A standard claim of Abrahamic theism is that at least one person will spend eternity in Hell. ‘The problem of Hell’ is how could this be, for what act could merit an eternal punishment? This paper discusses one response, the ‘divine injury response’, whereby we deserve an eternity in Hell because our immoral actions are an infinitely grave offence against God.

The ‘The Divine Injury Response’ section sketches the response and the two premises it relies upon. The ‘Injury’ and ‘Severity’ sections then discuss the two premises. I argue that the divine injury response is more defensible than one might think. However, it faces a serious objection, the successful resolution of which revolves around whether we interpret the problem of Hell as being an evidential problem or a logical problem.

The Divine Injury Response

Punishments should be proportional to the immorality of the act being punished; if I wantonly destroy your sofa, I deserve at most a financial sanction, but if I wantonly destroy your house, I deserve prison. Prima facie, this proportionality constraint conflicts with someone being condemned to Hell forever. To be punished for eternity, one must have done something to deserve an infinite punishment, but it is inconceivable that any act could justify such treatment—if you wait twenty-six million years, even Vlad the Impaler must have served his time! This is the ‘problem of Hell’ (Adams, 1975; Aikin & Aleksander, 2014; Kershnar, 2005, 2010; Lewis, 2007; Seymour, 1998, 2000, pp. 37–94; Talbott, 2021). (The referenced sources are Christian-centric, but the problem also appears elsewhere. For instance, both Ibn Taymiyaa (in Fanāʾ al-nār) and Ibn Qayyim al-Jawziyya (in Shifā’ al- ‘alīl) propose something very similar (Hoover, 2009, p. 191, 2015, pp. 209–210).)

One response to the problem of Hell is ‘the divine injury response’. Defenders include Aquinas (in Summa Theologica Prima Secundae, q74-76, q82-83, q87), Anselm (in Cur Deus Homo I.11–15, 19–25) and Jonathan Edwards (in Original Sin §130 as well as ‘The Justice of God in the Damnation of Sinners’). The divine injury response assumes:

Injury: Every immoral act includes God as an injured party. Even though others may be injured by the act, God is always amongst those injured.

Severity: Any immoral act that includes God as an injured party merits an infinite punishment.

Imagine an agent commits some immoral act. Perhaps it is deeply serious, such as murder. Perhaps it is more trivial, such as stealing a Wispa bar from a retail chain. The act has an impact on its immediate victim (e.g. the murder victim or, say, a shareholder who receives a smidgen less money in their dividend). Those immediate victims are finite beings, so the harm that they suffer merits only a finite punishment, no matter how bad the act. But, given Injury, the action is also an offence against God. Given Severity, the agent therefore deserves an infinite punishment. Crucially, no matter the nature of the immoral act—be it murder or be it pilfering—the agent deserves an eternity in Hell. So, God would send me to Hell for stealing a Wispa bar, and justifiably so.

It is worth noting how the divine injury response gels with, as well as runs against, certain theological and philosophical considerations.

It gels well with the idea that, in order to avoid Hell, we must seek God’s forgiveness. If immoral actions only injure those around us, then it is unclear why God is able to forgive us for those actions. After all, forgiveness is correctly sought from those who have been harmed, rather than someone independent of that harm. For instance, it would be wrong for a mugger appearing in court to seek contrition by apologising to the judge rather than to the victim. Given the divine injury response, we see why God’s forgiveness is relevant: the reason you are going to Hell is because of what you did to God; thus, it is unsurprising that only God can be the one to save you and why it is God whose forgiveness is relevant.

Likewise, the divine injury response captures the idea that even the mildest of immoral acts deserve an eternity in Hell. This connects with problems regarding vagueness (Kvanvig, 1993, p. 31; Sider, 2002). For instance, it is arbitrary for someone who commits n sins to go to Hell forever, even though someone who commits n-1 commensurable sins manages to avoid that fate. Given the divine injury response, anyone who commits even the slightest sin deserves to go to Hell. Thus, problems about vagueness are avoided.

The divine injury response won’t have universal appeal, however. One issue is that it conflicts with divine impassibility and immutability. That God is injured or harmed (by suffering moral outrage or becoming sad etc.) looks to be incompatible with the divine injury response. But this is not too worrying, since divine impassibility has notable detractors (Fiddes, 1992; Hartshorne, 1984; Shields, 1992). Taliaferro says that divine impassibility is the least popular of the ‘unfavourable Divine attribute of the 1950s and 60 s’ (1989, p. 217) and Mullins (2022, p. 162) says that denying impassibility is the majority view amongst Christian theologians and philosophers. So, even given that the response is inconsistent with divine impassibility, the divine injury response should still be of interest to a wide range of people.

The divine injury response is also ill at ease with the idea that those in Hell cannot leave Hell, for if the only thing keeping a person in Hell is their failure to seek God’s forgiveness, why can’t dead sinners seek out such forgiveness in their post-mortem state? It would be sophistic to simply state that dead people cannot seek absolution—there must be some justification given for such a claim! So those who endorse the divine injury response must either: (i) build into their soteriological theory an independent reason for why dead people cannot seek God’s forgiveness; or (ii) allow that people can leave Hell. Fortunately, neither option is all that implausible. Regarding option (i), we might believe that Hell is a timeless place where one suffers ‘eternally’ insofar as one suffers ‘timelessly’, frozen in an amber of torment (Effingham, 2015, pp. 40–44). Were Hell timeless, then the dead would be unable to seek forgiveness, since seeking forgiveness is an activity/process, and activities/processes must take place within time. Regarding option (ii), the idea that people might escape Hell has been historically popular (Bernstein, 2017, pp. 147–163) and, as long as not everyone escapes Hell, there is no problem in thinking that some could escape. (Not everyone can escape, for then no-one will spend eternity in Hell and therefore there would be no problem of Hell, thus no need to resort to the divine injury response in the first place.) Whilst it might seem strange that someone could have the capacity to escape Hell and yet never choose to seek forgiveness, precisely that claim has been endorsed by numerous people (e.g. Lewis, Swinburne, and Stump (Talbott, 1990, pp. 22–23), as well as Davis (1993, p. 156) and Seymour (1998)).

Having completed my exposition of the divine injury response, I next move to considering the two premises.

Injury

This section discusses three defences of Injury. The first defence, the disobedience defence, does not work. The latter two defences are more successful.

The Disobedience Defence

‘The disobedience defence’ of Injury comes from Anselm’s Cur Deus Homo (Bk I Ch XXI). God has instructed us to do certain things, in particular to do the things that are moral. By acting immorally, we are being disobedient. When you are disobedient, it is an immoral act against the agent to whom you should be obedient too—in this case, God. Thus, Injury is true.

The disobedience defence is a bad defence. It is unreasonable to punish someone for breaking an edict that they were unaware of or that they were aware of, but justifiably thought was fictional/unreal. Since many people are ignorant of God’s edicts (e.g. isolated tribespeople) or have justification to believe that those edicts are fictional (e.g. atheists), such people would not deserve to go to Hell forever (Adams, 1975, p. 442; Kershnar, 2010, p. 127; Kvanvig, 1993, pp. 40–50). Only a slim number of theists would allow that the isolated tribespeople and the atheists of the world avoid Hell, so I set the disobedience defence aside.

The Outrage Defence

Consider the first of two defences that do not run into the same difficulties. The ‘outrage defence’ says that God suffers some degree of moral outrage every time an immoral act is committed. Since moral outrage is harmful, God is harmed by any immoral act.

To understand the defence, start by considering cases of outrage that we are familiar with. Consider a case in which an agent directly witnesses some immoral act. For instance, I once witnessed a security van being robbed by men armed with steel bats. Whilst I was obviously not the primary victim of that crime, I was nevertheless harmed by their actions, since the very act of witnessing those events generated within me a negative and harmful effect. Whilst we might attribute that harmful effect to me fearing for my own safety, I doubt that this is the case, for I don’t believe I was ever at risk. Rather, I attribute (at least some of) the harmful effect to a sense of moral outrage. The mere fact that these events were taking place was, in of itself, enough to generate negative feelings within me that I would describe as harmful. To be clear: I recognise that the ‘harm’ is much less than that suffered by the primary victims, i.e. the security guards, but I see no reason to not describe it as harmful to some degree. The outrage defence has it that God is in the same situation. God, in being omniscient, witnesses every immoral act and thus is morally outraged—and thereby harmed—in precisely the same way i.e. Injury is true.

The outrage defence avoids both objections to the disobedience defence. Firstly, you can culpably cause outrage even if you do not know that you are being witnessed. Imagine that, believing I am not being observed, I publicly urinate on a wall. In actuality, I am being watched by a bus full of local nuns, who are looking in horror as I relieve myself on the side of their Church. In that case, I am culpable for the outrage and offence caused by my actions. Indeed, when punishments are meted out, mine should subsequently be stiffer. Or imagine that I intentionally torture an animal for pleasure but accidentally stream that act of torture live to network television, where it is then witnessed by thousands. The additional harm I inflict upon the viewers merits a greater punishment. If I torture a pigeon, I might get community service or a large fine; if it is streamed on national television, it would be (justifiably!) more severe.

Similarly, it is irrelevant whether you believe those witnesses to be fictional. Simply reimagine the example cases but add that I am accompanied by a friend who is a congenital liar. As I commit the immoral act, they (truly) tell me that nuns are watching or that I have accidentally turned on the live video feed. However, because they are a congenital liar, I justifiably believe that they are trying to deceive me and so I believe that the witnesses are fictional. My epistemically justified error does not downplay the harm the victims then suffer; even if I believe the witnesses to be fictional, their harms still justify an increased punishment.

Consider two problems with the outrage defence.

First problem: Not every agent is outraged by witnessing immoral actions. That said, God may be just such an agent, failing to be outraged by immoral actions.

Solution: I hypothesise that only imperfect moral agents fail to be morally outraged by immoral actions. God, in being perfectly moral, is thus outraged by all immoral actions. For instance, a Mafioso capo may witness terrible things committed by other people and fail to suffer from any moral outrage. However, the capo’s failure to feel outrage is clearly a moral imperfection—they are (in part) a bad person for not feeling outraged.

Along similar lines, minor immoral actions like stealing a Wispa bar may fail to generate moral outrage in witnesses. I suspect that this is, again, a function of one’s being morally imperfect. In a perfect moral agent, a minor immoral action should result in moral outrage—obviously, only a little outrage, but there should be some outrage (and some harm!). The perfect moral agent will have the correct amount of outrage for every immoral action, even if it is only a whit of outrage for a comparatively minor sin.

Second problem: People often feel unjustified outrage—not everyone who is harmed by moral outrage is such that the person inducing that outrage is immoral for having caused it. For example, in Footloose (Ross, 1984), a minister is unreasonably outraged about dancing. Or imagine that someone overhears a private conversation where I reveal my intention to vote for a particular political party, resulting in them being outraged. Clearly, I am not immoral just because they have suffered such outrage. These examples in place, it is not then obvious that I am immoral for causing outrage in God. After all, it is not my fault that God is a perfect moral agent capable of witnessing all of creation; maybe any harm God suffers is just His problem!

Solution: Whether you are morally responsible or not for outrage caused by one of your actions depends upon the morality of the act in question. Activities such as dancing or the private discussion of voting intentions are both morally permissible. Since they are morally permissible, any outrage those acts cause does not justify disapprobation. Capture that with the following principle:

For any action ϕ (where ϕ causes harm due to moral outrage in witnesses) and any agent α, α’s ϕ-ing is worthy of punishment in virtue of that harm only if α’s ϕ-ing is such that, had it not been witnessed, α’s ϕ-ing would have been immoral.

If that principle were true, the dancing and private discussion would not be cases where I was morally responsible for the outrage my actions caused, but we would still all be morally responsible for the outrage we cause in God.

The Love Defence

Consider a separate defence of Injury. The love defence plays on the harms suffered by people due to their personal connection(s) with the primary victim(s) of some given immoral act. For instance, a close friend of mine was murdered. That was deeply affecting and, clearly, harmful to me. Again: I am not the primary victim of that murder, nor am I even the person who suffered the most harm due to their emotional connection to his death. The claim is only that, as a matter of actual fact, I did suffer harm as a result of that action. And that harm I suffer—namely, the intense sadness at his loss—comes about because of the fact that I love my friend. And it is widely accepted that these sorts of harms should factor into the punishments we mete out, with murder cases being a particularly telling example. In murder cases (in the USA and UK), ‘family impact statements’ are read out post-verdict but pre-sentencing, the contents of which are factored into the resulting sentence. So, we already build into our existing modern legal systems the idea that harm to loved ones merits increased punishment. This fact is evidence for there being a moral stricture whereby harming someone by harming those they love itself merits punishment.

That said, consider God. God loves us all. If some agent α commits an immoral act against someone else, then they will therefore be harming someone that God loves. God is omniscient and thus witnesses that act, suffering some measure of sadness i.e. God is harmed. Given what has just been said, α is morally responsible for the harm that God has suffered i.e. Injury is true.

Thus, we have two defences of Injury. (Note that they are not incompatible, and God could be harmed both by being outraged by witnessing an immoral act and by being harmed in virtue of knowing that someone He loves has been harmed by that immoral act.)

Severity

This section introduces defences of Severity (‘Status vs. Character Defences’ section). It is here where the divine injury response runs into difficulties (‘Mapping Virtue to Punishment’ section). In the ‘The Ineffability Rejoinder’ section, I distinguish between two different versions of the problem of Hell, showing that the divine injury response is a good response to one version but not the other.

Status vs. Character Defences

Some defences of Severity are fairly unconvincing. For instance, in Summa Theologica (Prima Secundae, q. 87, a. 4, obj 2), Aquinas argues that injuries to a prince are worse than injuries inflicted on a commoner, thus, an injury to God—who is infinitely greater than a prince—must be infinitely worse i.e. Severity is true. In this day and age, most people are republican enough to find Aquinas’s argument unconvincing (Adams, 1975, pp. 442–444). Even an updated version, swapping out princes for elected officials, doesn’t work. Admittedly, we mete out increased penalties to those who injure our Heads of State, compared to those who commit similar actions against a regular citizen. But, whilst these increased penalties are warranted, they are the result of building a deterrent element into the penalty. We punish attacks on Presidents (etc.) more harshly in order to deter unreasonable constraints on free speech, political assassinations etc. We deal more harshly with those who throw eggs at politicians because, if we didn’t—and if politicians were egged more often than they are—then liberal democracy would suffer. Attacks on God are not in the same vein. Actions that harm God cannot threaten any such calamity; it is not as if the fabric of reality may be unwoven if I make God sad or morally outraged. Thus, there is no real analogy between harms visited upon God and harms visited upon Presidents. (For more on this, see Kvanvig (1993, pp. 29–31).)

But other defences of Severity don’t run into the same problem. Consider how the moral character of a victim, rather than their political status, factors into punitive decisions. Consider the following principle:

Moral Virtue: For all agents α and β and any action ϕ (where ϕ is an immoral action that injures β) if α was the agent that ϕ’d, then the level of punishment that α merits is partially a function of the moral character of β, such that the more moral β is, the greater the punishment α merits.

Moral Virtue is intuitively true. To see why, imagine two cases. In case one, I steal £100,000 from a duplicitous and unprincipled politician who squanders his ill-gotten gains on frivolous indulgences. In that case, I (clearly!) deserve to be punished. Meanwhile, in case two, I steal £100,000 from a hard-working doctor who volunteers for charities on the weekend and invests his money into socially responsible ventures. It strikes me that our intuition is that I then deserve a greater level of punishment than in case one. Moral Virtue thus seems intuitively true. Bear in mind that the claim is not that stealing from those of poor character merits no punishment, or even deliberately weaker punishments; rather, the claim is merely that the virtuous nature of the victim may incur upon me a correspondingly greater punishment. (Consider another example: If I steal the cigarettes of a serial killer in prison, I am doing a bad thing, it’s just that if I steal the cigarettes of the man who helped me fix my burst car tyre, I am doing a worse thing.)

It’s not hard to see why Moral Virtue might be thought to lead to Severity. God, in being perfect, is infinitely good—that is, God is of infinitely great moral character. Since Moral Virtue entails that you should receive a greater punishment based upon the character of the victim, if God is of infinite character, then you might believe that any offence against God merits a punishment of an infinite degree (e.g. an eternity in Hell). However, as we shall see in the remainder of this paper, things are not so simple.

Mapping Virtue to Punishment

To see the problem with moving from Moral Virtue to Severity, start by considering an example case in which you are writing me cheques. Obviously, every time you write me a cheque, I get richer. Imagine that you write me an infinite number of cheques; it would be impractical to cash them all, but perhaps Hilbert has opened a bank as well as a hotel, staffing it with cashiers who can complete the requisite supertasks. Even given that I have cashed an infinite number of cheques, it does not follow that I am now infinitely rich. This is because the value of the cheques that you write may keep decreasing in value. If they decreased in an exponential manner, then even if I cashed an infinite number of cheques, only a finite amount of money would end up in my bank account. For example, if the first cheque was for £250, the second was for £125, the third was for £62.50 etc. then the infinite number of cheques will only net me a total of £500.

The same thinking applies to the move from Moral Virtue to Severity. It is consistent to accept both Moral Virtue and that God has an infinitely great moral character whilst nevertheless denying that an action which injures God deserves an infinite punishment. All that needs to be the case is that the level of punishment one deserves progresses in a fashion similar to the infinite chain of increases to my bank account.

Imagine we could, for some given immoral act, quantify both the moral character of the victim and the degree of punishment that the perpetrator deserves, graphing them against one another on a chart. If the relationship between the victim’s character and the merited punishment is a linear relationship, then Moral Virtue entails Severity (see Fig. 1). (Note that the line does not start at the origin point because an immoral act committed against even someone of a quite corrupt character nevertheless merits punishment.) In that case, when I perturb God by stealing a Wispa bar, I do indeed merit an eternity in Hell.

Fig. 1
figure 1

Linear relationship

Full size image

However, the linear relationship is not the only possible relationship. It might be that the punishment one merits always increases, but only ever towards a hyperbolic limit (see Fig. 2). In that case, whilst outraging God or making God sad merits me more punishment than if I had only injured a mere mortal, the punishment I deserve is still only ever finite.

Fig. 2
figure 2

Limit relationship

Full size image

Worse, it is intuitive that it is the limit relationship that holds, rather than the linear relationship. If I urinate on a Church, I may deserve a fine of, say, £25. If I do it in front of nuns, I may deserve a slightly larger fine of, say, £35. But even if it was a bus full of Saints who saw me—and even admitting that this merits a greater punishment than the nuns witnessing me—surely I don’t merit a significantly greater punishment? Imagine that the accidental observer is as sinless and pure as you want, and you should nevertheless deny that I deserve to have my hand (or worse!) chopped off as a penalty. Similarly, imagine I see an irredeemably evil serial killer on the street, recently escaped from prison, who I then gratuitously assault. That merits some level of penalty (e.g. a suspended prison sentence). If I gratuitously assault a charity worker, I merit more (e.g. a custodial sentence). But you could imagine that the charity worker was as virtuous as you like, and you would never get to the stage where I should be doused in oil and have my skin burnt off, or have animals set upon me to gnaw on my flesh, or any sort of penalty even remotely similar to those I would apparently suffer in Hell. Along the same lines: Even if I hurt an agent with infinitely great moral character, I never deserve the awful fate of an eternity in Hell merely because I stole a Wispa bar.

In summary: The character of the victim may well factor into the level of punishment one deserves, but—if the limit relationship correctly represents that factorisation—then there is a ‘cap’ on how heavily punished I should be. Thus, we cannot move from Moral Virtue to Severity, and the divine injury response appears to be in trouble.

The Ineffability Rejoinder

A rejoinder is that I have not taken seriously just how unimaginably virtuous God is. God is infinitely virtuous and of infinite character. God is far more virtuous than any nun or nurse; compared to God, all of us—serial killers and saints alike—are vice-ridden debauchees. We can arguably only trust our intuitions about how punishment increases compared to the victim’s moral character when we are considering the character of other, flawed, mortals. If only we could get a grip on God’s infinite character—which, of course, our limited minds are incapable of!—then we would see that an offence against God does deserve infinite punishment. If only we could stretch our minds enough, we would see that the level of punishment deserved shoots off to infinity when we consider beings who have infinite moral virtue. The plotting of punishment vs. victim’s character would then be as depicted in Fig. 3. Whilst we have a firm grip on how the deserved punishment heads towards a limit when we consider mortal levels of virtue, as soon as we consider extreme levels of virtue, it ramps up to infinity.

Fig. 3
figure 3

(Eventually) hyperbolic relationship

Full size image

To evaluate this rejoinder, first draw an analogy between the problem of Hell and the problem of evil. The problem of evil has two forms: the logical problem (which concludes that it is logically inconsistent for God to exist alongside evil) and the evidential problem (concluding that evil is good evidence against the existence of God) (Beebe, 2022; Trakakis, 2022). It is widely accepted that driving a stake into the logical problem is easier than the evidential; conversely, the evidential problem is not as damning to the theist as the logical.

The same division can be applied to the problem of Hell. The logical problem of Hell concludes that there is no logically consistent explanation for God eternally punishing people. The evidential problem of Hell concludes instead that, given the evidence that we have to hand, we should not believe that God punishes people for eternity.

Once we recognise that there is this similar sort of division, we should come to see that the ineffability rejoinder is a panacea only for the logical problem. The ineffability rejoinder shows that there is no contradiction in believing that punishment levels increase as per Fig. 3; thus, there is no contradiction in believing that God metes out eternal punishments; thus, the logical problem of Hell is resolved. But, whilst you do not fall afoul of any logical contradiction, the best evidence we have available to us is that punishment levels progress as per Fig. 2 and not as per Fig. 3. That is, our best evidence bears out that harming God does not merit an infinite punishment and that Severity is false. Thus, the divine injury response does not resolve the evidential problem of Hell.

This division of the problem of Hell into logical and evidential versions also deals with other concerns similar to the ineffability rejoinder. Thus far, I have not only assumed that God is passible, but further talked as if the features that God has—features of feeling outrage, or sadness, or possessing a certain moral character—are of more-or-less the same ilk as features that other mortal beings have. But you might think that, whilst impassibility is false and that God suffers, He suffers in a totally different way than we do. That is, God does not suffer as mortals do, but has Divine Suffering. Similarly, whilst God is harmed by my actions making Him sad, it is a Divine Sadness. Similarly, He has Divine Outrage, and a Divine Moral Character, rather than being simply outraged or possessing a mere moral character. These indescribable divine characteristics are, at best, ‘analogous to’ the relevant mortal characteristics.

One might think that this can shore up the divine injury response. Whilst making someone sad might only ever justify at most a finite punishment, the claim would be that, for all we know, when we make someone Divinely Sad, it is such a great burden that it justifies an infinite punishment. Similar thoughts apply to outrage. And similar thoughts apply to Moral Virtue; i.e., we could revise Moral Virtue so that if we injure anyone of Divine Moral Character, we immediately deserve an infinite punishment.

But in those cases, the same worries that apply to the ineffability rejoinder apply here. I agree that there is no logical inconsistency in believing that inflicting Divine Sadness merits an eternity in Hell, and so I agree that these sorts of moves resolve the logical problem. But these moves do not resolve the evidential problem of Hell. By stipulation, we don’t really know anything about these ‘Divine’ characteristics. We are in the dark as to what Divine Sadness is like, and so are in the dark about what rules govern Divine Sadness and the level of punishment one deserves for inflicting it. Having assumed that these matters are beyond our ken, we must likewise assume that we can never have evidence about the relative punishment one deserves when these characteristics are involved. Hence, if these moves are introduced, then the evidential problem of Hell necessarily remains, even though the logical one is resolved.

Conclusion

The divine injury response to the problem of Hell has a lot to be said in its favour. Injury is defensible (assuming divine impassibility is denied). More problematic is the claim that injuries against God always merit an eternal punishment. Whilst there is something to be said in favour of punishments being greater when you injure someone with correspondingly greater moral character, this alone does not justify believing that you will deserve an eternal punishment when you injure God. Admittedly, there is some ‘wiggle-room’ in that there is no logical inconsistency in thinking that to injure someone with infinite character merits an eternity in Hell, but our best evidence supports otherwise. Thus, the divine injury response is only a solution to the logical problem of Hell, not the evidential problem.