created in the beginning, but rather different pictures of final creations. And, as we will see in the third section, I think it is obvious that Suan’s vision of the final creation—the institution of a (purportedly) 'perfect' hierarchy wherein the only 'good' of the eternally damned is their maintenance in existence—is irredeemably inferior to a universalist one.
But, briefly, to address counter-objection (c) to the soul-building theodicy, which Suan claims fails for three reasons. Upon scrutiny, however, it appears as if these are less three actual reasons than three red herrings that (once again) distract the reader away from the eschatological question at hand and lead him to the false conclusion that universalism is rendered problematic by imagining a possible world where God created everyone and everything perfect without any evil. If the universalist were to invoke the soul-building theodicy, however, in order to provide a merely provisional, possible yet plausible reason for why God might allow us to suffer (say, to acquire the goods of genuine rather than ready-made virtues which lead to genuine character development), then it is easy to see why Suan’s objections miss the mark. Take the first: horrendous evils aren’t required for virtue; therefore—what exactly? If the alleged issue is raised from the possibility of acquiring virtue with less “horrific” evils (presumably more than just less physical pain), one could easily retort with saying that it is metaphysically possible that God could create a world where humans acquire virtue and fall into the “natural evils” of their vice without the actualised reality of the “horrific” evil—“that one living death,” as George MacDonald put it in his great sermon “The Consuming Fire”—of eternal separation from God. (The classic distinction between “natural evils” and “moral evils” is largely useless here because human nature is a natural phenomenon.) Suan would likely readily grant this as a possibility (or maybe not?), just as I could readily grant that horrendous evils aren’t always strictly metaphysically necessary for cultivating virtues...but where would this lead us, other than to both agreeing with the claim—which he somehow turns into an objection, his second one (on which more in a moment)—that horrendous evils can allow deeper and more profound virtues?
The second alleged problem is that agreeing with the claim that horrendous evils allow deeper and more profound virtues “sets [universalists] on the trajectory to accept[ing]” the traditional view. I am going to tackle this in tandem with the argument of his third section—that classical theism is most compatible with the traditional view that not all will be saved—because they both address the metaphysics of freewill. Suan speaks in both places about a “maximal qualitative depth of virtue and vice.” Mackie’s world, he claims (and I agree), lacks this depth. So too, allegedly, does the universalist’s world. But his world doesn’t, because in it a certain number of maximally vicious individuals “offset” a certain number of maximally virtuous individuals. The obvious flaw in this reasoning—aside from its arbitrarily precise calculus—is that, far from solving the various problems of evil, it compounds various problems of evil, notably one from natural inequality (see Moti Mizrahi) in claiming that “God will [only] perfect the natures of those who exercise the virtue of love” (emphasis mine). For if the perfection of a select number of “virtuous individuals” is, apparently, “the reason why God allowed the trouble of the initial world” (emphasis mine), then one might reasonably question whether this is an acceptable "price" to pay, so to speak, for living in a world wherein the damned are virtually a means to an end of the elect. I think it certainly isn't worth it. Moreover, if there is nothing preventing those who might not choose to be perfected to choose other than a preexistent degree of ignorance of God's true nature, God's choice to redeem only "virtuous individuals" becomes arbitrary and 'elitist': why can't he also redeem vicious individuals by enlightening them with the truth about how God as the Good is the fulfilment of the deepest desires of their hearts as much as those who, because of circumstantial factors have been in a better state to choose God, have chosen him?
To conclude, let me end with a simple (albeit clichéd) example that David Bentley Hart uses in his recent book That All Shall Be Saved to illustrate why moral guilt is always qualified and can never be absolute. He presents the following dilemma regarding Adolf Hitler: either Hitler could, if he had been raised differently and exposed to different influences in his youth, have turned out differently; or he was congenitally wicked, and so from the moment of his conception was irresistibly compelled by fate to become Führer. In either case his guilt would be qualified: in the former case, he would be partly a victim of circumstance; in the latter a victim of fate. "These considerations do not excuse him, of course," Hart qualifies, "or make punishment for his evils unjust; he was himself in any event, and the self that he was certainly merited damnation" (p. 38-39). They do remind us, however, of his finitude and, in general, that the character of the worst of us is always (in part) the product of external contingencies beyond our control. And this is what makes talk of someone justly and freely damning themselves to eternal misery simply incoherent on a classical theistic image of God as the source, ground and end of all Good: there is neither true justice nor true freedom in the case of someone who has failed to reach one's God-given telos on account of their (not wholly culpable) self-delusion; there is only a perverted 'justice' and a slavery to the objectively irrational, a bondage which only God, as the patient and loving physician of our souls, can and will free them from…if he is to be a God of infinite freedom and love.