Out of respect for their freedom, God gives the damned what they want: life forever separated from God. Or so goes one defense of the moral justification of eternal conscious torment. On this view (call it a “free will theodicy of hell”), two states of affairs obtain: i) God allows some persons freely to reject him and the promise of life together with him, and ii) said persons genuinely desire life apart from God. But is (ii) coherent? What could it possibly mean to say that someone genuinely wants to be eternally separated from God, the ultimate satisfaction of humankind’s deepest desires?¹ In this paper I will argue that (ii) is coherent.

First, I will discuss two ways to motivate the claim that (ii) is incoherent. Neither motivation succeeds to show (ii) is incoherent, I argue, because the arguments are unpersuasive. Next, I will discuss two popular attempts to show (ii) is coherent, one from C. S. Lewis and the other from Michael Murray. While both begin on promising notes, both are incomplete attempts to show (ii)’s coherence. Thereafter I introduce Charles Williams’s romantic theology: it completes, I think, what is lacking in Lewis’s and Murray’s argumentation by appeal to what I will call ”progressive affective impotence” (PAI). PAI renders at least two plausible interpretations of (ii). Therefore, (ii) is coherent. To close, I discuss two objections.

But first, some prefatory definitions. By ”free will theodicy of hell”, I mean any attempt to give a plausible reason why God might allow some to be separated from him in hell, where the reason concerns some person’s free rejection of God’s efficacious offer of salvation. In a sentence, anyone in hell wants to be there (Davis 2016, p. 96; Davis 1990). The view of free will presupposed in this paper is libertarian free will, as this is the view favored in free will theodicy of hell debates. Finally, I presume hell is severance from God’s salvific grace.² More precisely, “hell is a separation from God as the source of true love, joy, peace, and light” (Davis, 1990, p. 178). According to Yandell (1992), “To be in hell is to have no positive contact with God, to not be an object of efficacious divine compassion or a beneficiary of effective divine mercy” (pp. 79-80). Being in hell cuts one off from God and his saving graces.³ Definitions in play, I can proceed.

I ask and answer on behalf of the skeptic whether (ii) is coherent. What follows are two arguments to show it is not. I argue that the arguments are unpersuasive.

First, (ii) is incoherent because it entails a contradiction. It describes a state of affairs in which rational agents postmortem both can and cannot rationally reject God for all eternity, because no rational person knowingly (i.e., rationally) errs and to reject God forever is to err. Ward (2007) argues that not all will choose hell, because both “[i]t is irrational to choose a course of life which it is better not to live” and “it is supremely rational to choose a life which is overwhelmingly worthwhile, since it offers endless bliss” (p. 34). Being rational creatures by nature, one will eventually come around: “a creature can fight against its true nature only for so long, but is likely to come to the discovery of [their being made for relationship with God] if given time and experience enough” (p. 34).⁴ Because no rational agent knowingly errs and to reject God for all eternity is to err, no rational agent can by dint of her nature desire life apart from God. Therefore, (ii) is incoherent.

Ward’s argumentation fails to show (ii) is incoherent. In a sense to be elaborated upon below, what rational persons would or would not choose may be irrelevant to the coherence of (ii). Possibly, some agent’s rationality becomes sufficiently disordered, distorted, atrophied,⁵ disintegrated that, postmortem, she becomes capable only of irrational decisions and desires. Moreover, even if some degree of rationality remains in an agent postmortem sufficient for her to retain her status as a rational agent, it remains possible that she, as a rational agent, makes an irrational decision against her better judgment. But the argument for this will have to wait until Section 3, after Charles Williams has been introduced.

Second, (ii) is incoherent, it can be argued, because it is inconsistent with God’s nature. Part of the motivation for such an argument can be found in the work of Thomas Talbott. God’s love overwhelms, compels, and reorients one’s natural desires such that they cannot help but will to be loved by God and desire to live eternally with him. Following Talbott (1990a), just as it would be unloving for a parent to let their child willingly be crushed by oncoming traffic, it would be unloving for God to let his children (including sinners) willingly separate themselves from him forever. What is more, if such separation were to occur, God would eventually see to it that it not be eternal: his love would ensure their redemption eventually (Talbott, 1990b, p. 232).

Talbott has also argued that proponents of a free will theodicy face a dilemma. What follows is a semi-formalized reconstruction of the dilemma (Talbott, 2022). 1)

Either S knows full well God’s nature, the promised benefits of eternal life with God, and the consequences of life apart from God; or S does not know.⁶

Neither Talbott’s dilemma nor his parent-child analogy succeed in showing the free will theodicist claim is incoherent. First, the dilemma is invalid: 5 does not follow from 1 through 4. All that follows is: whether S knows or does not know, S would not make a rational decision. However, that S would not make a rational decision either way is not sufficient to show that (ii) is incoherent, unless it is also incoherent that rational agents (assuming S is one) ever make irrational choices. Second, (3) and (4) assume divine judgment of the damned includes, not just whatever evidences and experiences were accessible to one while one was alive, but also evidence and experience postmortem – and this latter evidence overrides whatever former evidence one had.⁷ Nevertheless, all that follows from (3) and (4) is just that S is in no epistemic position to reject the true God, not that S is and will never be in any position to reject God and life with him on the basis of evidence and experience received and given postmortem or no. The latter claim, while stronger and would lend some indirect support to it, still does not justify the incoherence charge, for it remains possible that sinners experience God negatively postmortem. What this might mean will be discussed in the next two sections. Lastly, at best, Talbott shows that it is inconsistent to affirm both free and willing rejection of God postmortem and what he presumes a loving God would do for the damned (i.e., whatever follows from the sense of ”God”). The set is inconsistent, but that is not sufficient to show (ii) incoherent – maybe the sense of ”God” is the problem.

At any rate, neither Ward’s nor Talbott’s strategy succeeds in showing (ii) incoherent. What about reasons to think it coherent?

In this section I ask answer on behalf of the free will theodicist of hell whether (ii) is coherent. What follows are two arguments to show it is. I argue that the arguments are initially plausible, yet incomplete in some important respects.

First, (ii) is coherent because it’s imaginable. One can imagine some token individual with sufficiently defective moral sensibilities, dispositions, beliefs, and so forth; from here, one can extract from this image a type of person who has these defects, and then imagine oneself as an instance of this type. In part, this is C. S. Lewis’s argument in The Problem of Pain.⁸ Lewis (1940) invites readers to imagine “a jolly, ruddy-cheeked man, without a care in the world” and “who has risen to wealth or power by a continued course of treachery and cruelty, by exploiting for purely selfish ends the noble motions of his victims” (p. 77). The man funds his lascivious, extravagant lifestyle by these exploitations, “without a care in the world, unshakably confident to the very end… that his way of life is utterly successful, satisfactory, unassailable” (p. 77). No one, argues Lewis, should desire such a man be saved as he is, because such a desire expects God to forgive him; but “forgiveness needs to be accepted as well as offered if it is to be complete: and a man who admits no guilt can accept no forgiveness” (p. 78). Indeed, it is incoherent to suggest God could forgive this type of person: one who becomes like the jolly man has become self-absorbed, egoistic, and will be judged accordingly. “He has his wish,” Lewis (1940) writes, “to live wholly in the self and to make the best of what he finds there. And what he finds there is Hell” (p. 78). One should keep in mind, however, that the jolly man is a type; and any person may be its token. Hence “[i]n all discussions of Hell, we should keep steadily before our eyes the possible damnation…of ourselves” (p. 82). So, (ii) is coherent because it is imaginable: one can imagine another or oneself as a token of a type of person who refuses divine forgiveness.⁹

Lewis’s jolly man offers a poignant illustration of someone who satisfies some necessary conditions for rejecting life with God; however, Lewis’s overall argumentation fails to show that (ii) is indeed coherent. He merely asserts a (ii)-like claim, without explaining how it is possible. How does one come to be like the jolly man? Unless one can answer this question, it remains to be seen how it is coherent to claim that persons like the jolly man genuinely desire life apart from God. In this respect, Lewis’s (or this Lewis-like) defense of (ii) is incomplete.¹⁰

Second, (ii) is coherent because it describes a natural consequence of a suitably developed character. In his discussion of models of damnation, Murray (1999) offers a natural consequence model, according to which willing rejection of God is a natural consequence of having made choices which form one’s character so to will. “[W]e are habit formers,” Murray (1999) writes, “[O]ur behaviors can influence future behaviors by disposing us to desire or think in certain ways” (p. 297, italics in original). Dispositions eventually solidify into habits, and thus one becomes “set in their ways.” Once “set in their ways,” it is possible to imagine an upper limit at which saints and sinners become “maximally set in their ways. That is, they are disposed to act as lovers of God or lovers of self without fail” (Murray, 1999, p. 298, italics in original). According to Murray’s model, anyone may “become utterly self-absorbed and self-centered lovers of self, and thereby…merit hell as a natural consequence” (pp. 298-299). The natural consequence of egoistic character is an afterlife freely elected in which one may persist in their egoism.

Murray’s model adds a necessary condition for the coherence of (ii) lacking in Lewis’s imaginative illustration, yet falls short. One does need to be of a certain character for it to be possible to make the requisite, self-serving choices; but it remains unclear a) what it is about having such a character that makes such choices possible (and others impossible), and b) what relation these natural consequences have to acquiring said character. It is a necessary but not a sufficient condition for the coherence of (ii) that one is of such a character; it is also necessary to show how it is possible to become such a character.

Though Lewis’s and Murray’s respective models both identify necessary conditions for the possibility of choosing or desiring eternal life without God, their conjunction falls short of a sufficiently illuminating account of how it is possible to become someone who rejects life with God. Enter Charles Williams.

In this section I introduce to the present dialectic Charles Williams’s romantic theology. In particular, I articulate a concept ingredient to Williams’s romantic theology: progressive affective impotence. Progressive affective impotence supplies what is lacking in Lewis’s and Murray’s attempts to show (ii) coherent; namely, that some persons genuinely desire life apart from God.

What is progressive affective impotence (hereafter PAI)? PAI describes a stretch of time wherein some subject S lives in such a way that her capacities for empathic desire and rational decision atrophy. PAI can occur in at least two ways, summarized by PAIa and PAIb.

PAIa) As S’s faculties weaken, her empathic desires and rationality disagree until they cease to inform, to collaborate with, or to control one another at all. Faculties or powers effectively divorced,¹¹ S may utter genuinely that she wishes for eternity apart from God, because she has become powerless to desire otherwise: she cannot want any longer what she was able to want; namely, eternity with God.

PAIb) As S chooses to live in a certain way, negative emotions like hatred or enmity with respect to other persons, including God, begin to develop; especially when said persons extend charities like grace or forgiveness. Rather than weaken, S’s faculties heighten in response to perceived threats to her perceived self-image and self-interest.

Distinction between PAIa and PAIb should not suggest that they are mutually exclusive. As will be seen below, sometimes the two co-occur and mutually reinforce each other. The distinction merely describes two different ways one becomes progressively affectively impotent. What follows is a brief discussion of PAIa and PAIb in Charles Williams’s romantic theology in service of a defense of the coherence of (ii).

Williams demonstrates the coherence of (ii) in the life of Lawrence Wentworth, a military historian from his novel Descent into Hell. Wentworth falls for the darling yet unrequiting Adela Hunt. At various junctures, Wentworth observes Adela’s deferential subservience to others, a source of consternation for him. Over the course of the story, Wentworth becomes increasingly conscious of his festering obsession with Adela, faintly aware by dint of “[a] remnant of intelligence” “that this was the road of mania, and self-indulgence leading to mania” (Williams, 1999, p. 50). By Wentworth’s consent, his obsession ferments into possession.¹² Eventually, Wentworth discovers his rival, another historian, has been knighted and, for a brief moment, is presented with a choice, a fork in the road. He “could share [in another’s] pleasure,” but “[w]ith a perfectly clear, if instantaneous, knowledge of what he did, he rejected joy instead. He instantaneously preferred anger, and at once it came; he invoked envy, and it obliged him.” Williams elaborates as follows:

He knew that his rival had not only succeeded, but succeeded at his own expense; what chance was there of another historical knighthood for years? Till that moment he had never thought of such a thing. The possibility had been created and withdrawn simultaneously, leaving the present fact to mock him. The other possibility – of joy in that present fact – receded as fast. He had determined, then and for ever, for ever, for ever, that he would hate the fact, and therefore facts (pp. 80-81)

With hatred of fact came preference for deception, as Adela’s doppelgänger materializes before Wentworth. In every way she is “like [Adela], only better,” the phantom Eve to his perverted Adam (p. 82). Deliberate concentration of disordered faculties, facilitated by preference for illusion and deception, revokes possibilities for Wentworth’s return, his (attempts to) enjoy another’s joy. He wanted hell, because hell means Adela and reality obey him: “No people but his, no loves but his” (p. 88). Wentworth’s descent into hell culminates in his final stages of self-awareness, where word becomes sound, faces blur, and world becomes shapeless terror. Wentworth can no longer perform basic bodily functions; his faculties have weakened to the point of complacence, passivity, impotence. Once more, Williams (1999) elaborates in a most memorable passage:

He was sitting at the end, looking up an avenue of nothingness, and the little flames licked his soul, but they did not now come from without, for they were the power, and the only power, his dead past had on him; the life, and the only life, of his soul. There was, at the end of the grand avenue, a bobbing shape of black and white that hovered there and closed it. As he saw it there came on him a suspense; he waited for something to happen. The silence lasted; nothing happened. In that pause expectancy faded. Presently the shape went out and he was drawn, steadily, everlastingly, inward and down through the bottomless circles of the void (pp. 221-222)

Wentworth is a type. He represents the one who genuinely desires life apart from God and neighbor, and succeeds. Throughout his descent, Wentworth’s empathic desires and rationality disagree until they cease to inform, to collaborate with, or to control one another at all, as evidenced by Adela’s doppelgänger and his final moments of self-awareness. Wentworth becomes powerless to desire otherwise: he cannot want any longer what he was able to want; he has become too passive, too impotent affectively to attempt escape (notice that he merely expects something to happen, nothing specific), a compliant and complacent automaton. PAI yielded self-insulation, and thus Wentworth becomes unable to recognize once-familiar faces, unable to discern sound from Speech, unable to become self-aware.¹³ He becomes metaphysically what Lewis’s jolly man is socially. In these respects, Wentworth exemplifies PAIa.

Some may dismiss Wentworth’s case as extravagant, even if metaphysically haunting. No one, it might be objected, could actually intend their own self-destruction. However, Midgley (2001) has shown that one may indeed intend one’s own self-destruction as Wentworth does, “through perversion, recombination, and narrowing of natural desires” (p. 205). Midgley (2001) describes perversion in similar fashion to PAIa: “Their perversion…consists in retaining and cherishing them as obsessions, which become partially autonomous. These feed on the rest of the character, which atrophies, so that the individual disintegrates, though his detached desires retain their force” (p. 205). In this respect, “Self-destruction is thus a secondary, but seemingly inevitable, consequence of indulged resentment” (p. 205). Midgley adumbrates in abstract psychological terms what Wentworth experienced in metaphysical terms. In essence, PAIa conduces to faculty distortion, disagreement, and eventual separation by dint of obsession. Midgley confirms: Wentworth is a type.

So much for PAIa – what of PAIb? Recall that PAIb describes the following: As S chooses to live in a certain way, negative emotions like hatred or enmity with respect to other persons, including God, begin to develop; especially when said persons extend charities like grace or forgiveness. Rather than weaken, S’s faculties or powers heighten in response to perceived threats to her perceived self-image and self-interest.¹⁴ After her comments about self-destruction, Midgley (2001) mentions in passing the Fall. Williams elaborates, with special attention to a peculiar effect of the Fall: knowledge of good as evil. Knowledge of good as evil explains in part how one may come to despise or hate willing benefactors, like God or another charitable neighbor.

In He Came Down from Heaven, Williams (2005) interprets the Fall to entail, not just knowing more propositions or having new experiences, but a new way or mode of knowing, a new knowing as.¹⁵ Freedom of will, Williams insists, requires the possibility of alternatives, at some time t choosing either A or not-A. Since, prior to the Fall, humanity had no alternative to know but the good, humanity could only know good in a different way, as antagonism (Williams, 2005, p. 21). Relations altered under this new mode, including relation to God. When God confronts Adam and Eve, their shame compounds. “Unfortunately,” Williams writes, “the interrogation merely exhibits them as they are; a severe actuality is before them, and they dislike it. They know evil; that is, they know the good of fact as repugnant to them. They are forced into it” (p. 22). Knowledge of good as evil entails change in attitude toward the good. Adam and Eve, though presented with the fact of their nakedness and the fact of God’s dwelling with them in the garden, cannot stomach the facts. “The good of fact” disgusts them, embarrasses them, diminishes them. In knowing good and evil, Adam and Eve know good as an enemy.¹⁶

These two aspects of Williams’s romantic theology (progressive self-insulation and knowledge of good as evil) define at least to ways in which someone may become the sort of person who prefers damnation to eternal life with God. Incited by monomania, progressive self-insulation facilitates egoism or self-centeredness and self-obsession, narrowing the scope of moral decision-making to only those which concern oneself and no one else, per PAIa. Wentworth’s descent into hell corroborates this. As Swinburne (1989) writes, “The man who has blinded himself to the goodness of things is no longer an agent, one who chooses what to do in the light of beliefs about its worth. […] The man has ‘lost his soul’. Although there is no maximum to human goodness there is a minimum; and this is it” (p. 177; see also von Hildebrand, 2007, pp. 59-63). Williams also emphasizes the import of knowledge of good as evil. Knowledge of good as evil describes a perceptual shift which governs decision-making. Something objectively good is automatically perceived as a threat to the realization of one’s goals, self-image, etc., per PAIb. Weil (1997) summarizes the matter nicely: “The sin against the Spirit consists of knowing a thing to be good and hating it because it is good” (p. 125). Both aspects of Williams’s romantic theology, by no means mutually exclusive and often collaborative, exemplify PAI. PAI occurs as one persists in, and eventually insists upon, living solely for oneself unto their own self-destruction.¹⁷

PAI provides the context in which it becomes coherent to say some people genuinely desire life apart from God. How? Progressive self-insulation describes the process which any person, oneself included, may undergo as a consequence of their choices (a la Lewis): they reap the moral and spiritual fruits of their moral choices, however destructive, as a natural consequence (a la Murray). Also, knowledge of good as evil introduces into the dialectical context the logical (and psychological) possibility that one may prefer to know good as antagonistic to them.¹⁸ Given such a preference, it becomes possible to say that, postmortem, divine confrontation is a negative experience for someone with said preference. Proximity to God engenders fear and resentment of God. God has become their enemy, and no one in their situation wants to live forever with a moral, spiritual, metaphysical Threat. Hence, (ii) is coherent in the context of Williams’s romantic theology.

Having argued that Williams’s romantic theology supplies what is lacking in attempts to show (ii) coherent, I move to address two objections.

Objection 1: The suggested additions to Lewis and Murray are unpersuasive because they beg the question against the universalist. If someone really can become the sort of person who prefers damnation to eternal life with God via PAI in either form (or both), then it would be irrational for anyone to hope that all be saved eventually. However, as Reitan (2021) has argued, on probabilistic grounds it is rational to hope that all shall be saved eventually, given infinite opportunity for redemption (see also Gwiazda 2011). Since it is probabilistically certain that all will be saved eventually, it is not the case that it is or would be irrational for anyone to hope that all be saved eventually.

By way of reply, Williams’s romantic theology offers a provocative, possibly idiosyncratic counterpossibility. If Wentworth is a PAI-type, and Midgley’s and Swinburne’s conclusions about human wickedness and corruption are sound (as I think they are), then it is possible that God does not offer to Wentworth-type sinners infinite opportunity for redemption postmortem because it would be irrational for God so to do. Given how far such sinners would have descended into hell, perverted their desires, distorted their faculties, isolated themselves from charitable neighbors (including God), it is possible that they will have sufficiently dulled their consciences such that they are no longer capable of self-consciousness. Were God to offer infinite opportunity, they would not even be aware of what was being offered or to whom; like Wentworth’s experience, each opportunity for escape would simply be muddied, unintelligible sounds. So long as this is possible, then, the suggested additions to Lewis and Murray’s argumentation provide more reason to be suspicious of infinite opportunity arguments than to be suspicious of Williams’s romantic theology. Maybe God only offers a finite amount of opportunity, until the damned become unresponsive to his advances.¹⁹ In this respect, my argumentation does not beg the question against the universalist; it merely gives one reason to refrain from assenting to (some form of) universalism by dint of the (live) possibility that postmortem choices can be self-destructive. Possibly, some folks really do have a death wish.

Objection 2: Even if coherent, PAI describes an implausible, pessimistic view of human nature. No one can be so wicked! I have offered some reasons for thinking this phenomenology of human nature makes coherent the claim that some persons may genuinely prefer damnation/self-destruction to life with God. If that’s right, then the universalist owes the traditionalist/annihilationist an argument for thinking no one can genuinely prefer damnation to life with God, that no one can ever have a death wish – even postmortem. However, as Mele (1987; 2012) confirms, people (even rational people!) backslide; and it is possible that people backslide so far that they cannot want what they used to be able to want.²⁰

What of Talbott’s (2022) argument? Talbott (2022) argues that no one will choose life apart from God forever, because no fully informed person would so choose; and God’s love will compel, overwhelm, effect their eventual salvation. Williams has given reason to think Talbott’s argument unpersuasive. Suppose the damned do experience postmortem God’s loving advances and invitations. Whether the advances occur a finite or infinite number of times is irrelevant. Given PAI, any postmortem loving advances from God the damned will know as evil. They will not be positive, healing, restorative, salvific experiences; they will be negative, further confirmation that the damned have made the right choice – God’s loving presence is a threat to their self-image and self-preservation (cf. Ward, 2007, pp. 139-140). Why? Because they know good as evil, and have effectively severed themselves from the possibility of knowing good any other way on earth and in the afterlife. No amount of love, care, compassion, mercy, or information will effect their salvation anymore: their hearts are too weak, and only can wait, hate, and disintegrate (cf. Ward, 2007, p. 140). Probabilistically, even if God presented infinite opportunity for escape or restoration, they would not take it.²¹ The damned do not want to be loved. So, Williams has given reason to think Talbott’s argument unpersuasive. Some people can be so wicked.

In sum, I have argued that it is coherent to claim (ii), that some persons genuinely desire life apart from God. Initial attempts to render (ii) incoherent fail because they are implausible. Moreover, initial attempts to render (ii) coherent begin on promising notes, yet remain incomplete in some respects. Charles Williams’s romantic theology, I have shown, supplies what is lacking in the latter, relying extensively on the concept of progressive affective impotence. I defended my main thesis against some objections to close.

Williams could not imagine someone unwilling to be loved – and then he remembered the Fall. In the final analysis, it makes sense to claim that God can give the damned what they really want: hell.