In my opinion, the strongest defense of hell is the one associated with Thomas Aquinas and his followers, since it relies on an Aristotelian metaphysical framework which I broadly agree with (see my posts on that topic). This argument has been presented in a modern context by Ed Feser (2016a; 2024) and James Dominic Rooney (2024a; 2024b), and a related argument is given by Wahlberg (2022). In this post, I’ll show that these Thomistic arguments fail to make hopeless damnation for any person plausible. The failure of these arguments points to the plausibility of universal salvation, given certain fundamental Christian (and specifically Thomistic) commitments about God and humanity.
Background metaphysics
In a previous post on arguments for God’s existence, I tried to show that we can know from reason alone that (1) God desires our good and (2) God himself is our ultimate good. This grounds a solid belief in the fact that he “desires all people to be saved and to come to knowledge of the truth” (1 Tim. 2:4). Thomistic infernalists agree with this point. However, they argue that while God desires, all other things being equal, the salvation of each individual person, wider considerations could allow God to permit the hopeless damnation of some people. Specifically, they claim that the post-mortem repentance of a sinning person is a metaphysical impossibility, which even God couldn’t bring about.
To understand this claim, we have to first consider the relevant metaphysics. For Aristotelians, goodness is defined as the fulfillment of a thing’s “appetite,” which is the actualization of a potential within a thing’s nature. A thing’s potentials are only actualized to the extent that it exists, so goodness is interconvertible with existence, which is also interconvertible with truth (something is only true insofar as it exists in reality). These properties (existence, truth, goodness) are referred to as transcendentals, since they are shared by all being. Whatever is not good (i.e., evil) has no existence in reality; evil is the failure of a potential to be actualized when it should be (Oderberg 2019).
Moreover, for Aristotelians, what distinguishes rational beings like humans and angels from other creatures is our intellect and will. The intellect is what abstracts universals out of their particular contexts, thus allowing us to understand abstract concepts like triangularity and dog-ness which don’t correspond to any particular, material being. In order to abstract universals from their material contexts, the intellect itself must be immaterial. The will (also called the rational appetite) is what orders our actions toward ends that we choose, according to the intellect. If the will didn’t exist, then we would still be able to know things with our intellect, but we could never use this knowledge to do anything.
Our intellects are ordered toward (universal) truth; otherwise it would forever be tied down to particular contexts. This would be contrary to the intellect’s ability to abstract universals out of particular contexts. Furthermore, if the intellect were ordered toward anything other than truth – for example, survival – then we could never know something as true, merely as it is good for survival. This view is self-refuting, since someone who claims that the intellect is ordered toward survival is making a truth-claim – that it is true that the intellect is ordered toward survival.
Likewise, the will is ordered toward (universal) goodness; otherwise it would be tied down to some particular good. In that case, the will would be unable to order the multiple goods associated with human flourishing (friendship, sensory pleasure, bodily health). Moreover, the will can’t desire something that it doesn’t perceive to be good in some way (the “guise of the good”), for then we would be acting toward something we know to be nothing at all in reality. Since the intellect and will are ordered toward (universal) truth and goodness, it follows that the rational natures of humans and angels can only be completely fulfilled by truth in itself and goodness in itself, which is God. However, no created intellect can naturally comprehend the divine nature in itself (what some Christians refer to as the “beatific vision”); this requires God’s supernatural intervention to achieve.
So far, so good. I agree with all of this metaphysical background, and in the following sections I’ll take it for granted. To sum up: being, goodness, and truth are interconvertible transcendentals; rational beings like humans and angels have an immaterial intellect and will, which are ordered toward (universal) truth and goodness, respectively; therefore rational beings can only be completely fulfilled by God, who (as pure existence) is truth-itself and goodness-itself, and this requires supernatural intervention.
Post-mortem cognition and repentance
Why do Aquinas and modern Thomistic infernalists hold that repentance after death is impossible? First, let’s consider why Aquinas thought that angels – which in his view are immaterial – could never in principle repent after falling.
It belongs to the angels’ nature to have actual knowledge of everything they can know naturally, as we by nature have actual knowledge of first principles, from which we by a process of deductive reasoning proceed to acquire knowledge of conclusions. But angels do not have such a process of reasoning, since they intuit in the principles themselves all the conclusions proper to natural knowledge of them. And so as we are permanently disposed regarding knowledge of first principles, so the angels’ intellect is permanently disposed regarding everything it knows by nature. And since the will is proportioned to the intellect, it follows that their will is also by nature irrevocable regarding what belongs to the natural order. But it is also true that they have potentiality regarding movements to supernatural things, whether by turning toward them or by turning away from them. And so they can only have the change of moving from the order of their nature to things transcending their nature by turning toward or away from them. But since everything added to something is added to it according to the mode of its nature, it follows that angels persist irrevocably in turning from or toward a supernatural good. (De Malo 16.5)
This needs more than a little clarification. First, consider how we humans know things by our intellect: we observe things in multiple particular contexts, abstract out information about universals (e.g., that 2 + 2 = 4, that grass is green), and use this knowledge to reason our way to conclusions. This mode of knowledge is essentially embodied because it relies on sensory input. If angels are immaterial, then their reasoning isn’t temporally extended like this. They immediately have whatever knowledge is natural for them to have, including perfect self-knowledge.
Since the angelic intellect knows all natural things in a single act, the angelic will also chooses its end in a single act. Once an angel chooses its ultimate end, it can’t naturally gain any more knowledge that could cause it to reconsider its choice. Moreover, having perfect self-knowledge, the angel can’t be mistaken about its desires, and must identify with all of them and integrate them all toward its chosen end (Rooney 2024a, 4–6). Therefore, if it chooses an ultimate end other than God, it will be unable to reconsider or desire anything other than what it has chosen.
But if an angel has all natural knowledge, how could it choose something other than God, if a rational being can only choose what it perceives to be good? Simply put, because it has all natural knowledge, not supernatural knowledge. The angel doesn’t have perfect knowledge of goodness itself (i.e., God), since the beatific vision requires supernatural intervention. Unless it chooses union with God, no angel possesses the beatific vision. If it chooses otherwise, then the (now fallen) angel doesn’t gain the supernatural knowledge that would be necessary for it to reconsider its choice.
Therefore, even having all natural knowledge, it’s possible for an angel to wrongly choose some derivative (spiritual) good for itself as its ultimate end in place of God. At that point its choice is ‘locked in’ (so to speak) forever. Even God couldn’t alter its choice without destroying its personal identity, since the fallen angel, having perfect self-knowledge, fully identifies with its desires which have been fully integrated toward its perverted end. This is why it is impossible, in Aquinas’ view, for a fallen angel to ever repent.
What does this have to do with humans? According to Aristotelian anthropology, when a human dies (i.e., her matter and form separate), her form naturally continues to exist immaterially, due to its immaterial operations of intellect and will. Aristotle and Aquinas refer to the form of the living human as her “soul,” although it better corresponds to the Scriptural category of “spirit” (as I argued elsewhere). After her death, the human’s spirit is relevantly like an angelic spirit: it no longer knows via sensory input and discursive (temporally extended) reasoning – since that requires a body – but it instead immediately possesses all its natural knowledge, including perfect self-knowledge.
In the moment immediately after a person’s death, therefore, her spirit perfectly knows and identifies with all of her desires, and the will integrates those desires to the ultimate end that she chose before she died. Like the fallen angel, her choice becomes ‘locked in’ to the extent that even God couldn’t change it. The basic orientation of the will, either toward or away from God, is fixed at death. For some people, there will be a mixture of good and evil desires (since in our bodily state, without perfect self-knowledge, our desires may conflict). In that case, some desires will have to be modified, but in a state of perfect self-knowledge one’s desires will ultimately be made consistent.
For those who choose an ultimate end other than God, they will forever suffer evil (in the Aristotelian sense), since their rational nature will forever fail to fulfill its ultimate end. In addition to this terrible privation, they will suffer positive punishment. Consider that a failure to fulfill our nature, by acting in a disordered way, often results in some kind of unpleasantness: sickness results from overeating, a guilty conscience results from violating a moral duty, and so on. This usually serves the good purpose of directing us away from evils. However, for those who have fixed their will against goodness, they are also fixed in a state of unpleasantness that corresponds to the faulty end they have chosen for themselves. Since it’s God who orders things according to their natures, this punishment is “from God” in an ultimate sense (Aquinas, Summa contra Gentiles III.140; Feser 2016b).
This line of argumentation is used by Aquinas and modern Thomistic infernalists to prove the reality of hell and purgatory (for those whose desires are ‘mixed’ between good and evil at death). If people’s wills truly are fixed at death, then universalism might still be true – it’s theoretically possible that God leads each person to repentance just before they die. But this certainly seems unlikely, and ruling out the possibility of post-mortem repentance would strike a massive blow against Christian universalism.
Why the argument fails
First of all, I disagree with Aquinas that angels are wholly immaterial beings. This doesn’t have much bearing on his argument for hell, since we could simply consider what it would be like for immaterial creatures without conceding that any such beings exist. In fact, I’m perfectly willing to concede that if wholly immaterial creatures did exist, it would be impossible for them to repent after falling – I see no issue with this part of the argument. But because I don’t think angels are wholly immaterial, I disagree that it’s impossible in principle for them to repent (Case 2021). With Origen and Gregory of Nyssa, I hold that all created rational beings, up to and including the satan, can and ultimately will repent.
I also don’t necessarily agree with Aquinas and modern Thomists that the human spirit is conscious or active apart from the human body. It seems at least as likely to me that my spirit, although it continues to exist after my death, will be no more active than it is during deep sleep or syncope. (That would also account for the metaphor of death as “sleep” throughout the Scriptures.) In that case, the argument for hell fails because a person’s spirit after death won’t be in a state to set its final end, integrate its desires, and thereby become ‘stuck’ in that orientation. However, I’m willing to grant for the sake of argument that the human spirit is relevantly active after death, since I think the argument still fails.
The possibility of new experiences
First, as Rooney (2024a, 7) points out,
all that is fundamentally required for someone to persist in sin forever is simply that they make up their mind about what they want (where what they want is something incompatible with love of God) and never to encounter any reason sufficient to make them reconsider that decision, thereby prompting a potential new volition.
Both conditions are necessary for someone to persist in sin forever. While the first condition is fulfilled by the perfect self-knowledge of the post-mortem spirit, the second condition is much harder to fulfill. The very logic of the infernalist position requires that people undergo new (if unpleasant) experiences after death, which can form the basis of new knowledge and allow people to reconsider their choice. It seems like it would only be possible for a mind to have all natural knowledge all at once if it were completely atemporal and impassible, such that it could have no new experiences.
For this reason, the Thomist view of the fixity of the will after death seems like a merely Aristotelian holdover which doesn’t fit well into Christian theology. Aristotle believed that the mind, “when it is separated, is just as it is, and this alone is immortal and eternal, though we do not remember, because this is impassible” (De anima III.5). In other words, the mind after death is totally impassible and has no memories or new experiences. Aquinas and modern Thomists hold this idea in tension with Christian theology, since they believe that the human will is fixed after death and that people have new post-mortem experiences. But these views look like they’re mutually exclusive.
Moreover, this merely Aristotelian view doesn’t square well with the Christian doctrine of bodily resurrection. Even if the spirit is immutable when it’s disembodied, it will be embodied again in the future, and its re-embodiment will bring new perceptual experiences that should restore its mutability. Ed Feser (2016) argues against this conclusion:
But might not the resurrection of the body restore the possibility of a course correction? Aquinas answers in the negative [De veritate q. 24, art. 11]. The nature of the resurrection body is necessarily tailored to the nature of the soul to which it is conjoined, and that soul is now locked on to whatever end it opted for upon death. The soul prior to death was capable of change in its basic orientation only because it came into existence with its body and thus never had a chance to “set,” as it were. One it does “set,” nothing can alter its orientation again.
Feser’s (and Aquinas’) argument here is sound, based on his view that human spirits after death (like immaterial angels) gain all natural knowledge all at once. If that’s the case, then being embodied wouldn’t change a thing, since the person would already have all the knowledge they could possibly use to make a choice. But that’s precisely what I’m arguing against here. The human experience is essentially temporal and open to new experiences, so a human, unlike a wholly immaterial being, couldn’t possess all natural knowledge all at once. That’s just not within the realm of logical possibility.
Goodness and evil
This brings us to an even deeper problem with the Thomistic argument for hell: it implies that good and evil are on an ontological par. The human will can rest in good or evil eternally, depending on which one it chooses. Feser (2018) actually says this explicitly:
If the wills of the damned could change after death, then so too could the wills of the saved. Thus, they wouldn’t truly be saved any more than the former would truly be damned. They would forever be in danger of falling again into evil and facing punishment for doing so. The travails and instability of this life would never end. Hence, no hell, no heaven either.
This argument assumes an ontological equivalence between good and evil. If the human will can’t rest eternally in evil, then it can’t rest eternally in goodness either.
But this is totally contrary to the fact that evil is nothing in itself, being merely a privation of goodness – which Thomists agree with. This logically leads to the classical doctrine of “the guise of the good,” that people only desire what they perceive to be good in some respect. No one could choose that which they know to be nothing at all in reality. Thomists like Feser agree with this; in fact, they’re some of the strongest defenders of this view in the modern day (e.g., Feser 2014)!
To be sure, this doesn’t mean that people can’t be culpable for sinning. A person can do something that they perceive to be good in some respect, even if they know that it’s evil in other respects. Someone can culpably sin by voluntarily failing to consider the evil in an action, and only considering the good they perceive therein (Rooney 2024a, 2–3; citing Aquinas, ST I-II, q. 73). However, no one could be perfectly culpable in the way required for eternal damnation – that is, having perfect knowledge of the good and choosing something else as one’s end – it’s simply impossible. Because of the ontological imbalance between good and evil, someone who (culpably or not) chooses evil could always encounter or consider new knowledge that would lead them to repent.
Rooney (2024a, 8) argues against this conclusion on the basis that the only “decisive” reason that could bring a fallen spirit to repent is perfect knowledge of Goodness itself, the beatific vision, which “constitutes salvation and is not a necessary condition for salvation to occur.” I actually agree with this. However, my argument for universalism doesn’t rely on anyone encountering a decisive reason to repent. As long as there’s a non-zero chance of repentance, over a potentially infinite amount of time, the chance that any individual won’t repent will become infinitesimally small (Reitan 2022).
The “Autonomy Defense” and hell
Another Thomistic defense of hell has been offered by Mats Wahlberg (2022) in response to David Bentley Hart (2019). Wahlberg and Hart rightly reject the “free will defense” of hell (popularized by C. S. Lewis). According to the “free will defense,” in order for God to create a world with creaturely free will, he had to allow for the real possibility – outside of his control – that some people will reject God and choose evil forever. This defense is hugely problematic for a few reasons:
1. By casting creaturely free will as the freedom to choose evil, this defense implies that God (and the saints), who can’t choose evil, are radically unfree. But this is clearly wrong; God isn’t unfree, but the most free being, since nothing apart from him can constrain his action. Thus, freedom doesn’t imply the possibility to choose evil.
2. It’s incompatible with Jesus’ own statement that “the truth will make you free... everyone who does sin is a slave of sin” (John 8:31-35). This means that freedom is precisely the freedom to choose good, and those who choose evil are thereby unfree.
3. It ignores the metaphysics of good and evil, and the “guise of the good.” Rational beings can only choose what we perceive to be good in some way; moral evil is a result of the failure to (culpably or not) know and/or consider certain facts, and this comes from our finitude. This supports Jesus’ claim that those who do evil are unfree, and undermines the free will defense.
4. According to classical theism, God is the primary mover of all change in the world, and concurrently actualizes the being of everything that exists. God’s concurrent causality is not only compatible with freedom, but is necessary for anyone to do anything freely (Koons 2002; Grant 2010). Thus, there’s nothing contradictory with the idea of God causing someone to freely choose him, and classical theism is incompatible with the free will defense (De La Noval 2024).
5. Even if the free will defense succeeded in showing how someone’s choice to reject God could possibly be forever un-reversed, it doesn’t show that anyone’s choice would be irreversible. That would negate the concern for freedom that underlies the entire defense. But since the irreversibility of the choice is precisely what’s being debated, the free will defense is a red herring.
For these reasons, no Thomist or classical theist (or, for that matter, any Christian) should find the “free will defense” very convincing.
However, Wahlberg provides an alternative defense of hell – which he calls the “autonomy defense” – that seeks to sidestep these issues. First of all, this defense relies on the Aristotelian-Thomistic concern for preserving creaturely causality. God is the primary cause of all change, but created beings act as real (if instrumental) causes and have causal powers according to their natures. If creatures aren’t real causes, then this is occasionalism – the view that only God is a real cause – which leads to pantheism (that only God is real). Thus, there must be some measure of creaturely autonomy in order for us to say that there is a creation at all.
Due to our finiteness, every rational creature has the potential to fail to know and/or consider some moral facts, and as a result desire evil. The possibility of sin follows from God’s decision to create rational beings that aren’t himself. “Not even God could create rational creatures that are infallible or indefectible by nature” (Wahlberg 2022, 54). God could supernaturally intervene to remove this possibility of sin in two ways: by providing every rational being with the beatific vision from the start, or by actively and directly frustrating people’s choices to sin. Both of these options, however, deny creaturely autonomy. In order to avoid occasionalism, rational creatures must be able to sin. Otherwise, God hasn’t really created a new being at all; he’s simply play-acting with himself.
So far, I agree with this argument. I think the concern for creaturely autonomy provides a very plausible theodicy of moral evil which avoids the pitfalls of the free will defense. But Wahlberg goes further: he argues that friendship with God, like friendship with other humans, relies on the autonomy to be or not to be friends. If my friendship with another person is a sheer metaphysical necessity, then – Wahlberg claims – it is “gravely deficient” because it lacks the autonomous self-giving that characterizes friendship. Thus, if (as David Bentley Hart argues) every person will necessarily choose friendship with God, it’s really no friendship at all. There must be the real metaphysical possibility of eternal rejection of God, although this leaves open the possibility of hopeful universalism (Wahlberg 2022, 55–60).
Why I disagree
There are a couple reasons I disagree with this argument. First, I’m not sure I agree that friendship, even with other humans, relies on the autonomy to be or not to be friends. Suppose that there are two people who are just so compatible that in every possible world where they meet, they’re best friends. Would this make their self-gift to each other any less real and valuable? My intuition says no. Moreover, consider that on some accounts of social trinitarianism, the relationship between divine persons is precisely a metaphysically necessary interpersonal friendship. There are many valid critiques of social trinitarianism, but would anyone say that the divine persons aren’t really friends on such an account?
Second, even if it were the case that friendship with humans must be autonomous, it doesn’t follow that friendship with God could be autonomous. Wahlberg (2022, 62) responds to this critique:
we all agree that self-determination is a value-enhancing or good-making property of human friendships. It would therefore be very odd if our friendship with God lacked this characteristic. While friendship with God is certainly different from intra-human friendships, it cannot be different in the sense of lacking some fundamental property (compatible with God’s nature) that makes human friendships more valuable than they would be without it.
However, our relationship with God is so different from our relationship with humans that it could not be autonomous. God, unlike other humans, isn’t a being among beings, the kind of person who we might or might not love. God is the ultimate efficient and final cause of everything that exists, the one by whom and for whom are all things (Rom. 11:36). Simply because we exist, we’re already in a relationship with God, whether we like it or not. As David Bentley Hart (2019, 183) points out, “It is hard to exaggerate how large a metaphysical solecism it is to think of God... as an option that can be chosen out of a larger field of options”.
Within the “autonomy defense” is hidden the same mistaken assumption we saw earlier: that good and evil are on an ontological par. God is Goodness itself, and to say that someone might ultimately choose God or something else is to say that they might ultimately choose good or not. But this is a mistaken sort of voluntarism that assumes the human will might just latch onto anything whatsoever as its ultimate end. The human will is naturally oriented toward the good, and chooses evil only as a result of some defect (which might be culpable or not). Wahlberg (2022, 67) recognizes this, but simply holds it in “paradoxical” (contradictory??) tension with his autonomy defense.
Conclusion
Both of the Thomistic defenses of hell that I looked at here are problematic in their own way, but they share the fundamental error of implicitly assuming that goodness and evil exist on an ontological par. If evil has no real existence in itself and is merely a privation of the good, as Thomists believe, then the idea of a human’s will being ‘fixed’ upon evil without even the possibility of repentance is a metaphysical impossibility. The chance of anyone eternally choosing evil without ever reconsidering is incredibly low, and even less credible when we consider that God himself desires all people to ultimately repent and be saved.
To be sure, it’s logically possible that someone might choose evil forever, just not that they can be left without even the possibility of repenting. The question of whether anyone will choose evil forever is one that likely can’t be answered by philosophy alone; for this we have to turn to revelation. I believe that, whether or not an eternal hell is possible, revelation tells us that this world is not one in which any rational being will abide in evil forever. God will defeat sin and death and reconcile with every person (1 Cor. 15:20-28).
No comments:
Post a Comment