The nature of rationality

Some (inevitably perfunctory) comments on the nature of irrationality will be useful at the outset. ‘Irrational’, in its customary sense, means lacking good reasons for one's beliefs, demands, actions, etc., or as lacking sufficient evidence for one's beliefs, demands, actions, etc. The term is not used only in this sense, of course, but this is a necessary and minimal condition for its meaning: if one is irrational, then one lacks good reasons for one's beliefs, demands, actions, etc.Footnote ² The ‘good’ and ‘sufficient’ qualifiers are important, since one may have a reason for one's belief (demand, action, etc.) or evidence for one's belief and yet still have an irrational belief, viz. if the reason for it is bad or evidence insufficient.Footnote ³ If, for example, the only evidence I have for my belief that there's an invisible dragon in my garage is that I sensed its presence there, then, even though I have evidence for my belief, the belief is irrational because the evidence for it is insufficient.

This gloss on irrationality is a gloss on what is variously referred to as theoretical, evidential, or epistemic irrationality. One can lack good reasons for one's belief (demand, action, etc.) and still have a rational belief, i.e. a prudentially, instrumentally, or practically rational belief. In fact, for many – i.e. Pascalian or Jamesian wagerers – belief in God is a prudentially rational belief. Examples of such beliefs are easily multiplied. To use a hackneyed one, if I believe that I will survive an extremely risky life-saving operation, my belief, despite there being scant evidence for it – or, for that matter, despite there being evidence against it – is prudentially rational. I want to suggest, with an important caveat shortly to be elucidated, that God lacks both theoretical and prudential good reasons for His belief requirement, i.e. for stipulating that only believers will be granted salvation.Footnote ⁴ Below, then, even though I will speak simply of His lacking good reasons for this requirement, it should be understood that I mean that He lacks both theoretical and prudential good reasons for it.

Now, to be sure, rational people can make irrational stipulations (have irrational beliefs, etc.), and we wouldn't, on the basis of a single irrational stipulation, judge a person to be irrational. Someone who mistakes local irrationality for pan-irrationality is not only a shallow observer of human nature, but is guilty of the fallacy of composition: he or she infers from what is true of an aspect of an individual's character to the entire character. But people who make irrational stipulations do so because of weakness of will, akrasia, ignorance, and the like, psychological or epistemic failures to which God, by all accounts, is not vulnerable. Thus, if God makes an irrational stipulation, then He is indeed irrational – and not just to a greater or lesser degree. Any being who makes an irrational stipulation (one that, to all appearances, is permanent), despite possessing all the information relevant to that stipulation, and despite being fully knowledgeable (as we are to assume God is) about the nature of rationality generally, is a being who positively flouts rationality, i.e. is a being who is completely irrational. Although in what follows, then, I will speak of God's irrationality, and will not always explicitly qualify my discussion to indicate that He is completely irrational, such a qualification should be presumed.

The foregoing definition of irrationality is bound to strike some as exceedingly imprecise, but given that ‘irrational’ is, in the main, a multifarious, non-technical term that evolved for everyday purposes, producing a definition in terms of necessary and sufficient conditions would be well nigh impossible. Still, it will be desirable to fix terminology as much as possible. By and large, those guilty of acting irrationally (having irrational beliefs, etc.) have contravened certain standards of reasoning particularly radically or egregiously. As David Wong has put it: ‘The term “irrational” has a particularly strong sense, implying a disorder in reasoning processes, a violation of universally valid principles of reasoning that ought to be apparent to a normal reasoner’ (Wong (Reference Wong2006), 557). Peter Hacker claims, similarly, that the power of reason ‘exhibits irrationality only when it violates fundamental canons of reasoning, or when the agent is impervious to correction’ (Hacker (Reference Hacker2007), 199–200). The case for pan-irrationality will be augmented along the way using roughly these precisifications of ‘irrationality’. God is impervious to correction, at least as regards His belief requirement: this requirement, if we follow some theistic traditions on this, is something about which He is unconditionally intransigent. And, as I will show below, He violates certain other standards of reasoning.

A new evidential atheological argument

In this section I propound a new evidential atheological argument, i.e. a new argument providing good inductive evidence that a traditionally conceived monotheistic deity does not exist. The first premise of the argument is an important principle of some theistic traditions. It may be stated as follows:

• P1 Believing in God is a necessary condition for salvation.Footnote ⁵

This premise, put differently, says that God requires belief for salvation, where this is elliptical for ‘If God exists, God requires belief for salvation’, and ‘requires belief for salvation’ means ‘has intentionally made belief a necessary condition for salvation’. I will henceforth refer to this premise simply as ‘God's belief requirement’. The question of which theistic traditions treat belief as necessary need not detain us unduly, since it is not seriously contestable that some do. Christianity, for example, although it has never treated belief as sufficient for salvation, typically treats it as necessary (which is understandable given the many Biblical injunctions stating that it is necessary – e.g. John, 3.16–19; John, 5.24; Acts, 16.31; Romans, 10.9).

The matter of whether God, if He exists, has decreed that belief is sufficient for salvation need not be pursued here, either. Some theistic traditions presumably treat belief as sufficient, but the present argument does not require that it be sufficient; all it requires is that it be necessary. And, as previously remarked, it is uncontroversial that some theistic traditions treat it as necessary. According to these traditions, one will, as a matter of divine declaration, secure an eschatological reward only if one believes.

The second premise of the argument may be stated as follows:

• P2 God, if He exists, has the power to bring it about that every person believes that He exists.

That P2 is true seems scarcely disputable. I, at least, cannot imagine circumstances that would prohibit God – a being who, conceived along traditional monotheistic lines, is omnipotent – from revealing Himself in a way that brings it about that every person believes that He exists. After all, a human being, given sufficient time, resources, etc., could bring it about that every person (at least every person with the appropriate cognitive capacities) believes that he exists, and so there's no reason to suppose that God would be powerless to do it.Footnote ⁶

But God has plainly not brought it about that every person believes that He exists – countless people are, and countless people have been, unbelievers.Footnote ⁷ That is:

• P3 It is not the case that God has brought it about that every person believes that He exists.

It follows from P2 and P3, however, that God, if He exists, has elected, for whatever reason, to hide. More fully, since God has the power to bring it about that every person believes that He exists, and since He has obviously not brought about this state of affairs, God has elected to hide – i.e. He has elected to refrain from rendering evidence that would make His existence patent to everyone. Thus:

• P4 God, if He exists, has elected to hide.

But, on the basis of P1–P4, we can infer that God is irrational. For any being who, despite electing to hide, stipulates that only those who believe that He exists will be granted salvation, is irrational.

Proof: consider the following case. In a mountain cave lives an extremely powerful and wealthy magician who is especially fond of painting. Given his fondness for painting, and given that he has more money than he could possibly spend (or lose), he pledges that he will give $1,000,000 to any citizen of the town below who develops a sincere fondness for painting. However, he stipulates that, in order to receive the reward, one must, in addition to having a fondness for painting, have a sincere belief that he exists. He sends his assistant to the town below to inform the townspeople of his plan. The townspeople are, of course, intrigued by the idea, and they naturally suppose that engaging in painterly practices will arouse a fondness for painting. They also suppose, naturally enough, that obtaining evidence that the magician exists will engender a belief that he exists; and so, in effort to obtain such evidence, they trudge up the mountain to his cave. Shortly before their arrival, however, the magician makes himself, and all his possessions, undetectable (invisible, weightless, etc.) by any known test. Consequently, the townspeople are unable to obtain evidence that he exists. Still, he stipulates that anyone who wants $1,000,000 must have both a fondness for painting and a belief that he exists. Those who do not believe he exists – including those who, through no fault of their own, fail to acquire the belief – will be denied the reward.

The magician is irrational for (at least) the following reason: despite electing to hide, he stipulates that belief in his existence is a necessary condition for receiving the reward he has offered. More fully, he is irrational because, despite electing to hide, he promises to reward only those who have a fondness for painting and a belief that he exists. It is not irrational for him to stipulate that a necessary condition for receiving the reward is a fondness for painting – we typically don't regard as irrational philanthropists who provide endowments to their favourite arts organizations. And it is not irrational, arguably, for him to want the townspeople to believe that he exists – we can suppose that he expects their belief to ameliorate his persistent loneliness.Footnote ⁸ But it is irrational for the magician to stipulate that belief in his existence is a necessary condition for receiving the reward despite his decision to hide. For suppose a townsperson, Fred, prior to hearing anything about the magician's plan, developed a fondness for painting. Further suppose that, owing to the magician's hiddenness, he is unable to believe, try as he might, that the magician exists. Since Fred's unbelief is not a result of malice or vice, there is, it would seem, no good reason for the magician to deny him the reward. But since he does deny him the reward, the magician is irrational.

The magician's belief requirement is analogous to God's belief requirement, and the moral is therefore clear: God is irrational. He is not irrational for wanting people to believeFootnote ⁹ – at the risk of anthropomorphizing, perhaps He desires copyright recognition for creating the universe; or perhaps He simply desires, eminently reasonably, that people seek a loving relationship with him. But God, by our assumption, rewards only those who believe, and this is inconsistent with His supposed rationality. For there are, and have been, many people whose failure to believe is not a product of intellectual or moral vice; that is, there are, and have been, many inculpable unbelievers.

A person with a virtuous moral character who is unable sincerely to believe, try as he might, that there exists a hidden deity is an inculpable unbeliever. But simply in virtue of being an unbeliever he is denied the reward that deserving believers are given. This isn't rationally defensible. If Fred, an inculpable unbeliever, and Clyde, a believer, each have a virtuous moral character and, as a result, each consistently displays honesty, truthfulness, selflessness, generosity, loyalty, etc., then both deserve salvation. However, the proponent of the view that belief is necessary for salvation is committed to affirming that there is an intellectually plausible rationale for God's granting Clyde, but not Fred, salvation. No such rationale is in the offing.Footnote ¹⁰

To guard against a potential misunderstanding, I should note the distinction between inculpable unbelief and inculpable non-belief. An inculpable unbeliever is one who has been given the opportunity to believe in God but fails to take up the belief, though not because of a lack of intellectual or moral virtue. An inculpable non-believer, on the other hand, is one who does not believe in God because he or she is incapable of sophisticated propositional belief or has never had the opportunity to believe. Inculpable non-believers include infants, the cognitively impaired, and the hundreds of millions of people throughout history who never had the opportunity to acquire the concept of a monotheistic deity. Presumably God wouldn't condemn an inculpable non-believer, at least if the non-believer were innocent or had a virtuous moral character; and, in fact, most theistic traditions have exception clauses for those outside the range of the good news about God. This means that the problem of whether those who do not believe will be condemned, or otherwise denied an eschatological payoff, applies only to inculpable (and, of course, culpable) unbelievers. My claim is that the precept that propositional belief in God is a requirement for salvation for those who have had the opportunity to believe must be rejected, at least in some cases – e.g. in cases where the unbeliever has a virtuous moral character.

It is bound to be protested that anyone who makes a genuine, concerted effort to believe that a proposition, p, is true will, in fact, come to believe that p is true. But this is implausible. Some people are not suitably disposed to believe in hidden deities, and thus cannot believe that such deities exist no matter how hard they try. More generally, belief is not subject to direct voluntary control; it is not possible for a person to believe, by force of will, that p is true. To be sure, it is possible for a person to deceive himself into believing that p is true, but a person who has deceived himself into believing that p is true does not, arguably, actually believe – i.e. have the conviction – that p is true. Quite the reverse: he believes, at some level, that p is (probably) false. Thus, God, in electing to hide, has effectively debarred a large class of people from fulfilling a requirement for salvation – viz. belief – that He Himself has set out.

I am here verging on some difficult, quintessentially empirical questions, questions to which scientific answers are now only forming. But diverging at greater length on the topic of whether belief is subject to direct voluntary control is unnecessary, for even if we assume that it is, it does not follow that those who refrain from believing that there is a hidden monotheistic deity are condemnable (or otherwise blameworthy). Indeed, to suppose otherwise is to presuppose that, despite His hiddenness, God's belief requirement is, or can be, rationally justified. But that is precisely the issue under debate.

It bears emphasizing that God's having good reason to hide (if He has it) does not undercut this argument. Some theists claim that God values non-coerced belief, and since the manifest existence of God would unfailingly prompt belief, God has good reason to hide. Fair enough: let it be granted that God has good reason to hide.Footnote ¹¹ Even so, He's irrational because, despite electing to hide, He requires belief (by our assumption). It's the following conjunction, if true, that renders God irrational, not His decision to hide: God has elected to hide (for ex hypothesi good reason) and God requires belief.Footnote ¹²

Call the argument that theism founders on God's irrationality the irrationality argument for atheism, or IAA. A full formulation of the argument would go as follows, where T-assumptions are theistic assumptions:

• P1 God has intentionally made belief a necessary condition for salvation – i.e., God requires belief.Footnote ¹³ [T-assumption.]

• P2 God has the power to bring it about that every person believes that He exists. [Def. of omnipotence.]

• P3 It is not the case that God has brought it about that every person believes that He exists.

Therefore,

• P4 God has elected to hide. [(2), (3).]

Therefore,

• P5 God has elected to hide and God requires belief. [Conj. – (1), (4).]

• P6 God's belief requirement is rational only if God has good reason for His belief requirement. [Def. of rationality.]

• P7 God has good reason for His belief requirement only if He has good reason for requiring belief despite electing to hide.

• P8 It is not the case that God has good reason for requiring belief despite electing to hide.

Therefore,

• P9 It is not the case that God has good reason for His belief requirement. [MT – (7), (8).]

Therefore,

• P10 It is not the case that God's belief requirement is rational. [MT – (6), (9).]

• P11 It is not the case that God's belief requirement is rational only if it is not the case that God is rational.

Therefore,

• P12 It is not the case that God is rational. [MP – (10), (11).]

• P13 God is essentially omniscient. [T-assumption.]

• P14 God is essentially omniscient only if God is rational. [Def. of omniscience.]

• P15 It is not the case that God is omniscient. [MT – (12), (14).]

• P16 God exists only if God is omniscient. [(13).]

• C It is not the case that God exists. [MT – (15), (16).]

As we have seen, the nub of the difficulty for theism is that God requires belief despite electing to hide, and so a crucial premise is P8. To show that God has good reason for requiring belief despite electing to hide, and hence that P8 is false, the proponent of the view that belief is necessary for salvation must provide an intellectually plausible rationale for God's granting salvation only to believers. As I mentioned, however, no such rationale is in the offing.

Is it rational for God both to hide and require belief?

We can, I believe, promptly reject the suggestion that God, despite electing to hide, is justified in requiring belief because only belief will ensure that one be genuinely moral. Belief is neither necessary nor sufficient for morality: many unbelievers are highly moral, while many believers are highly immoral. I cannot, within the ambit of this paper, provide any kind of responsible discussion of the complex dialectic surrounding the nature of moral duty, but it seems to me a foregone conclusion that neither our negative duties (e.g. the duty to refrain from taking an innocent life) nor our positive duties (e.g. the duty to help someone in need) are grounded in divine commandments. As Plato argued centuries ago, if our duties are rooted in divine command – i.e. if God is literally the source or standard of our moral duties – then God's judgements regarding their nature aren't really judgements at all, but rather capricious stipulations. On the other hand, if God's judgements regarding their nature are not arbitrary, then they are based on something, which means that there must be something intrinsic to those duties, something antecedent to God's judgement, that makes them our duties. Either way, moral duties, whatever they consist in, are not the product of divine fiat.

Perhaps a better way for the proponent of the view that belief is necessary for salvation to reply is to maintain, à la sceptical theism, that God, despite electing to hide, does have theoretical or prudential good reasons for His belief requirement; only, given our inherent cognitive limitations, these reasons are beyond our ken. Alvin Plantinga, among others,Footnote ¹⁴ has argued that the claim that there is no good reason for God to do x implies that if there were a good reason for God to do x, we would be able to discern that reason, which is absurd. On Plantinga's view, there is an ‘epistemic distance between us and God’ (Plantinga (Reference Plantinga2000), 466–467), such that it's plausible to suppose that in many (perhaps most) cases we will be unable to discern God's good reasons for doing x. It might be claimed that considerations such as these ground an objection to IAA: the argument proceeds by tacitly assuming that we would be able to discern God's good reasons for requiring belief, but since this assumption is false, the argument fails.

A number of objections to this kind of argumentative strategy – at least as the strategy is pressed into service to rebut the problem of suffering – can be found in the literature,Footnote ¹⁵ but retooling and redeploying them here is unnecessary. There is a more effective reply. In saying that God, despite electing to hide, has good reason for requiring belief, our objector implies that God is rational. So our objector claims, in effect, that the charge of irrationality is unwarranted: God is rational; only, given that we are epistemically impoverished relative to Him, we are unable to discern His good reasons for requiring belief. My reply, however, is that this objection actually bolsters the charge of irrationality. For if, as our objector insists, God is rational, then we can ascribe to Him the belief that the following principle of rational etiquette – call it the principle of full disclosure, or PFD – is true:

• PFD In rational, non-confrontational, non-pedagogical discourses, investigations, inquiries, etc., it is always the case that, if possible,Footnote ¹⁶ rational beings should disclose their good reasons for doing x (believing x, etc.) to other rational beings seeking disclosure of their good reasons for doing x.Footnote ¹⁷

But since, by our assumption, it is not the case that we can ascribe to God the belief that PFD is true – He has not disclosed, and apparently will not disclose, His reasons for requiring belief – God is irrational.

Notice that it won't do for the proponent of the view that belief is necessary for salvation to attempt to parry this argument by claiming, as before, that God does have good reasons for rejecting PFD; only, given our inherent cognitive limitations, we are unable to discern what those reasons are. While it is true that we are free to extend the notion of rationality beyond its normal (admittedly sometimes vague) boundaries for specialized purposes, it is unacceptable to stretch it to a point at which it becomes legitimate to say that PFD is rationally rejectable. This might seem a terminological quibble, but it isn't.

In trying to uphold God's rationality, our objector attributes a property to God denoted by the term ‘rational’, but there is general agreement – in the philosophical and non-philosophical literature – regarding the meaning of this term, even if its precise meaning is occasionally unclear. Indeed, such general agreement is a necessary condition for a discussion of rationality to get off the ground in the first place. But on anyone's understanding of the meaning of the term ‘rational’ (and its cognates), PFD is not rationally rejectable. It seems to be part of the very concept of rationality that if we are secretive and unforthcoming in our non-confrontational and non-pedagogical engagements with one another, then we have failed to satisfy the demands rationality imposes on us: rationality demands openness. Otherwise put, to insist that PFD – and presumably any other principle of rationality one might adduce – is rationally rejectable (by God anyway) is to speak not of rationality, but of something else, perhaps of rationality*. But, of course, we are interested in rationality, not rationality*; and the charge here is that God is irrational, not irrational*.

The proponent of the view that belief is necessary for salvation ought to be stirred by these considerations, but there is an additional argument showing that trying to uphold God's rationality will just end up bolstering the charge of irrationality. As before, let us suppose, along with our objector, that God is rational. If He is rational, then we can ascribe to Him the belief that the following principle – call it the normative principle of rationality, or NPR – is true:

• NPR For any person A and any proposition p, A ought to believe that p, or engage in practices that will induce the belief that p, only if A is aware of either theoretical or prudential good reasons for believing that p.

Any rational being will believe (or at least ought to believe) that NPR is true – indeed, the principle seems all but tautological – and since, by our assumption, God is rational, He believes, or at least ought to believe, that NPR is true. But does He? Not obviously, for it would appear that we can ascribe to Him the belief that the negation of NPR is true:

• ~NPR It is not the case that for any person A and any proposition p, A ought to believe that p, or engage in practices that will induce the belief that p, only if A is aware of either theoretical or prudential good reasons for believing that p.

The belief that ~NPR is true can be ascribed to God because His stipulation that only believers will be granted salvation presumably applies to all people: it is meant to influence even those unfavourably disposed to the notion that there are theoretical or prudential good reasons for pursuing belief in God – e.g. those who consider Him to be completely irrational. But this is just to say that God believes that ~NPR is true, that sometimes a person should believe a proposition (e.g. that God exists) even if the person is not aware of any theoretical or prudential good reasons for believing it.

The proponent of the view that belief is necessary for salvation, then, is at risk of being impaled on the horns of a dilemma: either God believes both NPR and ~NPR, in which case He has contradictory beliefs and is thereby irrational – a being who knowingly harbours contradictory beliefs is uncontroversially irrational – or He believes that ~NPR is true, in which case He is irrational.

Of the two possible ways to avert the dilemma, the least plausible is to deny that a rational being must accept the truth of NPR. NPR does not seem rationally rejectable. There may be either theoretical or prudential good reasons – ‘external’ to A – that A ought to believe p, but if A is not aware of them, then such reasons cannot motivate A to believe p. But if such reasons cannot motivate A to believe p, it makes no sense to say that, nevertheless, A ought to believe p for those reasons. Rejecting NPR, then, commits one to rejecting the time-honoured and theoretically important principle that ‘ought’ implies ‘can’.Footnote ¹⁸ Thus, even though there is nothing manifestly amiss about the notion of external reasons – i.e. reasons that are independent of our beliefs, desires, motivations, etc. – such reasons, whatever they are, cannot motivate, and so it is misconceived to suggest that they have the normative punch required for rational belief formation.

The second way to attempt to avert the dilemma is to claim that God believes that NPR, but not ~NPR, is true. This strategy is equally problematic, however, because it forces the proponent of the view that belief is necessary for salvation into the unstable position of having to maintain that arguments for God's existence are relevant only to those who do not consider God to be irrational. The reason for this claim is that rational beings who have reflected on His hiddenness and His belief requirement and have thereby arrived at the conclusion that He is irrational will lack the motivation requisite for taking arguments for His existence seriously, since they will lack theoretical and prudential good reasons (in fact, they will think there are no such reasons) for believing that such arguments prove what they purport to prove. But if they lack the motivation requisite for taking such arguments seriously, they cannot take such arguments seriously, and so it makes no sense to say that they ought to take such arguments seriously.

This is a significant problem for the person who subscribes to the view that belief is necessary for salvation, for it means that she can employ arguments for God's existence only against those who haven't determined that God is irrational. But it would be grossly irresponsible – indeed, immoral – to employ arguments for God's existence against (perhaps excessively credulous) people in full knowledge of the fact that God is irrational.

It seems, then, that the dilemma cannot be averted without jeopardizing theism in ways that jeopardize it worse than the dilemma itself, and so the only recourse for the theist is to yield to the dilemma, both horns of which indicate that God is irrational.