Sceptical theism, the parent analogy, and many goods

Abstract

William R. Rowe argues for the low probability of the existence of God given our ignorance of the goods that come from apparently gratuitous evils. After exploring this argument, I present Stephen Wykstra's response, which is known as ‘sceptical theism’, focusing on the evocation of the so-called ‘parent analogy’. According to the parent analogy, God's knowledge, compared to ours, is analogous to the comparison between a parent's knowledge and a one-month-old infant's. After pointing out some difficulties with this analogy, I develop an improved version of sceptical theism. My main point is that the most valuable disinterested love and awe for God can be best developed in a world with two evils: our ignorance about most of the justifying goods and the apparent divine absence to sufferers. Moreover, since disinterested love and awe for God bring numerous benefits to human beings, God has good reasons for creating a world with these two evils.

Introduction

Rowe (1996) argues that certain types of evil are evidence that there is no omnipotent, omniscient, and perfectly good being that we call ‘God’. He thinks that evil can be justifiable if it leads to some greater good, which cannot be obtained without that evil. And yet, according to him, there are unjustifiable evils that provide evidence against the existence of God. Therefore, even if the existence of some evils can be justified by the greater goods they make possible, other evils do not have such a justification, namely, gratuitous evils. Rowe uses two cases of evil to argue for the existence of gratuitous evils: the suffering of a fawn because of a forest fire, whose many days of suffering are ended by death (I will call this case E1); and the suffering of a little girl in Flint, Michigan, who was severely beaten, raped, and strangled in early 1986 (I will call this case E2). On the basis of these two cases, Rowe argues that we are in a cognitive situation S wherein no good we know justifies God in permitting E1 and E2. Given cognitive situation S, we have E: that ‘no good we know of justifies an omnipotent, omniscient, perfectly good being in permitting El and E2’ (Rowe (1996), 263). Accordingly, Rowe assures us that we have the information that E is the case from situation S, such that, if E is the case, it can be used as evidence to infer the hypothesis that ‘~G: There is no omnipotent, omniscient, perfectly good being’ (Rowe (1996), 263).

Keeping with the sceptical theism pioneered by Stephen Wykstra, Rowe's argument can be convincingly challenged in two steps. Let H be whatever evidence we are considering and k the background information shared by both proponents and critics of the hypothesis. The first step is to accept the plausibility of Wykstra's Condition of Reasonable Epistemic Access (CORNEA) according to which,

[i]n cognitive situation S giving new input E, E is levering evidence for H only if it is the case that . . . the conditional probability of E on not-H – viz, P (E | not-H & k) – is below 0.5. (Wykstra and Perrine (2012), 392)

Wykstra and Perrine (ibid., 380–382) list three necessary conditions for evidence to be levering (ibid., 392). The first is that levering evidence – which is another way of referring to ‘robust evidence’ – must have information that, when obtained, changes the probability of hypothesis H concerning what it was before obtaining E. The second necessary condition is the requirement that there must be ‘evidence for or against a hypothesis that is of sufficient strength to shift the rational credibility of a hypothesis from one square state to another’ (ibid., 381–382), where square states signify doxastic attitudes: belief, non-belief, and disbelief. The third is that we must start from a neutral or agnostic position with respect to the hypothesis. The notion of leverage evidence determines the weight of particular evidence rather than all the evidence for or against a hypothesis. We must place each relevant hypothesis in the same square state of non-belief, Pr (H) = 0.5, and then see whether adding evidence to the relevant hypotheses leads to belief or disbelief in the hypothesis.

Is someone justified when inferring from S that there are no goods that validate God's permission to let E1 and E2 occur? According to CORNEA, in an epistemic situation where we have E, we are only justified in claiming that God does not exist if it is reasonable to believe that Pr (E/G & k) is less than 0.5. Wykstra's argument appeals to the parent analogy to demonstrate that Pr (E/G & k) is not less than 0.5. The parent analogy provides a reason for thinking that if God exists, our epistemic situation would be identical to S, where there is E. Proponents of the parent analogy believe that our knowledge in comparison to God's is analogous to the knowledge of a one-month-old infant in comparison to a parent. In the same way that a one-month-old infant is unable to know her parent's reasons for the pain they allow her to suffer (for example when administering a vaccine), there is no good reason to think that we could know God's intentions when allowing for specific evils. As Wykstra puts it,

[t]he linchpin of my critique has been that if theism is true, this is just what one would expect: for if we think carefully about the sort of being theism proposes for our belief, it is entirely expectable – given what we know of our cognitive limits – that the goods by virtue of which this Being allows known suffering should very often be beyond our ken. Since this state of affairs is just what one should expect if theism were true, how can its obtaining be evidence against theism? (Wykstra (1984), 91)

According to the parent analogy, Pr (E/G & k) is extremely high, close to 1. For even if there are goods that justify the divine permission of E1 and E2, our knowledge of those goods is as likely as a one-month-old infant knowing the intentions of her parents in allowing her to suffer. Since our knowledge of God's justification is unlikely, our cognitive situation would be precisely the same as the infant's. If the parent analogy is correct and God exists, then there is a good that justifies allowing for E1 and E2 even though we are ignorant of what this might be, such that Pr (G/E & k) is equal to Pr (G/k). Consequently, contrary to what Rowe thought, we are not justified in inferring that God probably does not exist from E: no good we know of justifies God in permitting El and E2. As a result, E would not be a good reason to think that God probably does not exist.

Criticism of sceptical theism has gone in two directions. The first criticism argues that as a principle of epistemic justification, CORNEA is extremely problematic. The second criticism states that, although CORNEA may be a good criterion of rationality, it does not challenge the argument against the existence of God because the parent analogy is problematic such that Rowe's argument satisfies CORNEA.¹

In the following, I aim to bring together the strongest objections made against the parent analogy based on its inability to demonstrate that Pr (E/G & k) is exceptionally high, followed by an answer to overcome these objections. I begin by presenting the most formidable objections to the parent analogy as set out in Rowe's criticism, which is picked up on in more recent discussions. I will then answer these objections by using DePoe's (2014) and Teeninga's (2017) insights.

Objections to the parent analogy

Rowe's argument is that Pr (E/G & k) is not high because, unlike adults, ‘a one-month-old infant hasn't developed the concepts necessary for even contemplating the proposition that good purposes may justify parents in permitting pains’ (Rowe (1996), 275). In other words, the gap between God's cognitive abilities and ours is not comparable to that between the parent's cognitive abilities and the one-month-old infant. To support this claim, Rowe continues,

[a]dult human beings, on the other hand, have the intellectual equipment to distinguish intrinsic goods from extrinsic goods, to distinguish different kinds of intrinsic goods, to recognize certain intrinsic goods as superior to others, to form an idea of goods that have never been experienced by living human beings on earth (e.g., total felicity in the eternal presence of God), and to make some reasonable judgments about what goods an omnipotent being would (or would not) be able to bring about without permitting various instances of horrendous suffering. . . . So, we know of many goods and we know of some of the very highest goods that human beings can experience. (ibid., 175)

Obviously, by noting that the gap between God's intellect and ours is not comparable to the gap between the parent's cognitive abilities and the one-month-old infant, Rowe makes a sound argument, for there is no good reason to think that we are like one-month-old infants compared to God in every conceivable way. Therefore, in reminding us of our cognitive limitation, the parental analogy does not lead to the conclusion that we would be ignorant of the divine goods that justify the occurrence of apparently gratuitous evils. Schneider (2020, 66) notes that, in biblical sources, God has given human beings numerous exhortations, warnings, commandments, and doctrinal explanations, which presuppose adult cognitive capacities for understanding. Furthermore, there are other sources of revelation belonging to other religions that contain these elements as well. It follows that there is a solid objection to the parent analogy as proposed by Wykstra; our cognitive limitations do not necessarily lead us to think that our intellect is like that of a one-month-old infant's in comparison to God's in every way.

According to Rowe, there is a more significant weakness in the argument based on the analogy between God and parents. This weakness lies in the difference between God and a loving parent when she allows her infant to suffer intensely for a distant good that could not be obtained otherwise. Loving parents pay attention to the infant during the process of suffering, comforting and expressing their concern and love for her in ways that the infant can perceive. When possible, they explain why the suffering is necessary. This is not the case with God, for the elements are not always present in any obvious way. Many human beings, including believers in God, suffer horribly without conscious awareness of God's presence, attention, comfort, concern, love, and explanation. But here we must ask, does the fact that loving parents are present with the infant, that they provide attention, comfort, concern, and love during suffering, influence the notion that we should know the justifying goods for which we suffer? I do not think so. Of course, we would not expect a perfectly good God to hide the fact of his existence from his children, causing them to suffer for unknown goods, to obscure his love and care from their view. Nevertheless, it does not follow that we should know the justifying goods of the evils that seem gratuitous to us, only that God is not analogous to loving parents as we would usually conceive it.

However, the fact that the loving parents would explain the reason for their child's suffering supports the idea that we too should know the goods that justify seemingly gratuitous evils. As some authors have recognized, it is for this reason that the parent analogy can be used to demonstrate that Pr (E/G & k) is extremely low. Recognizing that God can explain the goods that justify the evils experienced by people, would we expect a perfectly good God to hide the reason for their suffering from them? Hardly. Rowe's argument can be used to block the conclusion intended by the parent analogy, as he proposed. It can also be used to turn the tables on the sceptical theist by demonstrating that people should know the good reason for their suffering if a perfectly good God exists. In this connection, Dougherty notes that,

[t]he central problem for Wykstra's expanded defense of the Parent Analogy is that even if it is true that ability to plan for the future will increase in proportion to the scale of the three properties he mentions, the probability of Transparency will also increase in proportion to the scale of those three properties. . . . For the more benevolent a being, the more they would want sufferers to understand the reasons for which they are permitted to suffer. And the more wisdom they had, the more likely they would know how to do it. And the more power they had, the more likely they would be able to make it happen. (Dougherty (2012), 23)

Much human suffering is increased by not understanding the justifying goods for evils. Furthermore, if a loving parent would like to lessen her child's suffering by, when possible, communicating the good reasons why the child is suffering, one would assume that God would also want to communicate the good reasons why his children are suffering when possible. Once we have established that we have adult cognitive abilities and can grasp good reasons for things, we should be able to understand the good reasons for suffering. Nevertheless, even if we did not have such adult cognitive abilities, Schneider (2020) recalls that God could easily make comprehensible the good reasons for the evils – at least to reassure people who suffer – since he allows human beings to understand complex theological concepts. Therefore, we have no good reason to think that we cannot understand the reasons for evils, and even if we are incapable, God could make them more understandable in accordance with our limitations. So, we have no good reason to think that our cognitive limitations would prevent us from understanding the goods that justify evil and suffering. Dougherty further notes that Wykstra does not tell us why God would not give us cognitive faculties to understand the justifying goods of evil and suffering in the first place.

So far, we have five crucial difficulties posed against the parent analogy. The first concerns our cognitive ability. Even though we have limited cognitive ability compared to God's, we have intellectual equipment necessary to understand many things in our world as adult human beings, such as complex moral and theological explanations. Thus we should also understand why many sufferings seem to be void of justifying goods. The second difficulty is that, unlike a loving parent, God does not show himself to the sufferer in perceptible ways. This point gives rise to a version of the Problem of Divine Hiddenness: we would expect God to demonstrate his presence, care, concern, and love in perceptible ways to those who suffer, but he does not. In other words, we hope that if God allowed horrendous evil for the purposes of obtaining a justifying good, he would make his existence and love abundantly clear to us. The third difficulty is that God can make his reasons for apparently gratuitous evils understandable to us, and even if we cannot understand some things, he should be able to facilitate explanations for our understanding. The fourth difficulty is that God can create us with cognitive abilities allowing us to understand the good reasons for seemingly gratuitous evils. Finally, given that loving parents would explain the good reasons for suffering to their children, and we may be able to understand the reason for the evils that seem gratuitous to us, as a loving parent God would let us know the good reasons for seemingly gratuitous evil and suffering. Alternatively, given the second difficulty cited above, at the very least he should be perceptibly present to us.

According to the above observations, our cognitive situation of not understanding the good reasons for the existing evils and sufferings (as well as the absence of the divine presence to sufferers) is highly implausible given the belief that God is a loving parent. Given this difficulty, the theist's alternative is to abandon believing that God is a good parent. Since this alternative is unacceptable for the kind of theism at stake in this article, I intend to offer an alternative, beginning with the positive sceptical theism of DePoe.

Positive sceptical theism and the strategy of many goods

One way to reconsider Wykstra's sceptical theism and to overcome the difficulties posed above is to resort to another characteristic of the relationship between God and human beings following DePoe (2014).

DePoe's sceptical theism offers reasons for expecting E even if God exists. These reasons are premised on an idea given in Hick (1978) and Moser (2008; 2009; 2013), who argue that God has a good reason to create a world with evils that seem gratuitous to us and that this good reason is the positive value of mystery. In the first place, DePoe argues that the positive value of mystery creates an epistemic distance enabling a genuinely faithful response to God without a violation of the freedom of human beings. For if God's existence were uncontroversial and immediately known, human beings would be forced to submit to God, instead of submitting to him of their own volition. The existence of God should be noticeable only insofar as it makes rational belief in him possible without forcing his worship. Added to this, DePoe argues that seemingly gratuitous evil is the most effective and universal way to create this epistemic distance. According to DePoe, as Hick puts it, the positive value of mystery ‘contributes to the character of the world as a place in which true human goodness can occur and in which loving sympathy and compassionate self-sacrifice can take place’ (Hick (1978), 336). A world without evils that seem gratuitous to us, and especially without unjust suffering, would not lead us to sympathy or personal sacrifice, or to seek out community organization for help and services. These compassionate reactions are only possible because they presuppose that suffering is undeserved and bad for those suffering. If a criminal is receiving her just punishment or a patient is undergoing painful treatment that we know will cure her, we have no moral motivation to avoid either. Hence, for there to be a world where compassionate and sacrificial love exists, seemingly gratuitous evil must exist since it is a necessary condition for the extraordinary good of compassionate and sacrificial love. It appears then, that if God wants to have a world in which such extraordinary good is done by free creatures, he must allow the existence of seemingly gratuitous evils.

These two reasons can be categorized as second-order justifications for why God would create a world which includes apparently gratuitous evils. Now, since these reasons explain why God allows the evil of our ignorance about the first-order justifying goods – a kind of theodicy – we must submit them to the same criteria as alleged goods that justify the divine permission of existing evils. Specifically, we must subject the two second-order justifying goods, set out above, to criteria that are almost universally accepted in the debate about the problem of evil: the goods cited must be compensatory (greater) goods that make the occurrence of evil worthwhile, and could not in fact be obtained without the admittance of evils. In this connection, such second-order justifying goods are susceptible to the so-called ‘accommodation strategy’ (Schellenberg (1993)), which aims to demonstrate that certain proposed justifying goods can be obtained without the permission of certain evils. The literature is therefore replete with proposals for justifying goods made possible by evils that have been contested.² I want to demonstrate that positive sceptical theism can be integrated with another perspective resistant to the accommodation strategy.

I will argue that if there are second-order justifying goods made possible by apparently gratuitous evils and apparent divine absence to sufferers, even if the goods can be obtained without these, God would have sufficient reason to allow evils which appear gratuitous or could himself appear to be absent if those goods become sufficiently more valuable because of our ignorance or because of God's apparent absence. This is a strategy followed by Teeninga (2017), used to block Schellenberg's argument that the existence of ‘non-resistant unbelievers’ implies that God does not exist. Complementarily, he appeals to the strategy suggested by Dumsday (2010) that a cumulative case can be made to demonstrate that many goods result from the existence of non-resistant unbelievers. Furthermore, in Teeninga's strategy many goods become more valuable and more numerous, which are made possible by the non-resistant unbeliever. I believe that these kinds of goods are also made possible by two evils: our ignorance about most of the first-order justifying goods and the apparent divine absence to sufferers. Since God is a loving parent and values the good of his children, he would only allow these two evils if there have been many goods, as well as goods made more valuable by allowing for these two evils. As such, my many goods strategy should explain only two evils: our ignorance and the apparent absence of God. Of course, it facilitates my task of demonstrating that these many goods are compensatory, for, as will be shown, many goods are made possible only by two specific evils.

Following Teeninga's (2017) same line of argument, let W be a world exactly like ours and W′ a world wherein we know all or most of the first-order goods that justify evils and God is perceptibly present to sufferers but is identical to W in all other relevant respects. If W has a total value greater than W′, God has a justifiable reason for creating W instead of W′.

Sticking with Teeninga, although moving along slightly different lines, two things should be noted. First, it may be the case that none of the proposed justifying goods outweigh the good of understanding the first-order justifying goods of the evils permitted by God and perceiving his presence in suffering. However, what must be considered are the second-order justifying goods, which when understood individually appear non-compensatory but when taken together are compensatory; that is, they make possible the existence of more valuable and numerous goods. So, to demonstrate that my many goods strategy fails, one must argue that the total value of all proposed second-order goods in W is not compensatory given the two evils I have in view, which do not exist in W′. Second, our ignorance and the apparent divine absence need not be logically necessary conditions for the existence of any second-order justifying goods. By the strategy of accommodation it is possible that each of the proposed justifying goods can exist without our ignorance or apparent divine absence. However, it may be the case that the two evils are necessary for God to obtain a large amount of justifying goods at once. Furthermore, even though some of the goods can be obtained in W′, they can still count in favour of W if they are more valuable in W than in W′. My strategy also succeeds if the different means of obtaining such proposed goods causes a more significant number of evils than the two I have in view. Consequently, my many goods strategy succeeds even though all identifiable goods are not compensatory when considered individually and can be obtained in W′.

There, then, two tasks left for this article. The first is to identify individual goods, which alone may not outweigh the two goods present in W′ but collectively outweigh them. The second is to find goods that can be accommodated even in W′ but have a higher value in W. For my many goods strategy to be successful it must demonstrate that the total value of W is greater than W′. In the following, therefore, I propose that two compensating goods demonstrate that world W can outweigh world W′, even though these goods can only be obtained in world W. The justifying goods I propose are disinterested love for God and particular moral virtues, which can only be further developed by the admittance of the two evils I have in view. Before going into the argument, it is important to mark two canonical accounts in which these evils occur in the lives of two biblical characters. Turning to these accounts is not without necessity, for these justifying goods as they are derived from certain aspects of the human relationship with God were described before contemporary formulations of the problem of evil and sceptical theism. The first canonical account belongs to the Judaeo-Christian tradition, namely, the suffering of Job, while the second is restricted solely to the Christian tradition, namely, the suffering of Jesus Christ.

The suffering of Job

The canonical book of Job is about the life of a man who suffers without understanding the reasons for his suffering and who cries out for God's help without receiving his presence, care, or love. In this connection, Job appears to be a person who suffers the very evils I am considering in this article.

The prologue in the first two chapters of the book of Job is particularly relevant to my discussion, describing Job as a wealthy and prosperous man, God-fearing and of perfect moral virtues. After this description, the narrator proceeds to the meeting of the divine council. Among those present at the meeting is someone who has the task of observing the behaviour of human beings on Earth to bring accusatory charges against them. After God assesses Job as possessing perfect moral virtues, the accuser questions why Job loves God and lives a morally virtuous life, pointing out that it is possible Job conducts himself in the way he does because God has maximally benefited him, protecting him, his family, and his possessions from tragic misfortune.

The dialogue between God and the accuser ends with God giving permission to cause Job's suffering. While Job and his family live peacefully without knowing the decisions made in the divine council, suddenly their lives change dramatically. Four messengers tell Job of four disasters: the first describes the destruction of his oxen and asses; the second the destruction of his sheep; the third the destruction of his camels; and, as a climax, the fourth describes the death of his children. Job's immediate response to his losses and sufferings is that he blesses God's name, convinced that God is deserving of disinterested love through all of life's happenings, regardless of what is given and what is taken away. The narrator concludes by declaring that Job's virtuous character remained intact, informing us that the reason Job loved God and lived a morally virtuous life was not because of the benefits he received from God.

The narrator moves on to the second dialogue between God and the accuser, in which God allows him to cause physical pain to Job without taking his life to see if he will blaspheme against God. The text states that the accuser inflicted Job with terrible physical pain throughout his body. After his wife counsels him to declare God accursed, he responds by expressing that we must receive both good and evil from God. The narrator concludes the prologue by reiterating Job's morally virtuous character.

The challenge set by the accuser in the prologue guides the rest of the book: will Job remain loving towards God and maintain a morally virtuous character despite the misfortunes in his life and the absence of their explanation through divine presence?³ Balentine poses the question of the accuser in the following terms:

It is God who has sheltered Job behind a protective hedge; God who has determined the boundaries within which piety can flourish without distraction; and God who has chosen to respond to piety with blessing. But what if God construed another world, with different boundaries and other fortunes? If God did not create the conditions that make devotion natural and prosperity the norm, would God hear the praises of human beings who have other choices to make and different reasons for making them? Is God alone, apart from any reward or retribution, worthy of loyalty and devotion? The question is theological in the most profound way. What kind of God is God? J. Gerald Janzen has framed the critical issues clearly: Is God intrinsically worshipful? Is deity capable of creating a creature who, somehow, attains to such freedom and independence, such spiritual and moral maturity, as to be in a position to choose to offer God worship and service because of God's intrinsic worthiness to be loved? (Balentine (2006), 54)

What the accuser says has philosophical merit; Job chooses to love God and live virtuously under specific circumstances. What would he do if the circumstances were changed?

The accuser proposes that love of God and living the morally virtuous life can be explained only via the expectation of a reward. Yet, because Job is presented as someone possessing love for God and living a morally virtuous life without depending on divine benefits or expecting rewards, Job acts as a universal paradigm to be reached. The central question in the book of Job (and therefore in the world God created) is what role does reward play in love for God and the morally virtuous life? Similarly, Job's ignorance of the first-order justifying goods of his suffering and apparent divine absence is no different from our ignorance of the first-order justifying goods through which God permits particular sufferings. However, we also face the same question as Job: will we love God and live a morally virtuous life for no apparent benefit to ourselves in a world that does not guarantee clear rewards?

The suffering of Jesus Christ

The second canonical (strictly Christian) account is the so-called ‘cry of abandonment’ of Jesus Christ in his crucifixion. This cry of abandonment is in Mark's account of the crucifixion, where Christ's last words were: ‘My God, my God, why have you forsaken me?’ Bell (2020) argues, based on Adams's (1999) theodicy and Stump's (2012) telepathic connection theory, that Christ did not know the good reasons for his suffering and did not have the divine presence with him. According to Adams's theodicy, divine identification with all aspects of human suffering is a fundamental facet of Christian theodicy. According to Stump's telepathic connection theory, Christ identified himself with the mental states of every human being through a supernaturally amplified mirror neuron in his mind. Thus, there was an epistemic rupture in the cry of abandonment, a cognitive separation between God the Father and God the Son. For if God the Son, in the person of Christ, identified himself with all aspects of human suffering then he must have also experienced divine hiddenness, an aspect of human suffering. Given Bell's understanding of what divine hiddenness is, Christ had no epistemic access to God's good reasons for the existence of his particular suffering, such that when his suffering was intense, Christ felt abandoned by God. Just like Job, however, Christ maintained a love for God and conducted a morally virtuous life during the period of his suffering.

These canonical accounts demonstrate that the disinterested love for God and a morally virtuous character conducted for no apparent reward or benefit are shared by two traditions of theism, Judaism and Christianity. I argue that disinterested love for God can be better developed and more valuable in the cognitive situation of seemingly gratuitous suffering and the apparent absence of the divine. For other (albeit interconnected) reasons, I argue that a morally virtuous character can be better developed because of the existence of the two evils I have in view.

The positive sceptical theism of many goods

Let us return to the W and W′ worlds. In world W′, which is similar to our world W in all relevant aspects, there are pains and sufferings but we know all or most of the good reasons for justifying these. In world W′, when a person's suffering is intense and willing to seek God, God presents himself in a noticeable way. Now think about our world W, in which there is great pain and suffering, but where we do not know most good reasons for them; God does not present himself in a noticeable way to most sufferers, even those willing to seek him. Now we must ask, in which of these worlds is disinterested love for God best developed and most valuable?

Thomas Aquinas, quoting Aristotle, defines disinterested love for God as an authentic friendship without interested motives of benevolence (Aquinas (1947), II-II.23.1). Aquinas answers the question, ‘Whether out of charity God ought to be loved for Himself?’ in this way:

to enjoy is to cleave to something for its own sake. Now ‘God is to be enjoyed’ as he says in the same book. Therefore God is to be loved for Himself. . . . we love God, not for anything else, but for Himself. For He is not directed to anything else as to an end, but is Himself the last end of all things. (ibid., II-II.27.3)

It seems that the more valuable disinterested love for God is better developed in world W than in world W′. For in seemingly gratuitous suffering and apparently divine absence, it is only possible to love God without receiving any predictable rewards.

Suppose that in our world W, Job suffered as described in the canonical text: unaware of the reasons why he is suffering and without feeling the loving presence of God. In world W′, however, Job suffered in precisely the same way, except that he knew the first-order justifying goods of his suffering. He knew he was helping renew messianic hope in the early readers of the Bible, becoming a model of disinterested love for future generations. In world W′, because God would be present in a perceptibly loving way, Job's love for God could be sustained by what he receives from God, namely, the first-order justifying goods and his perceptibly loving presence. Nevertheless, in W, since Job had knowledge neither of these first-order goods nor of the perceptibly loving presence of God, he was only able to maintain love for God without receiving a reward in return. As an analogy, consider a person who charitably gives to poor people without receiving a reward. If charity were always rewarded, then she would not be better positioned to develop the most valuable disinterested love for poor people than she could be if charity were not rewarded. If charity were not rewarded, she could only be charitable if she was willing to be charitable without receiving a reward. So it is in a world where charity is not rewarded where the most valuable love for the poor would be better developed.

The idea is that we are in a better position to develop a more valuable disinterested love for another person in the situation where we do not receive a reward. One way for God to develop a more valuable disinterested love in us is by not providing cognitive access to the first-order justifying goods for the suffering we experience and by obscuring his presence from us. In these cases, we find ourselves in a situation where we can love God without expecting to receive a reward and hence in a situation where a more valuable disinterested love is promoted. In a world like W′, the most valuable disinterested love for God would remain undeveloped since we could easily choose to love God merely for the reward or owing to his loving presence. Now, understanding all or most of the first-order goods of our suffering and having the loving presence of God with us could render the highest disinterested love for him – an extremely valuable good – hard.

In a world wherein we are ignorant of most of the first-order justifying goods of sufferings and experience apparent divine absence, we are in a better position to develop a more valuable disinterested love for God. However, another question appears here: how good is it for human beings to develop a disinterested love for God? Based on the writings of Harry Frankfurt, Silverman (2019) has argued that disinterested love brings numerous goods to the one who loves. These goods (along with many others) provide the lover with pleasant and meaningful ends, contributing to her psychic integration, motivating her self-improvement, increasing her self-knowledge, and improving other relationships. Implied by disinterested love for God, these second-order goods can outweigh the goods of understanding all or most of the first-order justifying reasons and experience of God's presence in world W′. Therefore, it seems that the total value of W is greater than W′.

There are also other second-order goods in the set of compensatory goods, such as those made possible by the awe of God. Our ignorance of God's first-order justifying goods for permitting evils and their apparent absence to sufferers creates greater awe towards God. Ignorance of first-order goods creates a greater awareness of the radical difference between divine knowledge and human knowledge. If we knew all or most of the reasons why God acts, including the reason for his apparent absence, our perception of the gulf between divine knowledge and human knowledge would be less than in a world where we are ignorant of most reasons behind divine actions. This perception of the difference between the two types of knowledge awakens greater awe for God in the human being.

In favour of this last point, I appeal to recent research in social psychology on the awe of God. Shiota, Keltner, and Mossman proposed a definition of awe as ‘an emotional response to perceptually vast stimuli that overwhelm current mental structures, yet facilitate attempts at accommodation’ (Shiota et al. (2007), 944). Building on this definition, Krause and Hayward define awe ‘as feeling overwhelmed by the vastness of God's creations, power, wisdom, and timelessness coupled with the effort to assimilate these experiences into new mental models’ (Krause and Hayward (2015), 51). They hypothesize that people with more wisdom are more likely to feel the awe of God:

[t]o see why tolerance of ambiguity and uncertainty might also play a role in linking wisdom with awe of God, it is helpful to return to the definition of awe that we provided earlier. Recall that part of the experience of awe involves efforts to assimilate or accommodate the experience that triggered it. As Keltner and Haidt (2003) point out, this is a difficult task because a person's existing mental faculties do not allow them to fully grasp or make complete sense of what happened. In fact, when it comes to issues involving awe of God specifically, it seems unlikely that a full understanding can ever be attained. So the type of person who is more likely to reach out and embrace feelings of the awe of God when they arise is the individual who understands the limits of his or her own knowledge and has a relatively high level of tolerance for ambiguity. According to Baltes (2004), these qualities are important because he defined wisdom, in part, as understanding the limits of one's own knowledge. Moreover, this rationale is consistent with research by Mickler and Staudinger (2008), which suggested that people who are wise have a greater tolerance for ambiguity. (Krause and Hayward (2015), 53)

Another definition of awe is proposed by McShane:

Awe is an attitude that one typically takes toward things that one experiences as great in some way. . . . awe often involves feeling overwhelmed by its object, unable to grasp it. . . . The experience of awe often feels forceful, immediate, and involuntary: the awareness of the object and the perception of it as great often feels thrust upon one, unbidden and outside of one's control. . . . Awe, as I have described it here, shares many features of the sublime: both are usually described as powerful feelings in response to something perceived as great in some way that involve feeling overwhelmed and humbled by it. (McShane (2018), 473–474)

Our ignorance about most of God's first-order justifying goods for permitting evils and their apparent absence to sufferers creates a greater awareness of the difference between divine knowledge and human knowledge, which is capable of developing greater awe for God in the human being.

Furthermore, several recent experiments in social psychology demonstrate that the effects of awe are beneficial to those who experience it, causing them to behave in ways that durably benefit others. In these experiments it is shown that awe could:

1. (i) improve an individual's subjective wellbeing, promote the feeling that they have more available time causing them to be less impatient, more willing to volunteer their time to help others, and to be less materialist (Rudd et al. 2012));

2. (ii) promote collective engagement, being more generous, making more ethical decisions, and assuming prosocial values (Piff et al. (2015); Prade and Saroglou (2016));

3. (iii) increase the feeling of deep connection with others, causing a greater satisfaction with life (Krause and Hayward (2015));

4. (iv) improve mood and concern for others (Joye and Bolderdijk (2015));

5. (v) inspire environmental attitudes and commitments (McShane (2018));

6. (vi) increase creativity (Chirico and Gaggioli (2018));

7. (vii) increase honesty, humility, and happiness (Stellar et al. (2018)).

We therefore have strong evidence that awe contributes to the better development of well-being and valuable virtues in those experiencing it.

My argument is consistent with the fact that many believe they receive some benefit (first-order or second-order goods) from their relationship with God. W is a world where some of the benefits of loving and serving God are evident. While some people love God for his own sake, others choose not to, although they still enjoy the benefits. People who not love God for his own sake but still enjoy the benefits may be placed in epistemic ignorance about their suffering and a lack of awareness of God's presence. For once a given agent fails to love God for his own sake or needs to have his love for God perfected, a world like W is needed, wherein she can be in a situation to experience both evils mentioned above. In this situation, the agent has the choice to continue loving God or not. For example, at t₁ some people know they receive goods from their relationship with God, but the effects of apparently gratuitous suffering and the absence of the divine induce t₂ at which point a person can develop a disinterested (and most valuable) love of God. Although some may know and enjoy the goods of God's relationship at t₁, circumstances can change at t₂ to better develop their love of God. Furthermore, God can reward someone, making himself perceptibly present again at t₃ after a disinterested love has already developed, as he did to Job. There cannot be people at t₂ who know they benefit from their relationship with God since at this point God appears absent and distant. Furthermore, we enjoy greater awe because most of the first-order goods (enabled by the evils experienced by the agent) are unknown to human beings.

My strategy explains the morally virtuous attitudes present in the two canonical accounts above. At the end of the book of Job, when confronted with the superiority of divine knowledge, Job recognizes the limitation of his knowledge compared to divine knowledge and can forgive his friends who caused him grief during his suffering. According to Newsom (2003) Job's experience was of the sublime, which is connected with awe. Similarly, while experiencing ignorance of divine reasons and apparent divine absence, Christ forgave his executioners.

For all these second-order justifying goods that have been offered and are made possible by world W, we can once again conclude that the total value of W is greater than W′.

Conclusion

The most valuable disinterested love and awe for God can be best developed in a world with inexplicable sufferings and apparent divine absence. Moreover, since disinterested love and admiration for God bring numerous valuable benefits to human beings, God has good and unselfish reasons for creating a world with the two evils I have set out above. My strategy favours the apparent divine absence in a world where inexplicable suffering occurs as the most effective promotion of the love and awe of God, which in turn is capable of developing into the most valuable disinterested love (along with many other goods) in most people. The question of how many other objections raised against sceptical theism this strategy can help to overcome remains to be seen in future work.