Intellectual Humility and the Argument from Evil: A Reply to Zain Ali

Abstract

This is a response to Zain Ali’s critique in this journal of our presentation of a ‘right relationship’ normatively relativised ‘logical’ Argument from Evil. Our argument aims to show that the existence of horrendous evils (as defined by Marilyn Adams) is incompatible with the existence of the personal omniGod (a person or personal being who is all-powerful, all-knowing, and perfectly good), given certain reasonable judgments about what a personal God’s perfect goodness would imply about the way God relates to those caught up in horrendous evils. We reply to Ali’s main criticism that our assumptions about divine goodness are unjustified, and show a lack of intellectual humility. We defend the claim that, if God is a person, then God’s goodness is moral goodness according to our best human theory of what that implies. We accept that God’s situation as creator and sustainer of all that exists may justify ‘divine exceptionalism’: God’s personal moral goodness may be consistent with ways of relating to others that would fall far short of perfection in human-to-human relationships. But in that case, we argue, intellectual humility may be better served by accepting that God is so exceptional that God should not be understood as a person at all, which is the prevailing Muslim view, as Ali himself acknowledges.

We are grateful to Zain Ali for his thoughtful critique from the perspective of a reflective Muslim of our ‘Normatively Relativised Logical Argument from Evil’ (NRLAfE) (Ali 2024). Before responding, we will draw attention to the key features of our argument, as we recently set it out in Bishop and Perszyk (2023, pp. 29–34).1

1. A Normatively Relativised, ‘Logical’, Argument from Horrendous Evils

First, our argument purports to show, not that God does not exist, but that God according to a certain conception does not exist. That conception is of God as personal omniGod (that is, a person or personal being, who is all-powerful, all-knowing, and perfectly good). This relativization to a specific conception of God is a general feature of all Arguments from Evil (AfEs), which must proceed on the basis of (perhaps implicit) specific assumptions about what it would be like if God did exist.

Second, our purports to show that there is no God according to the personal-omniGod conception only on the assumption that certain specific normative claims hold true about what God’s goodness implies. Again, this normative relativization is a general feature of AfEs, which must all maintain that some facts about evil are either incompatible with, or at least (highly) unexpected in the light of, God’s goodness.2

Third, our argument purports to show that certain facts about evil are inconsistent with the personal omniGod’s existence. To use the established terminology, our argument is a ‘logical’ Argument from Evil (‘LAfE’), rather than one that claims to show only that our total available evidence makes it (highly) improbable that God exists (known as an ‘Evidential’ Argument from Evil (‘EAfE’)). LAfEs remain embarrassing to theists who may be reasonably unfazed by the improbability of God’s existence which a successful EAfE yields, because they rely on claims to special revelation to motivate their acceptance that God exists.

Fourth, our argument accepts that an Argument from Evil based on the mere fact that there is some evil will not succeed. To have a chance of success, an LAfE must focus on some more specific fact or facts about evil and claim that those facts are incompatible with God’s existence. But not just any more specific facts will do here, e.g., that people stub their toes, get paper cuts, and tell little white lies. Our focus, rather, is on horrendous evils, in Marilyn Adams’ sense of the term—that is, ‘evils the participation in which (that is, the doing or suffering of which) constitutes prima facie reason to doubt whether the participant’s life could (given their inclusion in it) be a great good to him/her on the whole’ (Adams 1999, p. 26).

Fifth, our argument concedes that, prima facie, the personal omniGod may have morally acceptable reasons for causing or allowing horrendous evils to occur, since that may be unavoidable even for an omnipotent agent if some outweighing good is to be achieved (for example, goods which presuppose the existence of morally significantly free beings). Following Adams again, however, our argument makes the normative assumption that, to be perfectly good, the personal omniGod would need to be good to each participant involved in horrors, not simply ‘balancing off’ those evils with an outweighing good, but ‘defeating’ them through bringing all the participants involved in horrors into the incommensurably great joy of everlasting life with God. We think that holding that God acts according to this scenario (we will call it ‘the Adams scenario’) is the only kind of reply theists have available, short of taking God’s goodness to be consistent with a more utilitarian account that would allow a perfectly good personal God to leave some horrendous evils undefeated. The Adams scenario takes on the burden of an understanding of eternal life as post-mortem personal existence in which, moreover, major developments can take place to defeat horrors. Both philosophical and theological objections to that understanding are possible, but our argument does not pursue them.

Instead, sixth, our argument expands the normative assumption made by Adams: we say that it may reasonably be held that the personal God’s goodness implies, not only that God will be good to each historical person, but also that God will relate to other persons in ways that are consistent with sustaining right relationships amongst persons. And we conclude that it is reasonable to hold that a person who first allows people to suffer horrendous evils and then, post-mortem, ultimately brings them into an everlasting relationship with himself, would not have acted in accordance with the ethics of right interpersonal relationships in so doing. Relative to that normative assumption, then, the existence of horrendous evils, even when supposedly defeated by the personal God according to the Adams scenario, is not consistent with God’s being perfectly, morally, good. Perfect personal, moral, goodness requires acting always in right relationship with other persons; plausibly, God’s first allowing horrendous evils and then defeating them post-mortem is not consistent with so acting; plausibly, then the existence of horrendous evils is inconsistent with God’s being perfectly morally good.

Finally, seventh, our ‘right relationship’ argument does not—as the ‘plausibly’ modifiers just used indicate—claim to be generally decisive; rather, it aims for the more modest conclusion that those who reasonably make the relevant judgments about what constitutes right interpersonal relationships, have reasons (that may be decisive for them) for concluding that, given horrendous evils, the personal omniGod does not exist. That modest conclusion is powerful enough, however, to show that the existence of the personal omniGod is open to reasonable doubt. In particular, theists who make the relevant normative judgments have reason for rejecting the personal omniGod conception of who or what God is, and for holding an alternative understanding of the concept of God. Thus, for these reflective theists, the Argument from Evil (on our version) serves as a prolegomenon to a discussion of what alternative understanding of God’s existence might be held. That is the way we deployed our argument in Bishop and Perszyk (2023), where we defended euteleological metaphysics as a coherent and religiously adequate alternative to ‘personalist’ understandings of theism that hold that God is, quite literally, a person or a personal being.

In this response to Ali, we will focus on what we take to be his central objection to our argument, namely, that the evaluation we make of God’s goodness in relation to participants in horrors is unjustified, and, indeed, shows a lack of intellectual humility. Ali does raise some independent objections, but we will set these aside here, noting briefly our reasons for doing so. First, Ali correctly observes that our NRLAfE cannot apply if there is no person-to-person relationship between God and humans because humans are not persons (see Ali 2024, p. 6). Ali motivates this startling suggestion by appealing to the idea that ‘the self’ is an illusion (as held by some Sufi Muslims, but familiar in Buddhism, and also to readers of David Hume). The question whether there is such a thing as the self is, however, a question about what it is to be a human person; rejecting a ‘substantial’ self, then, need not amount to denying that humans are persons. Second, Ali objects that our ‘normatively relativised’ approach will make perfect divine goodness ‘an ever-shifting goal post’ because ‘a normative commitment always seems to be at hand to doubt divine goodness even when there is no suffering’ (Ali 2024, p. 13). It is, of course, true that there can be (unlimitedly) many NRLAfEs (relative to the judgment that perfect goodness would make this patch of grass greener, an omnipotent personal God falls short, for example). For an NRLAfE to be at all plausible, however, the normative judgment to which it is relativised must be one to which some reflective theists or would-be theists are committed. As we have explained above, and will explore further in what follows, our argument rests on a quite specific normative evaluation, which we think many will find reasonable, as to how well God relates overall to participants in horrendous evils if God is a personal being in personal relationship with them.

2. Personal Divine Goodness and Moral Goodness

In his discussion of the notion of divine goodness (Ali 2024, Section 4), Ali first raises the concern that our argument assumes that personal divine goodness must be moral goodness, when alternative understandings may be available. In fact, of course, we agree that divine goodness is not moral goodness. On a ‘non-personalist’ account of theism, such as our own euteleological account, since God is not a personal being the idea that God is good in the way that a person is good does not get a grip—indeed, the more basic idea that God is ‘a’ being who ‘has’ goodness as a property does not get a grip either.3 But when God is understood literally as a personal being, we think it is conceptually necessary to accept that the personal God’s goodness is (or includes) moral goodness. According to our contemporary ‘forensic’ concept of a person, persons are agents who, when properly functioning under the right conditions, have moral responsibility for their actions and omissions, with both their behavior and their character accordingly open to moral evaluation. (Human persons may sometimes be beyond evaluation for their moral goodness—but that will always depend on some impairment that could not affect a personal God.)4

It might be urged that, although God’s goodness is moral goodness, the standards applicable to God may differ from those our best human theory of personal morality prescribes. That is the second concern Ali raises in discussing divine goodness. But unless God’s moral goodness is at least goodness according to our own highest moral ideals, we lack a proper sense of what it means to ascribe perfect moral goodness to God. That would be problematic, since our worship of God, if God is a person, needs to be based on God’s perfect moral goodness if it is not to reduce to mere submission to God’s superior power. Commenting on our claim that God’s actions under the Adams scenario might reasonably be judged to be morally flawed because God would seem to be contriving or manipulating his relationship with created persons, Ali says: ‘[s]uch a judgement is understandable within the context of interpersonal human relationships; however, God is not simply an individual within a community of individuals, God is held to be the being upon whom every other entity is dependent for their very being’ (Ali 2024, p. 8). In our view, however, this ‘metaphysical size-gap’ between God and creatures (to use an expression from Marilyn Adams)5 implies that God is not a personal being, since we take it that relating (or being capable of relating) to other persons in community is conceptually essential to being a person. Ali questions whether this is indeed a conceptual truth, citing (i) the 12th Century Andalusian philosopher, Ibn Tufayl’s, tale about a child raised by a doe on an uninhabited island who comes to discover the truth about human existence in God’s creation; (ii) ‘the first human ancestor who became self-aware’ (and was thus a person before anyone else was); (iii) the thought experiment of imagining waking up one morning to find that everyone else has disappeared; and, finally, (iv) the case of ‘a personal God who exists prior to there being any other persons’ (Ali 2024, p. 9). In response to these cases we think we may safely reply, either ((i) and (ii)) that it begs the question against us to regard the case as a real possibility, or else ((iii) and (iv)) that the case, though possible, is no counter-example because the capacity to relate to others in community is possessed despite the absence of other persons.6

3. Dominating Power and Healthy Relationships

We will assume, then, that the personal omniGod’s goodness is moral goodness according to our own best theory of the highest morality (derived, as theists believe, from God’s self-revelation of God’s goodness). Why, though, would it be reasonable to think that there is something morally flawed about God’s permitting horrors and then defeating them through bringing all caught up in them into ultimate heavenly joys? Our argument claims that God’s relationship with those creaturely persons would fall short of ideal mutually loving healthy interpersonal relationship. Might that not, though, be the fault of creaturely persons, with God having done nothing wrong? (Ali raises this concern, quoting a remark of Bruce Langtry’s, pointing out that a person’s action’s having a flawed outcome does not necessarily reflect badly on that person’s character. See Ali 2024, p. 10.)

We agree that the personal God might, prima facie, remain morally perfect in bringing about and sustaining imperfect personal relationships when their imperfection is due to wholly human wrongdoing. That view is implicit in our accepting that a free will defense establishes a presumption that God may be both good and all-powerful despite ‘moral’ evil (evil caused by the actions of morally significantly free creatures). As noted at the outset, however, we follow Adams in holding that a free will defense that articulates an adequate reason that God could have for permitting horrors cannot by itself vindicate God’s perfect goodness. To be perfectly good, God must also act to defeat the horrendous evils that result from the abuse of free will. God will be morally justified in putting up with moral evil for the sake of free will and the higher goods that presuppose it only if God also ‘mops things up’ with perfect justice and kindness, making sure that horrors that would otherwise render lives overall not worth living do not, ultimately, have that destructive effect. Adams emphasizes that the personal omniGod has infinite loving resources with which to achieve this victory over evil.

We object, however, that if God acts according to the Adams scenario (allowing horrors but defeating them post-mortem), then in the very setting up of a created order in which God knows that this scenario must play out, God knowingly brings about an unhealthy overall relationship between the person God is and the persons caught up, as victims and perpetrators, in horrors. True, God is not directly responsible for the wrongful actions of the perpetrators of horrors, but God is personally ultimately responsible for creating and sustaining in existence persons some of whom he knows will (or will very likely) perpetrate horrors. Indeed, the actions God takes to defeat horrors post-mortem are to be seen as motivated by God’s accepting ultimate personal responsibility for horrors. We say, though, that, if God were a morally perfect personal being, God would recognize that setting up a world in which horrors occur and are defeated post-mortem would be setting things up for a way of personally relating with participants in horrors that was inevitably flawed. Ironically, the ideal loving relationships that God aims to bring about could be achieved amongst creatures only if God’s own personal relations with participants in horrors ultimately fall short of that very ideal. A morally perfect personal God would thus—reluctantly, as we may imagine—turn away from creating a world with significantly morally free creatures, and, if he created at all, would create a world without horrendous evils. The existence of horrendous evils, we then conclude, shows that there is no personal omniGod.7

But where exactly is the flaw that allegedly makes the personal God’s relationships with creaturely persons less than fully healthy? Ali’s discussion presses us to explain further. As Ali’s example of the father who causes his daughter to suffer painful medical treatment and then compensates her when she is cured makes clear (see Ali 2024, p. 4), our morality allows that we may sometimes have good reason to cause or allow suffering we could have prevented. If the daughter’s illness is serious and cannot be cured by less painful methods, then putting her through painful treatment is justified, and sparing her the suffering with the result that she remains ill, unjustified. A loving father would certainly support his daughter while she is suffering and do whatever he could to counter its adverse impacts on her life (the father in Ali’s example does both these things). This kind of everyday example does not, however, parallel what the personal God does on the Adams scenario. On that scenario, there is a person, P (God) who, unilaterally and without mutual consent, sets up and sustains relationships with other persons in which P first (for the sake of outweighing ‘higher’ goods) causes or allows those persons to be involved in horrendous evils, and then provides ‘defeating’ compensation that ensures their lives are ultimately a great good to them as a whole.

Ali quotes a hadith on the changed perspective those who have suffered horrendous evils have once they attain eternal life with God (see Ali 2024, p. 10). Read literally, this hadith teaches that sufferers would deny before God that they had ever suffered or been distressed if they could only ‘dip once in Paradise’. As Ali himself makes clear, however, there can be no suggestion that, from the eternal point of view, horrendous evils did not really occur or were merely inconvenient. What is claimed to be true is that, from the eternal point of view, the potential of horrendous evils to render lives not overall worth living may be defeated. Our NRLAfE need not dispute the bold theist claim that a life marred by horrific suffering can yet ultimately be fulfilled in eternal union with God—for that claim as such does not commit itself, either on the question whether God is literally a person, or on the question whether a suffering person’s ultimate participation in God’s eternity involves that person’s everlasting continuation as a finite personal being in some supernatural paradisiacal world.8

The claim of our NRLAfE is only the claim that, when one understands God as a personal being, and the saving acts of God as involving personal relationships with suffering creaturely persons, then one has a Person who acts (in the Adams scenario) as P does. And we say that one may reasonably judge that any person who acts as P does relates to the other persons concerned in such a dominatingly high-handed and manipulative way that P’s relationships with those other persons fall short of the ideal for good healthy personal relationships—even though P had apparently good reason to allow the horrors, and then to defeat them, and even though P attempts to ‘make everything come right in the end’ by defeating those horrors. P’s way of relating to other persons on this scenario, in other words, seems to exhibit a pattern uncomfortably like the cycles of harming and reconciling found in some abusive intimate partnerships, or, even, the pathology of Munchausen syndrome by proxy, where a person with power over others secretly inflicts harm on them so as to gain approval for skill and virtue in making good on that harm.

4. Divine Exceptionalism

Sympathizers with Ali’s critique may here object that we are unaccountably ignoring a vitally important, but blatantly obvious, feature of God’s personal situation as creator of all else that exists—namely, that it is necessarily absolutely unique. They may complain that, in judging that God relates less than healthily to other persons in the Adams scenario, we are improperly projecting our dislike of high-handed and manipulative behavior in human-to-human personal relationships onto God’s uniquely situated relationship, as the supreme person, to all other created persons.

Notice that those who make this important objection may accept that God is a person in community and that God’s goodness must be perfect moral goodness in relation with other persons according to our own best account of what that requires. Their objection rests on a claim we will call divine exceptionalism—namely, that God’s necessarily and absolutely unique situation in the network of interpersonal relations permits types of action which would not be morally admirable in the human case.9 As a personal being, God is indeed a member of the same moral community of persons as creaturely persons are, and subject to the same moral standards (standards ultimately grounded on God’s very own nature); yet—the idea is—those same standards imply that God’s situation as personal creator is a uniquely morally exceptional one. Ali canvasses this idea of divine exceptionalism when he says that our judgments ‘about the necessity of mature, respectful and mutually trusting inter-personal relationship, seem reasonable for interpersonal relationships among peers, but not necessarily suitable for the divine-human relationship’ (Ali 2024, p. 8).

This is a telling objection. If God is a person, then it is incontrovertible that God is a uniquely exceptional person. If the personal omniGod acts according to the Adams scenario, then, God does so in the very act of creating a world that can achieve the highest forms of goodness, yet only at the cost of permitting horrendous evils—whose overall destructive effect on creatures’ lives God nevertheless successfully overcomes! God’s exercise of ultimate overall controlling power over others for their own (ultimate, and post-mortem) good may thus be claimed not to introduce any diminution in God’s personal moral perfection. Of course, we rightly deplore humans who exercise power over others in dominating ways, accusing them of being over-bearing, manipulative, or patronizing—of ‘trying to play God’! Such behavior is ruled out by the standards of our moral community. And God is a member of that community. But God is the supreme person on whom all else depends for its very existence. Thus, God’s unilateral exercise, for the sake of the overall good, of an ultimate personal power of control (prior to any possible consent from creaturely persons) cannot be morally flawed: God must ‘play God’!10

We will not attempt a rebuttal for this exceptionalist claim, since it suffices to point out that the modest role we have claimed for our argument is not undermined by it. As we noted at the outset, the incompatibility between horrendous evils and the personal omniGod’s existence depends on affirming a certain normative judgment—in effect, the judgment that a personal God would not be an exception to the general judgment that persons bring about less than perfect interpersonal relationships with others if they act in anywhere near as dominating a way as God does under the Adams scenario. Our ‘right relationship’ NRLAfE thus depends on judging that a perfectly good personal God would not be willing to ‘play God’ according to the Adams scenario. Personal omniGod theorists typically find that they attenuate the concept of a person in applying it to God—for example, when they deny (as they almost universally do) that God has a body.11 In the exceptionalist response to our argument, we see this same tendency to attenuate the concept of a person emerging in the moral sphere: God is a personal moral agent, but a uniquely exceptional one for whom actions may be consistent with perfect moral goodness, even though actions of that same general type would be flawed when done by creaturely persons. If such attenuation is necessary for the view that God literally is a person or personal being, it is reasonable to ask why it would not be preferable to admit that God is so exceptional a person that, really, God is not a person at all. We will pursue that question further by considering Ali’s remarks on intellectual humility and the risk of cognitive idolatry.

5. Intellectual Humility—And Who Has More of It

Ali comments on Q42:11 (‘that there is nothing like unto Him’) that ‘this can be seen as consistent with the view that God is a unique kind of personal agent’ (Ali 2024, p. 5). The general view in Islam, as Ali himself admits, is that God’s not being like anything else rules out God’s literally being a person, even though ascriptions to God of personal, and indeed, embodied, attributes are found in the Qur’an and—on certain Islamic views, such as those of the Hanbali traditionalists (e.g., Ibn Qudāma)12—are to be accepted without questioning how they can apply. Understanding God as a person, in personal relationship with creaturely persons, may thus ‘risk idolatry’, even though (for example) it is a revealed truth that God made a friend of Abraham and Moses. We think that the only way to avoid contradiction here—but an effective one—is to allow that personal attributions to God convey truths yet without conveying understanding of what makes those truths true. God sits on the throne, befriends Abraham and Moses, yes, indeed: but we are not to ‘read off’ from those truths an understanding that what makes them true is the existence of a person (unique, but of the same general kind as the persons we are) possessing the relevant personal attributes. Ali, however, thinks that we may understand God to be a person; we avoid idolatry by keeping in mind that God is a unique person amongst persons. This divine exceptionalism exempts God from potentially adverse moral judgments that might otherwise apply. This recognition that God’s person is unique and uniquely situated is required of us, Ali thinks, as a matter of intellectual humility.

Intellectual humility, as Ali says, is the virtue of ‘acknowledging the limitations of our cognitive capacities’ (Ali 2024, p. 12). It is not making oneself lower than one actually is—‘humbling oneself’—rather, it is a matter of accurate self-knowledge. Ali writes:

If indeed there is a personal God, we would have to acknowledge a sizeable epistemic gulf between God and created beings—we cannot so to speak, ‘step into God’s shoes’, or have a ‘God’s eye view’. Without a jury of peers, could we really put God’s character on trial? […] We simply may not have the requisite cognitive capacities and access to the relevant information that would allow for a reasonable assessment to be made (Ali 2024, p. 13).

Ali thus concludes that the normative judgment we make in our argument oversteps our proper cognitive limitations and thus lacks intellectual humility. We think we are interpreting him fairly as diagnosing that the judgment he regards as hubristic results from our failure to respect divine exceptionalism. But, then, we may retort that those who, like Ali, continue to hold that God is literally a person or personal being, are not going far enough in recognizing just how exceptional God is. Maybe it would lack intellectual humility not to make allowance for the unique status of God as a person amongst persons—but perhaps there is a more basic lack of intellectual humility in supposing that God is aptly understood as a person or personal being at all. The dialectic we embark on in proposing our NRALfE, then, can be seen as bringing it to light that, for it to be meaningful for us to regard God as a person, we have to treat God’s goodness as moral goodness according to the highest standards of personal, relational, morality that we know—and that this way of regarding God (with its accompanying construal of divine omnipotence and omniscience as the unlimited powers of a personal being) leads to the high paradox that such a God would, if he existed, refrain from creating a world of significantly morally free creatures. The remedy is to admit that our thinking about God as literally a personal being has led us into a muddle—a muddle that is hardly surprising if we have gone beyond our proper cognitive limitations in taking God actually to be a person. Cognitive repentance is in order! We find authentic intellectual humility in recognizing that the Alpha and the Omega, source and goal of all that is, was, and ever shall be, transcends all that belongs to the created world. We may yet believe (by faith, on the basis of purported special revelation) that personal and interpersonal existence is a key locus for the fulfilment of God’s purposes in creation, but, to avoid cognitive idolatry, we must deny that God is himself a node (even a uniquely great one) in that great complex of interpersonal relationships.

Who has the greater intellectual humility, the one who understands God as a person, but a quite unique one, or the one who says that God must be wholly beyond personhood? Those who agree with Ali in taking the former view that regarding God as a uniquely exceptional person is intellectual humility enough can point, we think, to two important kinds of motivation for their view. Our case for the more radical view of what intellectual humility requires can then be strengthened if we are able to show these motivations not to be well founded.

The first motivation is that any alternative to understanding God as the personal omniGod will be open to some form of the Argument from Evil, so we may not get any further ahead by shifting to some non-personalist alternative. And the second motivation is, that, anyway, we cannot make coherent sense of the central religious practices of worship and prayer unless we understand God as a personal being. Now, we agree that we cannot make sense of worship and prayer unless it is apt for us humans to relate in personal ways to God, and to use personal language in addressing and in speaking about God. That does not entail, however, that we must understand that God actually is a person.13 In our book defending a non-personalist, euteleological theism we argue (Bishop and Perszyk 2023, Ch. 6) that worship and prayer make sense on a ‘non-personalist’ view of God. As well, we consider (Ch. 5) how evil can exist in the creation on a euteleological understanding, arguing that that understanding is not undermined by any ‘logical’ version of the Argument from Evil. We will not elaborate further here, however. We will be content to conclude that—since this issue of what intellectual humility requires is at least moot—we have responded adequately to Ali’s objections to our claim that, for those who make a certain reasonable judgment about what the perfect moral goodness of a personal omnipotent and omniscient God would imply about ideal interpersonal relationships, the existence of horrendous evils implies that there is no such God.