Deontological sceptical theism proved

Abstract

In this article, I argue that sceptical theists have too narrow a focus: they consider only God's axiological reasons, ignoring any non-axiological reasons he may have. But this is a mistake: predicting how God will act requires knowing about his reasons in general, and this requires knowing about both God's axiological and non-axiological reasons. In light of this, I construct and defend a kind of sceptical theism – Deontological Sceptical Theism – that encompasses all of God's reasons, and briefly illustrate how it renders irrelevant certain charges of excessive sceptical and how it evaporates equiprobability objections. Furthermore, I put forth a simple argument in favour of Deontological Sceptical Theism, which shows that everyone (at least currently) ought to endorse it.

Introduction

Arguments from evil against the existence of God claim that some known fact about evil is (at least some) evidence that God – an omnipotent, omniscient, and perfectly good being – doesn't exist. Sceptical theism is a popular response to such arguments: it's thought that sceptical theism undermines many arguments from evil, rendering the most common (and perhaps most powerful) arguments for atheism without bite. However, sceptical theists have traditionally had too narrow a focus: they have zoomed in on God's axiological reasons and have ignored non-axiological reasons that God may have. Their focus is too narrow because predicting how God will act requires knowing more than God's axiological reasons. Instead, it requires knowing the weight of reasons God has in general, and this includes both his axiological and non-axiological reasons. And so when considering how likely some fact (evil or otherwise) is given theism, we need to consider God's reasons simpliciter, not just his axiological reasons. In light of this, I propose what I call Deontological Sceptical Theism, defend an argument in its favour, and illustrate some of its upshot.

This article proceeds as follows. First, I provide a brief overview of a (or perhaps the most) popular form of sceptical theism, and show that it's an axiological thesis. After this, I provide an explanation of what I call Deontological Sceptical Theism, offer an argument for its truth, and illustrate some of its upshot. I finish by considering three objections to Deontological Sceptical Theism: (i) it entails moral scepticism, (ii) it undermines natural theology, and (iii) it's undermined by equiprobability principles. I show that all of these objections fail.

Axiological sceptical theism

Arguments from evil often claim that if God exists, he wouldn't permit an evil unless it's required for a greater good. However – these arguments go – there are evils for which we recognize no greater good. And since we recognize no greater good produced by these evils, we may infer that there probably is no such good. Therefore, God probably doesn't exist (e.g. Rowe (1979) and (1996), and Ekstrom (2021)).

Sceptical theism, if true, is thought to undermine arguments from evil, such as the one mentioned immediately above. There are different kinds of sceptical theism (e.g. Cullison (2014), Howard-Snyder (2009), Dougherty and Pruss (2014), and Wykstra (1984)), but what unites sceptical theists is their rejection of inferences from our lack of recognition of a greater good connected to some instance of evil to the conclusion that there (at least probably) is no such greater good. While there are different kinds of sceptical theism, the kind that I'll be outlining in this section is Bergmann's (2001, 2009, 2012, 2014) – the apparently most popular kind.¹ What does Bergmann's sceptical theism amount to? It amounts to an affirmation of the following sceptical theses:²

ST1: We have no good reason for thinking that the possible goods we know of are representative of the possible goods there are.

ST2: We have no good reason for thinking that the possible evils we know of are representative of the possible evils there are.

ST3: We have no good reason for thinking that the entailment relations we know of between possible goods and the permission of possible evils are representative of the entailment relations there are between possible goods and the permission of possible evils.

ST4: We have no good reason for thinking that the total moral value or disvalue we perceive in certain complex states of affairs accurately reflects the total moral value or disvalue they really have. (Bergmann (2012), 11–12)

The first thing to notice about these theses is that they are axiological: they are about value (goods), disvalue (evils), and their connections. Indeed, in Hendricks (2019) I suggest that ST1, ST2, and ST3 can be stated in the following axiological way:

We have no good reason for thinking that the goods and evils that we know are connected to some instance of evil are representative, [with] respect to [value], of the actual goods and evils that are connected to said instance of evil. (Hendricks (2019), 116)

So, Bergmann's sceptical theism – at least on one natural interpretation – is axiological in nature. As such, I will call this Axiological Sceptical Theism.

There is, I think, good reason to think that Axiological Sceptical Theism is true (e.g. Hendricks (2020a) and (forthcoming), Hudson (2006) and (2014)). Moreover, there's good reason to think that Axiological Sceptical Theism, if true, undermines many important arguments against theism that make use of axiological claims. For example, Rowe's (1996) argument from evil (and many others) relies on an axiological assumption, namely, that there are probably evils that aren't (weren't and won't be) required to bring about greater goods or prevent equally bad or worse evils. Law's (2010) evil-god challenge makes the axiological assumption that there is (in some sense) just too much good in the world for an evil-god to exist, and argues that this is a threat to traditional theism. Moreover, Draper's (1989, 2013) Humean argument from evil – according to him – depends on the assumption that the prima facie good, evil, and neutral states of affairs we recognize are correlated with ultima facie good, evil, and neutral states of affairs. And the common-sense problem of evil (e.g. Dougherty (2008) and (2014)) claims that certain evils seem, in the phenomenal conservative sense, like they aren't required for the production of outweighing goods or the prevention of equally bad or worse evils. Finally, the argument from divine hiddenness depends on claims about what the greatest good for finite creatures is and how to bring about this good (e.g. Schellenberg (2015)).

All of these arguments are vulnerable to Axiological Sceptical Theism: Bergmann (2001) and Draper (2013) argue that Rowe's evidential argument from evil (and others like it) succumb to Axiological Sceptical Theism; Hendricks (2018b) argues that the evil-god challenge succumbs to Axiological Sceptical Theism; Draper (2013) argues that if something like Axiological Sceptical Theism is true, then his argument from evil is in trouble;³ Bergmann (2012) provides an error theory for why some might find the common-sense problem of evil compelling, and argues that Axiological Sceptical Theism is still relevant;⁴ and Bergmann (2009) and Hendricks (forthcoming) claim that the argument from divine hiddenness is undercut by Axiological Sceptical Theism. My purpose here is not to defend any of these claims or arguments. Instead, I claim only that Axiological Sceptical Theism is thought by many to have significant implications with respect to the above arguments against theism. However, it's worth noting that almost no one has argued that Axiological Sceptical Theism, if true, doesn't have these effects. The typical (nearly ubiquitous) procedure is to argue that Axiological Sceptical Theism is false ⁵ – not that arguments from evil (etc.) are immune to it.⁶ So Axiological Sceptical Theism is an important thesis: if true, it appears to undermine a significant number of arguments against theism.

Deontological sceptical theism

While Axiological Sceptical Theism is important and (at least arguably) undermines numerous arguments against theism, its focus is too narrow: its exclusive focus on (dis)value (i.e. axiology) is a mistake. This is because in order to predict how God will act, we need to know the weight of his reasons simpliciter for (and against) an action – not just his axiological reasons and their weight. Furthermore, some arguments from evil are explicitly stated deontologically. For example, Tooley's equiprobability argument from evil (e.g. Tooley (2019) and (2021), and Plantinga and Tooley (2008)) focuses on rightmaking and wrongmaking properties, and claims that the recognized wrongmaking properties of some evil E make it likely that E is all things considered impermissible. And Mooney (2017, forthcoming) has explored deontological formulations of the argument from evil that make use of side-constraints – non-axiological and non-consequentialist constraints on God's actions.⁷

More generally, those who reject an axiological consequentialist ethic won't be appeased by the scepticism of Axiological Sceptical Theism: they might hold that there are non-axiological side-constraints on God's actions, such that even if an action of God's produced a greater good or avoided an equally bad or worse evil, it would be impermissible. Or they may hold that God has non-axiological reasons for or against actions, meaning that we have to take them (non-axiological reasons) into account when considering how likely God is to allow some event (evil or otherwise). So, focusing exclusively on axiology is a mistake: it doesn't address important arguments from evil and, crucially, it doesn't tell us enough to predict how God will or would be likely to act.

For the purposes of this section, we may understand a reason for (or against) an action A to be a consideration in favour of (or against) A. (I will go into more detail below about the nature of reasons.) As a rational being, God will do whatever he has the most reason to do. However, value is just one reason God has for action: if we know the value of an action A, that doesn't necessarily tell us whether or how likely it is that God will perform A, since there could be non-axiological reasons for (or against) A. For example, suppose that I promise to meet Sarah for lunch tomorrow, and suppose that this would produce 1000 units of good. After making this promise, suppose that Sally invites me to play spikeball (unbeknownst to her) during the time I'm supposed to have lunch with Sarah, and suppose that playing spikeball with Sally would produce 1000.1 units of good. Even though playing spikeball with Sally produces more good – more value – I still ought to have lunch with Sarah: my reasons still favour having lunch with Sarah. And this is true even though it's an axiologically worse action.⁸

This is why it's important that we don't focus simply on the (dis)value of actions: they are not the sole determining factor of how God (or humans) likely will or ought to act.⁹ Indeed, if all we know is the (dis)value of permitting some state of affairs, we don't know how likely it is that God will permit it, since we need to know something about the total weight of God's reasons to know that, and that requires knowing something about both God's axiological and non-axiological reasons. And this tells us that Axiological Sceptical Theism has too narrow a focus: it focuses on one kind of reason God has instead of his reasons simpliciter. While I suspect that Axiological Sceptical Theism came to be framed in this way because of popular formulations of evidential arguments from evil (especially Rowe (1979)), now is not the time to point fingers (accusing in print is one thing, but pointing fingers is another). In the section below, I will attempt to ameliorate this issue by providing an account of sceptical theism that isn't too narrow, and showing that we – at least currently – should accept it.

Deontological sceptical theism stated

So much for Axiological Sceptical Theism. Below I will explicate and defend Deontological Sceptical Theism. What distinguishes Deontological Sceptical Theism from Axiological Sceptical Theism? Whereas Axiological Sceptical Theism involves scepticism about value, Deontological Sceptical Theism involves scepticism about the weight of reasons God has in favour of (or against) permitting particular states of affairs. We may state its sceptical component as follows:

Deontological Sceptical Theism: For any evil state of affairs we know of E, we have no good public antecedent reason to think it's likely that the known weight of justifying and requiring reasons God has in favour of (or against) permitting E resembles the actual weight of God's justifying and requiring reasons in favour of (or against) permitting E.

Some explanation is in order here. By ‘public reason’, I mean a reason that is, in principle, equally available to everyone. All other reasons are private. For example, that a sample is random is a public reason for thinking it isn't biased. Or, that everyone agrees about X is a public (weak) reason for thinking X is true. Or, that I found my computer in a dumpster is a public reason for thinking it probably isn't a nice computer. Now, take seemings. Let's say that if it seems that p to S, that p has the appearance of truth for S. For example, it might seem to S that a stick put halfway into a stream is bent. This is a perceptual seeming. Or, it may seem to S that 4 is greater than 2. This is an intellectual seeming. Or, it may seem to S that burning cats is morally wrong. This is a moral seeming. In all these cases, the proposition has the ‘feel of truth’ for S (Tolhurst (1998), 298–299).¹⁰ Given this characterization of seemings, the fact that it seems to S that p won't count as a public reason, since it's more directly available to S than it is to others – seemings are private in the sense that a seeming is not equally available to all. Or, the fact that S finds it intuitive that p won't count as a good public reason, since S's intuition is not equally available to all. More generally, we may say that private reasons are one's own mental states, and that reasons aside from one's own mental states are public reasons.¹¹

By ‘we have no good reason for thinking X’, I mean that we do not currently recognize any reason as being a good one for thinking X. This qualification is important, because there is a sense in which we can have a good reason for thinking X and yet not recognize it as being a good reason for thinking X. For example, a child might see that the car in front has turned on its left turn signal. That's a good reason to think that the car will turn left, but the child doesn't recognize it as a good reason for thinking that the car will turn left – it just looks like a blinking light to the child. And so given our understanding of ‘having good reason’, we may say that, in this case, the child has no good reason to think the car will turn left. It should be clear, then, with respect to Deontological Sceptical Theism, that we're concerned only with good reasons that we recognize as being good reasons.

A good antecedent reason should be understood as a good reason that sets aside our beliefs about God's existence or non-existence. I focus on good antecedent reasons because one's beliefs about God will influence whether one has a good reason to think the known weight of God's justifying and requiring reasons is likely to resemble their actual weight. For example, theists may hold that for any instance of evil that is, given what we know, weighted in favour of impermissibility, we have good reason to think that the known weight of God's justifying and requiring reasons doesn't resemble the actual weight of God's justifying and requiring reasons, since God's existence entails that all such evils are permissible – God wouldn't permit something that's impermissible, after all. So, the focus on antecedent reasons, in effect, rules out Moorean responses to arguments from evil.¹²

By the ‘known weight’ of God's justifying reasons and requiring reasons I mean the recognized weight of all of God's reasons that we recognize. And by the ‘actual weight’ of God's justifying and requiring reasons, I mean the weight of all of God's justifying and requiring reasons – both recognized and unrecognized. Obviously, on this understanding it's possible that the known weight of God's justifying reasons and requiring reasons matches the actual weight of his reasons.

For the sake of simplicity, I will often forgo distinguishing justifying reasons and requiring reasons and just speak of reasons.¹³ For example, I may just speak of God's reasons for performing X. In those cases, ‘reasons’ should be understood to denote both justifying and requiring reasons.

The focus of Deontological Sceptical Theism is on God's reasons as opposed to reasons simpliciter, since reasons for you or me might not be reasons for God, and vice versa. For example, Murphy (2014 and 2017) argues that ensuring human well-being may be a requiring reason for you or I to act but not a requiring reason for God to act.¹⁴ Moreover, as Anderson (2012, 36) points out, God's position as creator makes it such (or if he exists would make it such) that he will have different reasons than his creation. For example, God has a reason (even if very weak) to refrain from interfering with his creation, thereby permitting his creation to bring itself from chaos to order. This is because creation bringing itself from chaos to order is a (even if very weak) good. However, created beings have no such reason, since they are part of creation: their bringing about order is part of creation bringing itself from chaos to order. Or, God has a reason (even if weak) to refrain from constantly interfering with his creation in order to show respect to his creation's autonomy, whereas created beings have no such reason – they are part of creation and so their actions won't interfere with creation being autonomous (to a degree) in relation to God.

Finally, we may say that there are two possible final deontological positions with respect to the weight of God's reasons for permitting a state of affairs: overall impermissible or overall permissible. And if the known weight of God's reasons for permitting E resembles the actual weight of his reasons, then the known position of the weight of God's reasons is the final deontological position. For example, if the known weight of God's reasons for permitting E is overall impermissible, then the known weight of God's reasons resembles the actual weight of his reasons if and only if overall impermissible is the final deontological position with respect to the weight of God's reasons for permitting E. If the final deontological position is overall permissible, then there is no such resemblance.

For the sake of readability, I will shorten ‘we have no good public antecedent reason to think it's likely that the known weight of justifying and requiring reasons God has in favour of (or against) permitting E’ to ‘we have no good public antecedent reason to think it's likely that the perceived weight of God's reasons for the permission of E’. And I will shorten ‘the actual weight of God's justifying and requiring reasons in favour of (or against) permitting E’ to ‘the actual weight of his reasons’. Understood in these terms, we may simplify Deontological Sceptical Theism to the following thesis:

Deontological Sceptical Theism Simplified: For the permission of any evil E, we have no good public antecedent reason to think it's likely that the perceived weight of God's reasons resembles the actual weight of his reasons.

The nature of reasons

Reasons play a central role in Deontological Sceptical Theism. But what are they exactly? Reasons, in general, may be understood to be considerations in favour of, or against, some action (e.g. Scanlon (2000)). There are different kinds of reasons: there are requiring reasons and justifying reasons.¹⁵ When an action has a requiring reason against it, we may say that it has ‘requiring weight’ against it, and how weighty it is depends on how strong the reason is. Similarly, when an action has a justifying reason for it, we may say that it has ‘justifying weight’ for it, and how weighty it is depends on how strong the reason is.

A requiring reason for A is a consideration in favour of the impermissibility of ~A, and also a consideration in favour of the permissibility of A: if A is required of me – if ~A is impermissible – then A is permissible.¹⁶ For example, that my kids are hungry is a requiring reason for my feeding them – it makes it such that, in the absence of a sufficiently weighty justifying reason for not feeding them, it's impermissible for me to not feed them.¹⁷ A justifying reason for A is a consideration in favour of the permissibility of A: in the absence of a sufficiently weighty requiring reason against A, it renders A permissible (but not required). That is, a justifying reason for A just counts in favour of its permissibility – it doesn't count in favour of ~A being impermissible or of A being required. For example, that eating cookies would taste good is a justifying reason for my eating them, but it isn't a requiring reason for me eating them: it pushes my act of eating cookies towards being permissible, but it doesn't push my eating cookies towards being required. Or, suppose that by risking serious harm and running into a burning building you could save a dog from perishing in the flames. That you could save the dog is a justifying reason for risking harm to yourself, but it isn't a requiring reason for risking such harm (Gert (2012), 612). Or suppose that a grenade lands in the middle of a group of soldiers. A soldier, S, could jump on the grenade, saving her fellow soldiers but taking her own life. That jumping on the grenade would take S's own life is a justifying reason for her not jumping on it, but it isn't a requiring reason for her to do so (Tucker (forthcoming a)). And so, for any pair of potential actions A and ~A, there will (potentially) be requiring reasons for A, requiring reasons for ~A, justifying reasons for A, and justifying reasons for ~A.

Whether A is permissible or impermissible is a function of the weight of the requiring reasons against it and the justifying reasons in favour of it: if the justifying reasons for A outweigh the requiring reasons against A, then A is permissible. If not, then not.¹⁸ And, per the above discussion, we can see that for any action A, there will be more kinds of reasons in favour of its permission than against its permission: justifying reasons for A and requiring reasons for A push it towards permissibility, whereas only requiring reasons against A push A towards impermissibility. (Justifying reasons for ~A push ~A towards permissibility, but that doesn't push A towards impermissibility.) In this sense, then, reality is biased towards permissibility – there are more kinds of reasons that count in favour of an action being permissible than count against its permissibility.

Additionally, there are (at least) three other reasons for thinking reality is biased towards permissibility. First, there are default justifying reasons in favour of every action performed by an agent: there's always a justifying reason for performing an act in virtue of the agent's autonomy, for example.¹⁹ Second, a relatively standard (and extremely plausible) view in ethics – one which I'll be assuming for this article – is that there can't be prohibition dilemmas: it can't be both that A is impermissible and that ~A is impermissible. In other words, one will never find oneself in a situation where all actions open to one are impermissible. This suggests a bias towards permissibility since one can't find oneself in a situation where all acts open to one are impermissible.²⁰ And third, the default status of an act is permissible: if there are no requiring reasons against the permissibility of A, then A will be permissible. This is because of the fact that A is only impermissible if there's a requiring reason against it and no sufficiently weighty justifying reason in its favour. A requiring reason against A, therefore, is a necessary condition for A being impermissible. And so if there is no such reason, it will be permissible. This, then, is an explicit bias towards permissibility: an action starts off as permissible until it acquires a requiring reason against it.

In summary, then, reality is biased towards permissibility because (a) there are more kinds of reasons in favour of the permissibility of an action than there are against it, (b) there aren't prohibition dilemmas, (c) there are default justifying reasons in favour of an action's permission, and (d) the default status of an action is permissibility.²¹ Or, at the very least, I will be operating under the assumption that reality is biased towards permissibility – if one wants to reject this thesis to object to my argument, one may. But it will be an uphill battle.

The upshot of deontological sceptical theism

So much for Deontological Sceptical Theism. What's its upshot? Here, I'm going to be quite brief, because I take these implications to be fairly clear. The upshot of Deontological Sceptical Theism is this: it undermines all evidential arguments against theism that rely on the claim that for some fact F, the known or perceived weight of God's reasons against permitting F makes it unlikely that he will permit F or makes it less likely that God would permit it than that F would obtain given a competing hypothesis. This is because, per Deontological Sceptical Theism, we have no good public antecedent reason to think that the perceived weight of God's reasons resembles the actual weight of his reasons. And this means that we have no good public antecedent reason to think that the perceived weight of God's reasons for permitting F resembles the actual weight of his reasons for permitting it. In effect, Deontological Sceptical Theism renders the perceived weight of God's reasons useless. For example, claims that the perceived weight of God's reasons for preventing the Holocaust make it likely that he would prevent the Holocaust and that, therefore, the Holocaust's occurrence is evidence against theism will succumb to Deontological Sceptical Theism. This is because it (Deontological Sceptical Theism) entails that we have no good public antecedent reason to think that the perceived weight of God's reasons for preventing the Holocaust resemble the actual weight of his reasons. But then the perceived weight of God's reasons for preventing the Holocaust doesn't give us good reason to think that God would likely prevent it. In other words, that the known or perceived weight of God's reasons against allowing the Holocaust favour him preventing the Holocaust doesn't make it likely that the actual weight of his reasons favours him preventing it. Deontological Sceptical Theism undermines this sort of move. Indeed, it will undermine any move like this for any fact that is claimed to be evidence against theism. So, Deontological Sceptical Theism renders the perceived weight of God's reasons useless.

There is, of course, a clear way for proponents of these arguments to avoid Deontological Sceptical Theism: they can appeal to private reasons, which Deontological Sceptical Theism doesn't make any claims about. However, this will make these arguments into versions of the common-sense problem of evil (see endnotes 6 and 11), and these arguments will fare as well or poorly as those arguments.²²

Deontological sceptical theism proved

Recall the simplified version of Deontological Sceptical Theism:

Deontological Sceptical Theism Simplified: For the permission of any evil E, we have no good public antecedent reason to think it's likely that the perceived weight of God's reasons resembles the actual weight of his reasons.

In this section, I will present a simple argument for Deontological Sceptical Theism – the Default Argument for Deontological Sceptical Theism.

The Default Argument for Deontological Sceptical Theism is quite simple, and goes like this: Deontological Sceptical Theism is true definitionally until and unless we are given a good public antecedent reason for thinking it's likely that the perceived weight of God's reasons for permitting E resembles the actual weight of his reasons for permitting E. This is because, again, we are understanding ‘having a good reason to think X’ as meaning that we recognize a reason as being a good one for thinking X. Therefore, until and unless we're given such a reason, we should accept Deontological Sceptical Theism – Deontological Sceptical Theism is the default position. And hence we should accept Deontological Sceptical Theism until and unless we're given reason to reject it. I leave it to opponents of Deontological Sceptical Theism to produce such reasons – although, below I will briefly examine three.

Objections to deontological sceptical theism

So, Deontological Sceptical Theism is the default position. That doesn't, however, mean that we should endorse it – the Default Argument for Deontological Sceptical Theism only claims that Deontological Sceptical Theism is the default position, not that it should be one's position after all things are considered. This is because if one has a good reason to think Deontological Sceptical Theism is false, then – even though it's the default position – one should reject it. As such, I will consider (very briefly) three such objections in this section. My discussion here is brief because my aim isn't to produce a novel response to these objections – while that's an article worth writing, it's a different article. Instead, I will simply highlight (i) a way in which Deontological Sceptical Theism isn't vulnerable to some charges of moral scepticism or moral paralysis, (ii) that there is a clear limit to how much scepticism Deontological Sceptical Theism can unleash in the domains of natural theology and theodicy,²³ and (iii) that equiprobability objections to Deontological Sceptical Theism fail.²⁴

Moral scepticism

The first objection I will consider – one that is bound to occur – says that Deontological Sceptical Theism entails an excessive amount of moral scepticism: if we accept it, then we have no grasp of our actual moral reasons and so have no clue how we should act. That is, if the perceived weight of our reasons for or against an action isn't indicative of the actual weight of our reasons, then we can't know how we should act. But we do, in fact, have a grasp of what actions we should (and shouldn't) do and we do have a grasp of the actual weight of our moral reasons. And so we have reason to reject Deontological Sceptical Theism.

What are we to make of this charge of moral scepticism? Not much. This is because, in brief, this objection is misguided: Deontological Sceptical Theism is about God's reasons, not our reasons – that we have no good public antecedent reason to think it's likely that perceived weight of God's reasons for the permission of some state of affairs A resembles the actual weight of his reasons for permitting A doesn't entail that for some other action A* open to us, that we have no good public antecedent reason to think it's likely that the perceived weight of our reasons that in favour of (or against) A* resembles the actual weight of our reasons in favour of (or against) A*. And so Deontological Sceptical Theism doesn't (at least in itself) result in moral scepticism.

Theodicy and natural theology

Another objection that might be made is that Deontological Sceptical Theism prevents us from predicting how God will act – since God will act in line with his reasons and we have no good public antecedent reason to think it's likely that the perceived weight of God's reasons for permitting some state of affairs A resembles the actual weight of his reasons for permitting A, this means that we can't predict whether God will permit A. But this threatens both the practice of theodicy – identifying reasons for why God allows evil – and natural theology – using facts about the natural world or reason to provide evidence for God's existence.

With respect to this issue, I will punt: this is a complex issue that I discuss at length in Hendricks (forthcoming), arguing that sceptical theism is compatible with both theodicy and natural theology. But if I'm wrong and these objections are successful, then so much the worse for natural theology and theodicy: it's (I think) pretty clear that Deontological Sceptical Theism is true, and it's less clear that some theodicy is successful or that some piece of natural theology is. That said, it's worth mentioning from the outset that this objection, if successful, wouldn't undo all natural theology – there's a clear limit on how far this scepticism may bleed. For example, modal cosmological arguments (e.g. Pruss and Rasmussen (2018)), modal ontological arguments (e.g. Plantinga (1974)), the Kalam cosmological argument (e.g. William Lane Craig (2000)), contingency arguments (e.g. Gale and Pruss (1999)), moral arguments (e.g.  Layman (2002)), and so on don't rely on an analysis of God's reasons for action and so are not threatened by Deontological Sceptical Theism.²⁵ And so even if Deontological Sceptical Theism undermines some natural theology, it doesn't do away with it completely or perhaps even mostly – this is a clear limit on the scepticism that can result from Deontological Sceptical Theism in this domain.

Equiprobability objections

Some have objected to sceptical theism by using equiprobability principles (e.g. Draper (1989), (2013), (forthcoming), and Swinburne (1998)). They claim that an unknown good and an unknown evil are equiprobable, and that this either shows that sceptical theism is false or that it loses its bite. This is because if we assume that there's a non-zero probability of there being no unknown good or unknown evil connected to some evil E, then the probability that there's an unknown evil connected to E must be less than 0.5, and so the probability that there's a greater good connected to E must also be less than 0.5.²⁶ This is clearly a problem for sceptical theism: this kind of move makes it such that one may infer from our lack of knowledge of an unknown good connected to E to the conclusion that there probably is no such good (since the probability that there is such a good has to be less than 0.5), which clearly conflicts with the sceptical theist's position.

In my view, this is the most serious kind of objection to sceptical theism on the market, and it's a shame that it hasn't been discussed much.²⁷ But Deontological Sceptical Theism isn't vulnerable to this objection.²⁸ This is because an unknown reason in favour of some evil E and an unknown reason against E aren't equiprobable: it's more likely that there's an unknown reason in favour of the permission of an evil E than that there's an unknown reason against the permission of E. This is due to what we've seen above: reality is biased in favour of permissibility. There are more kinds of reasons that count in favour of the permission of an action than there are against its permission. Moreover – as we also saw above – there are default justifying reasons in favour of all actions, prohibition dilemmas are impossible (meaning that reality favours permissibility), and the default status of an act is permissibility. These facts show that reality is biased in favour of permissibility. And this makes it more likely that there's an unknown reason in favour of permitting E than that there's an unknown reason against E, and hence the equiprobability claim won't hold for Deontological Sceptical Theism, and so equiprobability objections can't get off the ground.

To be clear, I'm not making any claims about how much more probable it is that there are unknown reasons in favour of God permitting some evil than that there are unknown reasons against him permitting it. My point is far more modest: I only claim that the fact that reality is biased towards permissibility breaks any claim of equiprobability about unknown reasons for (or against) God permitting some evil – that there's even a sliver of bias in favour of permissibility is enough to break that claim. And this means that equiprobability objections fail.

Conclusion

In this article, I've shown that (i) Axiological Sceptical Theism, though efficacious, has too narrow a scope: we need to focus on God's reasons simpliciter, not just his axiological reasons; (ii) Deontological Sceptical Theism is the default position: until someone can provide good reason to reject it, it's definitionally true; (iii) Deontological Sceptical Theism doesn't result in moral scepticism or paralysis, nor does it rule out natural theology; and (iv) Deontological Sceptical Theism isn't susceptible to equiprobability objections. So, what's the upshot of this? The upshot is, first, that we should all endorse Deontological Sceptical Theism, and second that all arguments that claim that some fact F about the world is evidence for atheism because the perceived weight of God's reasons make it unlikely he would permit F (or make it less likely that God would permit F than that F would obtain on a competing a hypothesis) fail.