On sin-based responses to divine hiddenness

Abstract

While sin-based responses to divine hiddenness arguments are a road less travelled, they do nonetheless have a number of defenders in the contemporary divine hiddenness literature. I begin this article by exploring the various strategies that have been employed to attempt to motivate such accounts. What none of these strategies seem to take into account, however, is a cluster of facts about the correlation (or lack thereof) between a person's propositional attitudes about God and the degree to which that person displays the relevant moral and intellectual virtues. This article aims to fill this lacuna by mapping out the options available to defenders of sin-based responses in trying to cope with this cluster of facts. I argue that there may be resources available for preserving some aspects of the sin-based approach, but that taking stock of the aforementioned facts will ultimately require the positing of causal factors besides sin in order to generate a sufficient explanation of the phenomenon of non-belief.

Introduction

J. L. Schellenberg's argument from divine hiddenness alleges that there are some human beings whose non-belief in the existence of God is not due to any fault of their own, and that this situation is incompatible with the existence of a perfectly loving God (Schellenberg 1993; 2007; 2015). The dominant species of response to this argument consists in offering what is in effect a theodicy: a claim that there is some greater good for the sake of which a loving God might allow some people to be in a suboptimal evidential position with regard to the existence of God for at least part of their earthly lives. The less popular species of response – though it is certainly not without its defenders – alleges that non-belief is actually the result of culpable resistance to God. Defenders of this latter approach employ a variety of strategies to attempt to motivate the approach, which I examine below. What none of these authors seem to take into account, however, is a cluster of facts about the correlation (or lack thereof) between a person's propositional attitudes about God and the degree to which that person displays the relevant moral and intellectual virtues. This article aims to fill this lacuna by mapping out the options available to defenders of sin-based responses in trying to cope with this cluster of facts. I shall argue that there may be resources available for preserving some aspects of the sin-based approach, but that taking stock of the aforementioned facts will ultimately require the positing of causal factors besides sin in order to generate a sufficient explanation of the phenomenon of non-belief.

Hiddenness arguments and sin-based responses

Undoubtedly the most widely discussed argument from divine hiddenness (ADH) is the one developed by Schellenberg across multiple books and articles over the past three decades. The exact manner in which some of the premises are expressed by Schellenberg has been subject to evolution over the years, but the gist of the argument has remained essentially unchanged since the argument's first appearance (Schellenberg 1993), and can be stated as follows:

Premise 1: If a perfectly loving God exists, then for all created persons who are cognitively capable of relationship with God, it is the case that every such person who is not resisting God is in a position at all times to enter into a personal relationship with God just by trying.

Premise 2: If a person S is in a position to enter into a personal relationship with R, then S believes that R exists.

Sub-conclusion: Therefore, if a perfectly loving God exists, then for all persons who are cognitively capable of relationship with God, it is the case that every such person who is not resisting God believes that God exists; that is, if a perfectly loving God exists, then there are no non-resistant non-believers.

Premise 3: In fact, there are non-resistant non-believers.

Conclusion: Therefore, no perfectly loving God exists.

Schellenberg's is not the only ADH under discussion. Charity Anderson (2021) has advocated a reframing of the hiddenness debate around probabilistic versions of the hiddenness argument, in parallel with discussions of the problem of evil in which it has become typical to treat the topic probabilistically (e.g. Draper 1989). Where ‘H’ is some observed fact about the world – for example, the fact that belief in God and non-belief are rather unevenly distributed across the globe, or the fact that pre-axial humans lacked a monotheistic concept of God – a probabilistic hiddenness argument alleges that H is significantly more probable on atheism than it is on theism, and hence that H strongly confirms atheism over theism (e.g. Maitzen 2006; Marsh 2013).

The majority of replies to Schellenberg's ADH target Premise 1, in effect seeking to offer theodicies that sketch various possible greater goods for the sake of which a perfectly loving God might permit there to be non-resistant non-believers.¹ On the other hand, a substantial minority of responses to Schellenberg's ADH appeal to human sinfulness, calling Premise 3 into question.

Schellenberg's ADH is structurally parallel to the logical argument from evil (see Mackie 1955) in the sense that, given the truth of the argument's conceptual premises (Premises 1 and 2), the success of the argument requires only that there be at least one non-resistant non-believer in the entire history of the world. An attack on Premise 3 looks like an uphill battle, to put it mildly. But as with the logical problem of evil, the plausibility of Schellenberg's empirical premise (Premise 3) is matched by the vulnerability of its conceptual premises. Accordingly, just as probabilistic arguments from evil grant the logical compatibility of God and evil but question whether God would permit the amount and kinds of evil we actually observe (see Howard-Snyder 1996b), probabilistic arguments from hiddenness may simply grant that theism is logically compatible with the existence of some non-resistant non-belief but call into doubt whether God would permit as much non-resistant non-belief as we seem to find in our world.

We can distinguish, then, between two kinds of sin-based account of non-belief: universal and partial. Universal accounts claim that every instance of non-belief is caused by sinful resistance on the part of the non-believer. Only universal sin-based accounts are relevant to Premise 3 of Schellenberg's ADH. Perhaps because of the dominance of Schellenberg's deductive framing of the hiddenness problem in the contemporary literature, the sin-based accounts on offer do almost always purport to be universal in scope.² Partial accounts may yet be relevant to probabilistic hiddenness arguments, however. Partial accounts claim that some non-belief is caused by sinful resistance. Presumably, the extent to which a partial sin-based account constitutes a worthwhile response to a probabilistic hiddenness argument is a matter of the extent to which that account persuades us that some non-negligible portion of the non-belief that we initially thought was non-resistant is actually due to sinful resistance. The non-negligible portion could either be a portion of the total number of non-believers or a portion of the set of causes that are individually necessary and jointly sufficient for the non-belief of an individual or group of individuals.

But what reasons are there to think that non-belief is ever caused by sinful resistance towards God, let alone that all non-belief is so caused? I suggest that the extant literature defending sin-based accounts of non-belief can be divided into broadly three camps as regards the strategy that is employed for answering this question.

In the first camp belong those sin-based accounts that essentially consist in a detailed description of the sinful human condition and an unpacking of its alleged consequences as regards belief in God. Typically such a description draws heavily upon the work of leading figures in what might be termed the Augustinian tradition within Christian theology. William Wood (2013) and V. Martin Nemoianu (2015) draw extensively on Blaise Pascal. Alvin Plantinga (2000, ch. 7), Ebrahim Azadegan (2013), and Tyler Taber and Tyler McNabb (2018) take their cue from John Calvin. Despite varying emphases, the common thread in all of these descriptions is the idea that the sinful human self is fatefully consumed with the love of that which is not God, who is in fact the only genuine source of human fulfilment, and it is this primordial disorder of the will that distorts all aspects of human life, including our cognitive faculties. Our thinking about the matter of God's existence is perhaps especially susceptible to the distorting influences of sin, according to these authors, because of the enormous axiological significance of God's existence. Beyond providing a description of what a notable strand of the Christian tradition has historically asserted about fallen human nature, these authors don't obviously offer anything that could be regarded as positive evidence that this really is the way things are with humans. I suggest that the most charitable way to interpret these authors is as offering what Plantinga (2000, 168–170) terms a model, which is to say, a story that is epistemically possible: true for all we know. Peter van Inwagen has distinguished between a theodicy and a defence as two modes of responding to problems of evil, where the former consists of an account that purports to be ‘the real truth of the matter’, whereas the latter ‘may or may not be true, but which, the teller maintains, has some desirable feature that does not entail truth – perhaps (depending on the context) logical consistency or epistemic possibility (truth-for-all-anyone-knows)’ (van Inwagen 2006, 7). My suggested interpretation of the aforementioned authors would give their accounts the status of defences, in van Inwagen's sense. This status is considerably less epistemically impressive than that which they would possess if compelling positive evidence for their truth were presented. As van Inwagen notes, though, a defence in this sense is commonly taken to have some value in the fields of law, history, and the sciences (van Inwagen 2006, 7).

A second way of trying to motivate sin-based accounts of non-belief is to scrutinize Schellenberg's standard for what would count as an epistemically responsible investigation into God's existence, and to argue that probably no non-believer satisfies Schellenberg's standard. Douglas Henry (2001) and Robert Lehe (2004) have both pursued this strategy. The crux of Henry's move against Schellenberg is the thought that the more a person has the requisite strong passion for truth and the desire for a well-justified belief about God's existence, the less likely it is that such a person will ever be content to conclude the investigation and land on a settled negative position on the matter. Lehe concurs with Henry but goes further, contending that the more that someone exhibits a strong commitment to intellectual honesty, the more reluctant such a person will be to believe that their investigation has been sufficiently free from subconscious bias and self-deception. In a sense, both authors are attempting to shift the burden of proof onto Schellenberg with regard to the question of whether there are any non-resistant non-believers. Again, though, given the truth of Schellenberg's conceptual premises, all his ADH needs is the existence of at least one non-resistant non-believer in the entire history of the world. As such, Henry and Lehe's efforts against Premise 3 of Schellenberg's deductive ADH look very strained, but they may yet have some value if construed as partial sin-based accounts to be offered in response to probabilistic hiddenness arguments, along the lines sketched earlier. That is, they may succeed in reducing our confidence that there are quite as many non-resistant non-believers as we initially supposed.

A third way to try to motivate sin-based accounts is to appeal to alleged positive evidence of a widespread human tendency to wilfully play down clues of God's existence. Mark Talbot (1989) has suggested that the testimony of some converts to theism from non-theism, that they had been wilfully suppressing evidence of God's existence during their prior non-theistic state, provides significant inductive support for the general claim that all non-theists are engaging in such suppression. Jake O'Connell (2011) has offered a range of real-life cases in which a person or group fails to believe that God exists (or to believe some more specific claim such as that Jesus is divine) despite encountering what they take to be a miraculous event. O'Connell suggests that the individuals in these cases constitute a good inductive sample of humanity more generally, such that it is reasonable to conclude that the tendencies that lead people to avoid coming to a theistic conclusion in the face of such evidence are characteristic of humans as a whole.

Doubtless there are many critical questions that can be asked of these three strategies for trying to establish the plausibility of sin-based accounts of non-belief, but as we shall see in the next section, none of these strategies anticipates a certain noteworthy set of data that appears to be at odds with sin-based accounts.

A family of troublesome facts

There is a cluster of facts that is highly unexpected given sin-based accounts of hiddenness, and that none of the authors mentioned in the previous section considers. Sin-based accounts explain a person's lack of theistic belief in terms of that person's moral and intellectual vices. Such accounts, therefore, lead us to expect that there would be a tight correlation between a person's propositional attitudes about God – that is, a person's beliefs about whether God exists and has various attributes – and the degree to which that person displays the virtues of openness to relationship with God and conscientiousness in investigating the evidence for God's existence (which we can henceforth refer to as ‘the relevant virtues’).³ In short, sin-based accounts predict that the more someone displays the relevant virtues, the more likely it is that the person believes that God exists. The problem is, there are several observations that severely call the existence of such a correlation into question, as we are about to see.

The first is something that Jason Marsh (2013) has particularly called to our attention, namely, the phenomenon of non-belief in pre-axial humans who lacked a concept of God that was close to monotheism. Marsh's thought is that such individuals can hardly be blamed for failing to believe in a being of whom they have never conceived, or at any rate, a being whose reality their surrounding culture has never deemed a remotely live option. In short, given that there is no reason to think that pre-axial humans were inherently more sinful than post-axial humans, it seems extremely strained to view the non-belief of pre-axial humans as the result of intellectual or moral vice.

Another fact in the nearby vicinity that has been discussed in the recent literature concerns the uneven spatio-temporal distribution of theistic belief and non-belief, which Daniel Howard-Snyder calls the phenomenon of ‘cultural nonbelief’ (Howard-Snyder 1996a, 452). Stephen Maitzen contends that such unevenness renders sin-based accounts wholly implausible because sin-based accounts are committed to the existence of a tight correlation between theistic belief and the relevant virtues:

Those contributors who explain all non-belief in terms of the epistemic or moral defectiveness of non-believers never address the question ‘Why does that defectiveness vary dramatically with cultural and national boundaries?’ Moreover, given the widely held assumption that, generically speaking, epistemic and moral defects are evenly distributed among the world's peoples, it is hard to see how that question could be answered . . . These dispositions thus resemble other fundamental human characteristics, such as the ability to hear; despite marked differences among individuals, we don't find entire countries whose citizens are nearly all deaf. (Maitzen 2006, 180, 184)

In short, the concern here is that the flaws of character that sin-based accounts contend are the explanation for non-belief would have to be spatio-temporally distributed so as to show up with much greater frequency in some regions of the globe and some stretches of history than others. Not only does this defy common sense, it also appears to be contrary to clear biblical affirmations of the universality of sin's effects upon humanity.⁴

A third and related concern is the thought that the virtues and vices that are relevant to whether someone believes in God, according to sin-based accounts, are not notably present to a greater degree in the lives of believers in God as compared with non-believers. Indeed, Schellenberg contends that there are some reflective non-believers – that is, people who possess the relevant concept of God but who, after examining the evidence, end up not believing that such a being exists – whose ‘investigations are exemplary, even supererogatory, and match in quality those of the most scrupulous of their opponents’ (Schellenberg 1993, 67). Moreover, he suggests that some believers are palpably morally and epistemically deficient and yet don't as a result end up becoming non-believers: ‘[A]s the lives of many believers testify, it's possible to hang on to belief in God and preserve its comforts while through self-deception blunting or subverting the moral force of its implications. So we already know that resistance to God doesn't always lead to nonbelief’ (Schellenberg 2015, 82). The contrastive claim here that seems to jeopardize sin-based accounts of non-belief is that some non-believers are at least as morally and epistemically virtuous in the relevant respects as some believers. As Dustin Crummett aptly puts it, what we have here is the problem of ‘at-least-no-more-culpable-than-anyone-else non-belief’ (Crummett 2015, 46, n. 3).

Clearly, the three themes just outlined are interrelated. I suggest that from this interconnected web we can draw out three distinct objections to sin-based accounts of non-belief:

Objection A: Pre-axial non-believers cannot plausibly be viewed as having resistantly rejected a monotheistic God given that they either possessed no concept of such a God or had no cultural plausibility structures that would make belief in such a God a live option.

Objection B: Many reflective non-believers have been sincere and conscientious in their investigations of theism, or at any rate, no less sincere and conscientious than some of their believing counterparts.

Objection C: Regarding the moral and epistemic traits of character that sin-based accounts claim are determinative of whether someone believes in God, it isn't plausible that the populations of spatio-temporal regions in which theistic belief is concentrated exhibit virtuous and vicious traits at substantially greater and lesser frequencies (respectively) than the populations of spatio-temporal regions in which theistic belief is sparse.

In the remainder of the article, my aim is to explore the options available to defenders of sin-based accounts in responding to objections A, B, and C. The aim is not so much to defend sin-based accounts as it is to try to ‘count the cost’ of defending them. Two clarifications are in order before we go any further.

First, one might wonder how to square the emphasis on unmerited grace as the basis for salvation – which is typical of sin-based approaches – with idea that is at the heart of objections B and C, namely, that sin-based accounts predict that believers in God would exhibit the relevant virtues of moral and intellectual character to a greater degree than non-believers would exhibit them. Given that the Augustinian theological stream in which sin-based accounts of non-belief are situated strongly rejects the idea that salvation is attained through good works (where ‘good works’ can be taken to include the acquisition of good traits of character), don't these objections effectively strawman sin-based approaches? I would suggest not, given the Augustinian tradition's equally clear emphasis on the idea that the intellectual and moral character of a recipient of unmerited divine grace undergoes a process of positive transformation. Plantinga is representative of this tradition when he writes that, ‘By virtue of the work of the Holy Spirit in the hearts of those to whom faith is given, the ravages of sin (including the cognitive damage) are repaired, gradually or suddenly, to a greater or lesser extent’ (Plantinga 2000, 243–244). While Plantinga's qualification (‘gradually or suddenly, to a greater or lesser extent’) may suggest a slight weakening of the correlation between someone's being a believer and her exhibiting the relevant virtues of intellectual and moral character, it doesn't by any means eliminate the expectation of such a correlation given sin-based accounts of non-belief.

Second, I want to emphasize that the focal issue of the present article is whether and to what extent sin might contribute to an explanation for why there would be non-belief in God in a world in which Christian theism were true. I am not concerned here with the question of whether non-belief is an appropriate target of divine punishment.

Rejecting an unknown God

Objection A: Pre-axial non-believers cannot plausibly be viewed as having resistantly rejected a monotheistic God given that they either possessed no concept of such a God or had no cultural plausibility structures that would make belief in such a God a live option.

This objection turns crucially on the sort of awareness of God a person would need to have in order to be able to reject or turn away from God, which rejection the Augustinian tradition takes to be the universal condition of fallen humanity. The objection, I suggest, envisages a person needing to have a fairly determinate monotheistic concept of God in order to be able to reject God.⁵ If that is the correct way to think about things, and if it is true that people in the pre-axial period lacked such a concept,⁶ then those people cannot be viewed as having rejected God. What are the options for defenders of the sin-based account? I suggest that there are at least two.

The first is to appeal to Molinism and invoke counterfactuals about what such people would have freely done in the nearest possible worlds in which they possess a theistic God-concept. This sort of approach is taken by William Lane Craig (1989) in addressing the question of the fate of the unevangelized. In essence, Craig's proposal is that God has providentially arranged the world so that unevangelized regions of space-time are populated only with individuals whom God knows would have freely rejected the Christian gospel if they had been exposed to it. The proposal in view for present purposes is similar: in God's providence, cultures that lack a theistic God-concept are populated only with individuals whom God knows would have freely rejected God if they had had a theistic God-concept. The crucial point is that according to this proposal, the explanation for why these individuals lack a theistic God-concept is a fact about how they freely respond to God in the nearest possible world(s) in which they do possess a theistic God-concept.

This approach will strike many as being too much to stomach. First, there is the fact that the metaphysics of Molinism are deeply controversial, and it would be better for defenders of sin-based accounts to be saddled with as few controversial metaphysical commitments as possible.⁷ Second, there is the sheer implausibility of claiming that the billions of humans throughout history who have not been exposed to monotheism would all have rejected it had they been so exposed. But even setting aside those issues, there still remains the fact that such an individual has not, on the proposal under consideration, actually rejected God, and yet the Augustinian tradition in which sin-based accounts are situated is crystal clear in its contention that all human beings start out their existences in a state of opposition towards God.

A second option is what we might call the de re move. This is the move of counting people who don't have mental states that they would represent to themselves as being about God as nevertheless having mental states that are de re about God; that is, they have mental states that they would represent to themselves as being about X, and in actual fact X = God. Mental states may be taken to include both cognitive attitudes such as beliefs or credences and affective attitudes such as longings or loathings. The de re move is one that has a venerable history, though it is more commonly deployed as a way of explaining how saving faith might be available to those who haven't been exposed to the Christian message (an idea to which we shall turn later). The proposal in view here is that people who lack a monotheistic God-concept can nevertheless be viewed as having rejected God, in virtue of their having mental states that are de re about God. William Wood's Pascalian account of sin as self-deception could well be read as invoking the de re move:

We long for stable happiness, and so we implicitly long for God, the sole source of happiness. Similarly, our restless desire to know is also an implicit longing for God, the source of all truth. (Wood 2013, 5)

[A]t a deeper level, every sinner does deceive himself. He deceives himself about the fact that he implicitly recognizes that God is the highest good and the true source of value. He also deceives himself about whether he is deeply and fundamentally satisfied even though he is estranged from God. (Wood 2013, 11)

If God is truth, it follows that to turn away from the truth is to turn away from God, which is the very paradigm of a sinful action. It is therefore not surprising that the structure of self-deception should mirror the structure of sin. (Wood 2013, 12)

Wood's suggestion, then, might be read as saying that people's attitudes towards the source of true happiness or towards truth itself be counted as de re attitudes towards God, who is identical with these things. Other things towards which people have attitudes that might be counted as de re attitudes towards God include the cause of the universe, the ground of one's being, the source of one's blessings, the moral lawgiver, the creator of some part of the world, and so on.

The de re move seems preferable to the Molinist move in that the de re move does not require any particularly controversial metaphysics whereas the latter does. Moreover, if it is plausible to suppose that people without a monotheistic God-concept can have attitudes that are de re attitudes towards God, this move offers a way to see all humans as actually having rejected God, rather than merely having done so in some possible worlds.

The main area of vulnerability for the de re move is the question of whether the ascription to a person S of a de re attitude towards God really is plausible, particularly when the X that is the object of S's attitude is conceived of by S as having a quite different set of properties from the properties that God in fact has.⁸ Suppose, for example, that S has an attitude of fear towards the creator of the stars, whom S believes to be a gigantic being with the body of a lizard, the head of an eagle, and a capricious temperament. Could such an attitude credibly be construed as a de re attitude of fear towards God? That seems dubious. More generally, the kinds of beings who have been thought to be responsible for creating the world (or parts thereof) in various hunter-gatherer societies are in some cases so different from the God of monotheism as to put quite a strain on the de re move.

The more promising avenue for the de re move, it seems to me, is to focus on human moral beliefs and attitudes as the locus of de re engagement with God. Robert Adams, among others, has made the case that the concept of moral obligation is implicitly social, which is to say, obligations are always ultimately owed to someone. As Adams puts it,

We do not really have the concepts of moral obligation and guilt until we can make some sort of distinction among the things we do that strain relationships . . . I believe it is not childish, but perceptive and correct, to persist in regarding obligations as a species of social requirement, and guilt as consisting largely in alienation from those who have (appropriately) required of us what we did not do. (Adams 2002, 240)

There are a couple of reasons why a person's engagement with the moral realm might be a more promising place for locating de re attitudes towards God. For one thing, there is a fairly strong consensus among the relevant empirical practitioners that belief in moral realism – the belief that moral obligations are objectively binding – is something like a human universal, and one which is very difficult to suppress.⁹ For another thing, given that what we are presently interested in is trying to make sense of how a person's rejection of God could be due to sinful attitudes on the part of the person – an inherently morally freighted notion – it makes sense to look for the relevant de re attitudes towards God in the area of a person's thought life concerning the moral realm. It seems to me not too much of a stretch to suppose that even a person who has no conscious thoughts about to whom or to what her moral obligations are owed might be viewed as having attitudes de re towards God when, for example, she begrudges her moral obligations. The extent to which this preserves the universality of sin-based accounts will depend on how plausibly widespread the relevant de re attitudes are. At any rate, I suggest that defenders of sin-based accounts are not without resources when it comes to responding to objection A.

Virtuous doubters and vicious believers

Objection B: Many reflective non-believers have been sincere and conscientious in their investigations of theism, or at any rate, no less sincere and conscientious than some of their believing counterparts.

Turning to objection B, the heart of the issue here is that there are some non-believers in the categories of reflective doubters and ex-believers who appear to be at least as open to relationship with God as some believers are (in the sense that they would welcome such a relationship were it available), and whose investigations into the truth of theism appear to have been at least as sincere and conscientious as those of some believers. To reiterate, it is this openness to relationship with God and conscientiousness and sincerity with respect to investigating the evidence concerning God's existence that are principally what I mean when talking about the relevant moral and intellectual virtues of character. Other areas of a person's moral and intellectual character will often be germane to the question of whether that person really is open, sincere, and conscientious as regards the God question. For instance, a person who is flagrantly abusive towards others while claiming to want relationship with God might well be thought to be actually insincere. Conversely, someone who gives herself to a life of serving others and who claims to have an unfulfilled longing for relationship with God might very well be deemed to be sincere as regards her claimed desire for relationship with God in virtue of the total picture we have of her life.

Objection B has the following form, then, where ‘N’ denotes the property of being a non-believer regarding God, ‘T’ denotes the property of being a believer in God, and ‘V’ denotes the relevant moral and epistemic virtues of character:

1. (1) Some Ns possess V to at least the degree that some Ts do.

2. (2) But sin-based accounts of non-belief predict that any N possesses V to a lesser degree than any T does.

3. (3) Therefore, sin-based accounts are strongly disconfirmed.

As always, the options for the defender of an allegedly empirically disconfirmed theory are either to dispute the empirical claim or to argue that the theory doesn't actually predict what its detractors claim it does. The first option – rejecting (1) – seems like a tough bullet to bite. If one considers the most intellectually conscientious and morally upstanding non-believer one knows, and then asks the question of whether this person exhibits the relevant openness, sincerity, and intellectual integrity to a lesser or greater degree than, say, a well-heeled televangelist who fleeces vulnerable people for money, the answer is blindingly obvious. But, turning to the second option, do sin-based accounts of non-belief really predict that the least virtuous theist would always be more virtuous in the relevant ways than the most virtuous non-theist? It seems not.

The Augustinian tradition from which sin-based accounts flow, and the New Testament picture of human beings on which the former purports to be based, doesn't view propositional assent to the existence of God as the heart of the matter. Or, put another way, mere propositional belief that God exists and has specific attributes is woefully insufficient for standing in right relation to God. Jesus is reported as saying that, ‘Not everyone who calls me Lord will enter God's kingdom. The only people who will enter are those who do what my Father in heaven wants’ (Matthew 7:21). The Epistle of James remarks that, ‘You believe that God is one; you do well. Even the demons believe – and shudder’ (James 2:19). Passages like these strongly suggest that something else is much more important than the mere belief that there is a God with such-and-such attributes: namely, having a certain set of dispositions to act in certain ways as well as displaying certain affective attitudes such as love and trust towards God.

Let us use the term saving faith to denote the cluster of cognitive states, affective attitudes, and dispositions to action that are involved in being rightly related to God. Let's suppose for now that propositional belief that God exists is included in this cluster (we will shortly consider how things look if that assumption is dropped). The thought, then, is that by no means everyone who possesses the belief that God exists – let alone everyone who professes belief that God exists – has genuine saving faith; propositional belief that God exists is not sufficient for having saving faith. We can call this the insufficiency move. Where a believer is someone who has propositional belief that God exists (and perhaps also professes that belief), and a non-believer is someone who lacks the propositional belief that God exists, the thought is that the Augustinian tradition is not committed to the empirical prediction that any non-believer possesses the relevant moral and epistemic virtues to a lesser degree than any believer. It doesn't predict that the most virtuous non-believer you know is less virtuous than the oily televangelist.

Despite the relief that is available to defenders of sin-based accounts in light of making the insufficiency move, there are yet further challenges. If, for every non-believer who appears morally and epistemically on a par with or superior to some believer, the believer in question is thereby deemed to lack genuine saving faith, then the circle of those regarded as possessing genuine saving faith will surely end up very small indeed. To put the point another way, the more the gap between mere propositional belief and saving faith is exploited in order to deal with the hardest cases of apparent virtue parity between non-believer and believer, the harder it will be for anyone to count as having genuine saving faith. What's more, in light of the fact that the view we are currently considering remains committed to the necessity of propositional theistic belief for saving faith, individuals who have suffered severe religious trauma that has left them psychologically incapable of believing that God exists or is worthy of worship will be deemed culpable for their non-belief. This is an especially implausible consequence of the view, given the growing body of empirical data indicating that the far-reaching psychological effects of religious trauma are typically not under the control of the sufferer (see Panchuk 2018).

One option is simply to double down: affirm that the degree of relevant moral and epistemic virtue displayed by the world's most conscientious non-believer is not sufficient for possessing genuine saving faith; thus, all believers with an equal or lesser degree of the relevant moral and epistemic virtues will be deemed not to have genuine saving faith, however things might appear. How big a bullet this seems to be depends on how relevantly virtuous one thinks the most virtuous non-believer is.

The other option is to suppose that propositional belief that God exists is not only insufficient for saving faith, but also not necessary. We can call this the non-necessity move. While this might sound like a surprising move, many philosophical accounts of the nature of faith on offer in the contemporary literature maintain that S's level of intellectual confidence that p is true need not rise to the level of belief in order for S to have faith that p.¹⁰ Such non-doxastic accounts of faith, as they are called, are perhaps the dominant variety today. The gist of the arguments offered for the non-doxastic view of faith is that unless we suppose S can have faith that p despite having a level of intellectual confidence below the threshold for belief that p, it is hard to make sense of various cases in which S perseveres in acting in certain ways in spite of considerable uncertainty about p. Daniel Howard-Snyder, for example, offers the case of Yehuda, who becomes unsure about whether there is a God but nevertheless continues to pray and study Torah. Howard-Snyder contends that it is plausible to think that Yehuda still has faith that God exists, despite falling short of having the belief that God exists (Howard-Snyder 2013, 365). Presumably the only way that adopting a non-doxastic account of saving faith would help in responding to objection B is if such a move enables the defender of sin-based accounts to deem the harder cases of apparent virtue parity between believer and non-believer to be cases where both the believer and the non-believer have saving faith. In this way, the circle of those who have saving faith need not be drawn so very tightly.

The non-necessity move is not without its drawbacks for defenders of sin-based accounts. What is required for this move to work is not merely a non-doxastic account of saving faith, but a fairly radical such account: one that allows that saving faith is compatible not only with a level of intellectual confidence that falls somewhat below the threshold for belief that God exists, but also is compatible with outright disbelief. (Keep in mind that objection B concerns reflective non-believers and ex-believers.) This worry could be mitigated by softening the move so that it only applies to those non-believers who hope that God exists, where hope is a cognitive attitude (and plausibly also an affective attitude) that is incompatible with outright disbelief that p. This version of the move would be somewhat more plausible than claiming that outright disbelievers can have faith in God.¹¹ The cost of softening the non-necessity move in this way, of course, is that it then applies to fewer cases of apparent virtue parity between non-believer and believer.

But there is another significant cost for sin-based accounts besides this, and it is a cost that applies to any version of the non-necessity move. By allowing that there exist people who are equipped with a monotheistic God-concept and who lack the belief that such a God exists and who nevertheless have saving faith in God, this move constitutes a significant concession on the part of sin-based accounts: namely, it concedes that there are some people whose lack of theistic belief is not explained by a lack of the relevant moral and epistemic virtues of character. After all, if such individuals possess the relevant moral and epistemic virtues to a great enough degree to count as having saving faith, it would make little sense to attribute their lack of propositional theistic belief to a lack of the relevant moral and epistemic virtues. The making of this concession will thus create the need for a distinct, non-sin-based explanation for the non-belief of these individuals. Sin-based accounts that avail themselves of both the insufficiency and the non-necessity moves therefore cannot be universal accounts of non-belief; at most they can be partial accounts.

Sin's spatio-temporal distribution

Objection C: Regarding the moral and epistemic traits of character that sin-based accounts claim are determinative of whether someone believes in God, it isn't plausible that the populations of spatio-temporal regions in which theistic belief is concentrated exhibit virtuous and vicious traits at substantially greater and lesser frequencies (respectively) than the populations of spatio-temporal regions in which theistic belief is sparse.

This final objection that we shall consider in essence demands of the sin-based explainer of non-belief: ‘Are you really saying that some entire regions of the globe and some entire eras of history are more sinful than others?’ The objector, in other words, supposes that the sin-based explainer of non-belief is committed to thinking that the explanation for the uneven spatio-temporal distribution of theistic belief is that sin is similarly unevenly spatio-temporally distributed.

It is important to see that there is more than one sense in which sin might be unevenly distributed throughout human populations. One way would be for the inborn tendency to sin – that is, original sin – to be exemplified to a greater degree in some human populations than others. I take it that an orthodox Christian cannot embrace such an idea. Another way would be for it to be the case that some cultures have developed social structures – including legal frameworks, economic systems, norms of etiquette, habits of thought and behaviour, patterns of devotion, and what James K. A. Smith (2009) has called ‘cultural liturgies’ – that are especially hostile to the things that God values. This is the notion of social sin. Kathryn Pogin offers the following apt characterization of social sin:

Our social structures and epistemic resources share a mutually dynamic relationship. That relationship may be fruitful, as when community enables knowledge-sharing and discovery, or inhibiting, as when polarization, oppression, or widespread prejudice, disorients us from the truth. Where our social environments are marred by systems of domination and oppression, and where our epistemic resources develop within that same socio-political order, there is a risk that our very abilities to recognize and will good may be nearly inescapably compromised. (Pogin 2021, 160)

To be clear, Pogin herself doesn't seek to apply the concept of social sin to the issue of non-belief. The thought we are considering here, though, is that the lack of monotheistic belief in a given culture might be explained in terms of that culture's having collectively strayed especially far from alignment with the Good, which is identical with God.

A couple of observations are worth making regarding this proposal. The first is that unless defenders of sin-based accounts are prepared to argue that social sin is fully reducible to individual sin, this proposal won't yield a picture on which every individual non-believer's non-belief is attributable to their own individual sin. Another point is that the proposal under consideration seems to imply that, in general, societies that manifest the greatest degree of social sin will also be the least monotheistic. But there are some examples of societies that don't fit this pattern – Nazi Germany comes readily to mind.¹²

As with objection A, another option here is to make a de re move. The aim of the de re move is different here, however. We can call this the second de re move. With regard to objection A, the first de re move aimed to explain how people who lack a monotheistic God-concept could nevertheless be said to have rejected God. With regard to objection C, the second de re move aims to show how it is possible for a person living in a culture that lacks a monotheistic concept of God (or in which belief in such a God isn't a culturally live option) to stand in right relation to God, or in other words, to have saving faith in God.¹³ The idea, then, is that even if monotheistic belief is somewhat unevenly spatio-temporally distributed, saving faith in God may well be far more evenly distributed; hence, sin-based accounts need not postulate unevenness in the distribution of the relevant moral and epistemic character traits.

The view that saving faith in God could be had by people who don't consciously realize that their faith is in God has had a number of advocates in recent decades.¹⁴ Some have suggested that one finds hints of this idea in Jesus's parable in Matthew 25:31–36, where, in response to people who at the last judgment are surprised to learn that they blessed the king during their earthly lives, the king tells them that, ‘Just as you did it to one of the least of these who are members of my family, you did it to me’ (Matthew 25:40). As with the non-necessity move that was considered with respect to objection B, the second de re move does have the upshot that a distinct, non-(individual)-sin-based explanation will be required to explain the lack of propositional monotheistic belief in these people who are deemed to have de re saving faith in God. Again, if such individuals possess the relevant moral and epistemic virtues to a great enough degree to count as having saving faith, it can hardly make sense then to attribute their lack of propositional monotheistic belief to a lack of the relevant moral and epistemic virtues. Hence, making the second de re move has the result that a sin-based account of non-belief can at best be partial in scope, not universal.

Conclusion

While sin-based responses to divine hiddenness arguments are a road less travelled, they do nonetheless have a number of defenders in the contemporary divine hiddenness literature. We have seen that there are a variety of strategies that have been employed to attempt to motivate such accounts. What none of these strategies anticipate or seemingly have the resources to cope with, however, is the set of data that underlies objections A, B, and C. In this article I have sought to assess whether sin-based accounts can cope with these data.

I have suggested that with respect to objection A, the first de re move may offer a way to view at least some pre-axial people as culpably resistant to God despite their lack of a robustly monotheistic concept of God. The extent to which this move preserves the universality of sin-based accounts will depend on how plausibly widespread the relevant de re attitudes are. I have argued that objections B and C plausibly require the making of moves – specifically, the non-necessity move and the second de re move – which involve giving up on the general claim that every non-believer's lack of propositional monotheistic belief is attributable to the sin of the individual. In effect, the upshot of those moves is to reframe the discussion such that the datum to be explained by sin is not a person's lack of explicit propositional belief that God exists, but rather, the failure of some individuals to exhibit the cluster of affective attitudes and dispositions to action that are held to be constitutive of saving faith.

According to my assessment, then, universal sin-based accounts, which is to say, accounts according to which every non-believer's lack of propositional monotheistic belief is the result of sinful resistance by the individual, are not viable in light of the data that underlie objections B and C. Recall that Schellenberg's ADH only needs there to be one non-resistant non-believer in the entirety of human history, given the truth of the conceptual premises of that argument (Premises 1 and 2). If my assessment is correct, then offering a response to Schellenberg's conceptual premises remains vital for anyone wishing to deny the soundness of that argument. Refuting Schellenberg's conceptual premises is far from the end of the matter, however. Just as deductive arguments from evil have given way to probabilistic arguments from evil, showing that a loving God might permit some non-resistant non-belief naturally leads us to move on to the question of how much? Partial sin-based accounts, which I have suggested may remain viable in light of objections A, B, and C, could make a worthwhile contribution to the discussion around probabilistic hiddenness arguments insofar as they are able to make it plausible that some non-negligible portion of the non-belief that we initially thought was non-resistant is actually due to sinful resistance.