Is theism compatible with pointless non-resistant non-belief?

Abstract

Perry Hendricks (2025) argues that theism is not only compatible with what he calls ‘pointless atheism’ (instances of non-resistant non-belief that do not serve a greater good) but also makes it expected. His case combines the Responsibility Objection (RO) – the view that God permits non-resistant non-belief because it’s required for theists to bear responsibility for bringing others into relationship with God – with a William Hasker-inspired argument concerning motivation and rationality. Hendricks’s core argument can be expressed in two distinct yet interrelated ways: a ‘motivation’ formulation and a ‘rationality’ formulation. I examine each in turn. I argue that, even granting (RO) and the rest of Hendricks’s assumptions, each formulation fails. (RO), together with a few further assumptions to which Hendricks also seems committed, leads to conclusions that undermine rather than support his argument. Thus, we have at least as much reason to reject as to accept his conclusion, and without further clarification and support, his case remains incomplete.

Introduction

In a recent contribution to this journal, Perry Hendricks (2025) argues for the compatibility of theism and what he calls ‘pointless atheism’. By ‘pointless atheism’, Hendricks (2025, 590) means ‘pointless non-resistant non-belief’: any instance of non-resistant non-belief that God, if God exists, antecedently knows it to be certain or probable that he could prevent without forfeiting some greater good or permitting an equally bad or worse state of affairs.¹ The term ‘pointless atheism’ is, however, doubly infelicitous, so I depart from Hendricks’s usage. First, by dropping the qualifier ‘non-resistant’, ‘pointless atheism’ invites readers to forget that it refers to pointless non-resistant non-belief. Second, Hendricks (2025, 595, n. 1) uses ‘atheism’ as a cover term for any non-believer, whether a disbeliever or someone who simply lacks belief that God exists; it is therefore easy to miss that he is not speaking only about atheists. Hendricks’s terminology may therefore lead to needless confusion without any corresponding benefit. Henceforth, I use ‘pointless non-resistant non-belief’ for Hendricks’s ‘pointless atheism’, and ‘non-resistant non-belief’ for his ‘non-resistant atheism’.²

Hendricks assumes ‘two prominent views about evil and divine hiddenness’ (Hendricks 2025, 590) to make his case. The first is the Responsibility Objection to Schellenberg’s (1993, 2007, 2015) Hiddenness Argument. This objection holds that God wants to foster a specific kind of responsibility: namely, theists helping non-believers come to know God (Crummett 2015; Dumsday 2010; Swinburne 2004). More specifically, Hendricks (2025, 593) assumes the following central claim of the Responsibility Objection (RO):

RO: God allows some non-resistant non-belief because it’s required for humans to be responsible for bringing others into relationship with God.

The second is William Hasker’s (1992, 2004, 2008) reasoning on pointless suffering.³ Hasker’s basic claim is that God would not prevent all instances of pointless suffering, as doing so would undermine human morality. The idea is that if people knew every instance of suffering was permitted by God for a greater good, their motivation to prevent harm would significantly weaken. By preventing the suffering, they would know they would be foreclosing the greater good(s) for which that suffering was necessary. Hendricks says he wants to explore the implications of Hasker’s view ‘when it’s paired with the Responsibility Objection’ (Hendricks 2025, 592). He adopts a similar line of reasoning but shifts the focus from morality to proselytisation. Hendricks argues that if God eliminated all pointless non-resistant non-belief, theists’ motivation and/or reason to lead others to belief in God would significantly diminish.

To be clear, Hendricks’s purpose is not to defend the Responsibility Objection (RO) or Hasker’s reasoning, but to assume their success to ‘see what follows’ (Hendricks 2025, 591). Together, he thinks these views lead to the conclusion that theism is compatible with pointless non-resistant non-belief.⁴ In other words, Hendricks offers a Conditional Argument (CA):

CA: If (RO) and (something like) Hasker’s view are correct, then theism and pointless non-resistant non-belief are compatible.

In fact, Hendricks claims that if the antecedent of (CA) is true, then theism entails pointless non-resistant non-belief, meaning that ‘we should expect there to be’ pointless non-resistant non-belief ‘if theism is true’ (Hendricks 2025, 593).

In this paper, I critique Hendricks’s argument. While others (Schellenberg 1993, 191–199, 2007, 211–213; Blanton 2025, 709–710) have argued that the antecedent of (CA) is false, here I grant its truth and raise a different objection. I show that even if we assume the antecedent of (CA) is true, the consequent is doubtful. So the conditional is rendered doubtful as well. In this way, I argue that Hendricks’s argument contains the seeds of its own failure. More specifically, I show that Hendricks’s assumption of (RO), together with a few other plausible assumptions to which he also seems committed, undermine the support for his conditional.

In the next section, I lay out Hendricks’s overarching argument for the compatibility of theism and pointless non-resistant non-belief. Since the argument can be formulated in two distinct but interrelated ways, I formalise the first and offer a corresponding critique before turning to the second and doing the same.

Hendricks’s overarching argument

Hendricks’s conditional argument hinges on a central assumption borrowed from the Responsibility Objection. As he puts it, ‘God wants us to be responsible for helping others come to know him’ (Hendricks 2025, 593). By us, Hendricks means theists; by others, he means non-believers (Hendricks 2025, 595, n. 6). I will call this Hendricks’s Key Claim (KC):

KC: One of God’s important aims is for theists to be responsible for helping non-believers come to know God.

I include the word ‘important’ because, on Hendricks’s view, if God is willing to permit a pointless bad state of affairs to achieve that aim, then presumably it must be a relatively important aim of God’s, not merely something it would be nice for God to allow.

From (KC) Hendricks develops his argument for the compatibility of theism and pointless non-resistant non-belief. He argues that if God’s important aim is for theists to be responsible for helping non-believers to come to know God and

1. (i) God wants theists to remain motivated to do so,

and/or

1. (ii) God wants theists to be rational in doing so,

then God must allow pointless non-resistant non-belief. Hendricks writes that ‘it doesn’t matter which way we frame [the]…argument – although, I’ll note I find it more plausible when framed in terms of rationality and reasons’ (Hendricks 2025, 592). Let’s call (i) Hendricks’s ‘motivation-formulation’ and (ii) his ‘rationality-formulation’. In what follows, I argue that both formulations rest on assumptions that call (CA) into question.

The motivation-formulation stated

Let’s begin with Hendricks’s motivation-formulation. In this version, Hendricks’s central claim is that if God prevented all pointless non-resistant non-belief, many theists’ motivation to proselytise would be significantly undermined. He states the argument as follows⁵:

Suppose it’s true that God would allow some bad state of affairs only if it resulted in a greater good. Non-resistant [non-belief], of course, is a kind of bad state of affairs. Given this…it follows that all instances of non-resistant [non-belief] are cases in which either (i) they will come to know God or (ii) their continued lack of relationship with God will result in a greater good. So, any case in which one might try to help a non-resistant [nonbeliever] come to know God will be unmotivated (or, less motivated): we know that if the non-resistant [nonbeliever] remains [a nonbeliever], a greater good will come from it. And so we need not intervene to help this person come to know God – after all, if we don’t, the world will be better! (2025, 593)

Although the argument appears straightforward, it will be helpful to unpack it for two reasons. First, it relies heavily on assumptions drawn from Hasker, so readers unfamiliar with his work may miss much of what Hendricks takes for granted. Second, several of these assumptions are central to my critique of Hendricks’s argument. I therefore propose that we reconstruct a formalised version of the argument as follows:

1. (1) If God prevents all pointless non-resistant non-belief, then theists know (or reasonably believe) this.

2. (2) If theists know that God prevents all pointless non-resistant non-belief, then many theists’ motivation to proselytise is significantly undermined.

3. (3) God does not bring about or govern the world in such a way that a substantial portion of theists have their motivation to proselytise significantly undermined; God wants theists to be motivated to proselytise.

4. (4) If God wants theists to be motivated to proselytise, then God must permit pointless non-resistant non-belief.

5. (5) Therefore, God must permit pointless non-resistant non-belief.

To reiterate, this is an expanded version of Hendricks’s motivation-formulation. Although not every premise is stated explicitly by him, they are, I suggest, implicit and required for his case to succeed.

Consider premise (1). Hendricks does not state this premise, but he writes: ‘we know that if the non-resistant [non-believer] remains [a non-believer], a greater good will come from it’ (Hendricks 2025, 593; emphasis added). So it is implicit in his formulation. A natural question, however, is why Hendricks thinks that, if God were to prevent all pointless non-resistant non-belief, we, or at least theists, would know this. After all, could God not prevent theists from knowing it, especially if such knowledge would undermine their motivation to proselytise?

Here, Hendricks implicitly relies on two considerations drawn from Hasker: one about another central divine aim, and one about the impossibility, or impermissibility, of divine deception. The first is that one of God’s central aims is to lead people to know that he exists and that he is both loving and just (Hasker 1992, 39). If the claim ‘God prevents all pointless non-resistant non-belief’ expresses a truth grounded in God’s loving and just nature, then God could not achieve the aim of bringing people to knowledge of his existence without their recognising that truth about him. Therefore, if God were to prevent all pointless non-resistant non-belief, theists would know this. If one were to object that God might prevent all pointless non-resistant non-belief while also preventing theists from knowing or inferring that he does, Hendricks could reply, following Hasker’s second consideration (Hasker 1992, 39), that God cannot deceive us. If God were to prevent all pointless non-resistant non-belief, he would not ‘cover his tracks’, so to speak. That would involve a form of divine deception, which is either impossible or impermissible for God.

I think there are significant problems with these arguments.⁶ But since Hendricks is offering only a conditional argument, I’ll grant them, along with premise (1). What’s important to note is that premise (1) is crucial to Hendricks’s case. For if God were to keep theists from knowing that he was preventing all pointless non-resistant non-belief, even while in fact doing so, their motivation would not be undermined in the way Hendricks suggests.

Consider also premise (3). Hendricks assumes that God wants theists to be motivated to proselytise. He does not, however, claim that all theists would have to have their motivation completely undermined for his argument to succeed. If he did, the argument would falter, since some theists in the actual world hold that God prevents all pointless non-resistant non-belief and yet their motivation to proselytise remains undiminished. Rather, Hendricks assumes that God would wish to avoid a situation in which a substantial portion of theists would have their motivation to proselytise significantly undermined. This reading is supported by Hendricks’s (2025, 594) counterfactual: if groups of theists were to have their motivation undermined in the way he describes, that would be sufficient reason for God to permit pointless non-resistant non-belief. Moreover, Hendricks relies on Hasker, and Hasker (2019, 60) does not assume that all theists would need to have their motivation to prevent suffering completely undermined for his argument to work; it would suffice if a significant portion were to have much of their motivation undermined. Again, as with premise (1), I will simply assume that premise (3) is true.

Now consider premise (2). If theists were to know that God prevents all pointless non-resistant non-belief, many would have their motivation to proselytise significantly undermined. Hendricks’s support seems to run as follows. In the actual world, many theists are largely motivated to proselytise because they believe that preventing non-belief in God makes the world better (Hendricks 2025, 592). But if God were to prevent all pointless non-resistant non-belief, theists would know this and infer that they could not make the world better by proselytising (Hendricks 2025, 593). Therefore, if theists were to know that God prevents all pointless non-resistant non-belief, one of their main motivations to proselytise would be undermined, and many would be significantly demotivated in their proselytising.

The exact details of why theists would believe they cannot make the world better need not detain us. Indeed, Hendricks makes some errors in support of this claim⁷; I will address the main one in a later section. What matters for our purposes is the general line of thought. Hendricks’s rough idea is that, if theists were to know that God prevents all pointless non-resistant non-belief, they would believe that, by proselytising and preventing an instance of non-resistant non-belief, they would invariably foreclose the greater good for which that non-resistant non-belief is necessary. And if they were to take themselves to be invariably foreclosing that greater good, they would conclude that they cannot make the world better by proselytising.

This is why Hendricks thinks that God must permit pointless non-resistant non-belief (i.e., premise [4]). If theists were to know that God permits pointless non-resistant non-belief, they would know there are some instances of non-resistant non-belief that are not attached to any greater goods. And if they were to know this, they would know that preventing those instances would not foreclose any greater good, so they would be able to make the world better by proselytising. Therefore, their main motivation to proselytise would not be undermined.⁸

The motivation-formulation assessed

With Hendricks’s motivation-formulation clarified, let’s turn to criticism of it. As with the other premises, I’ll simply grant premises (2) and (4). That is, in what follows, I do not deny any of the premises or claims just discussed. Instead, I draw attention to the main assumption on which Hendricks’s reasoning is predicated, and which is easy to miss, namely that helping non-believers come to know God requires God to allow non-resistant non-belief. This is Hendricks’s (RO) assumption. Without it, his argument unravels. For without (RO), we lack any good reason to think God would permit non-resistant non-belief in the first place. More importantly, without (RO), the rest of his argument looks highly suspect: why would God specifically require pointless non-resistant non-belief to keep theists motivated to proselytise if non-resistant non-belief is not itself required for proselytising?

But as I will argue, (RO), when its implications are fully understood and combined with a few other plausible assumptions to which Hendricks also seems committed, conflicts with premise (3): the truth of their conjunction would mean that God would bring about a situation in which a significant proportion of theists lose much of their motivation to proselytise. In other words, the truth of (RO), together with these further claims, points to the falsity of premise (3). Hence, insofar as Hendricks is committed to these other plausible claims, his commitment to (RO) undermines his own argument. Thus, we arrive at a reductio of his position.

First, let’s draw out the implication of (RO) that I’ve been hinting at. To do this, consider three permutations of (RO):

RO¹: God allows non-belief (resistant and non-resistant) in God because it’s required for theists to be responsible for bringing others into relationship with God.

RO²: God allows resistant non-belief in God because it’s required for theists to be responsible for bringing others into relationship with God.

RO: God allows non-resistant non-belief in God because it’s required for theists to be responsible for bringing others into relationship with God.

In what follows, I argue that Hendricks must reject (RO¹) and (RO²) and restrict himself to (RO) for his argument to succeed. This is significant, because it entails that Hendricks must accept that resistant non-believers cannot be proselytised.⁹ From there, I’ll argue that this proposition, namely that resistant non-believers cannot be proselytised, together with two other plausible claims to which Hendricks also seems committed, leads to the undermining of his motivation-formulation.

Consider (RO¹). Hendricks’s argument cannot rely on (RO¹) for two main reasons. First, there is no good reason to think that both resistant and non-resistant non-belief are required for theists to be responsible for bringing others into relationship with God. Resistant non-belief in God (understood as a culpable, self-deceived state of epistemic error in which the non-believer has resisted belief in God; see Schellenberg (1993, 27–28; 2015, 54–55)) more than suffices for this responsibility.¹⁰ So there is little reason for Hendricks to accept (RO¹) except on an ad hoc basis. After all, why would God permit a bad state of affairs like non-resistant non-belief to achieve a good when that same good could be achieved without permitting that bad state of affairs, or by permitting a less serious one?

In fact, non-belief itself, whether resistant or non-resistant, does not appear to be required for bringing others into a relationship with God. So long as mere propositional belief that God exists does not compel faith in God (where faith is a fiducial state in addition to propositional belief), a different taxonomy could suffice for the relevant responsibility: resistant and non-resistant believers, with non-resistant believers bringing resistant believers into a loving, positive relationship with God (‘even the demons believe, and shudder’). Or, as one reviewer of this paper reminds us, perhaps union with God, or even salvation, ‘hangs in the balance on doctrinal issues’, such that non-belief is not required for the good of proselytisation.¹¹ Thus, it is difficult to see how Hendricks’s acceptance of (RO¹) would not be ad hoc, at least without further defence.

But suppose Hendricks could properly motivate (RO¹). Or perhaps he would simply say that he is assuming (RO¹) for the sake of his argument (which, again, is only a conditional argument). The second reason Hendricks cannot use (RO¹) for his argument is this: it would imply that resistant non-believers could be successfully proselytised. If that were the case, Hendricks’s argument would face a serious problem. This is because resistant non-belief could play the same functional role that Hendricks alleges pointless non-resistant non-belief plays in his argument. Recall the role he attributes to pointless non-resistant non-belief: on the version of theism he targets, he claims that, by proselytising and thereby preventing non-resistant non-belief, theists also foreclose the greater good for which that non-resistant non-belief is necessary; therefore, theists would believe they could not make the world better by proselytising. Pointless non-resistant non-belief, then, is meant to sustain theists’ motivation to proselytise by assuring them that not every instance of non-resistant non-belief is tied to a greater good, so they will not always be foreclosing a greater good by proselytising. In short, theists will retain the belief that they can make the world better by proselytising if they believe God allows pointless non-resistant non-belief.

But resistant non-belief can serve the same functional role. Of course, resistant non-belief is permitted by God for a greater good: respecting the non-believer’s autonomy and dignity in shutting out, resisting, or ignoring God’s existence and overtures (Schellenberg 1993, 27–28). The theist already knows this. She also recognises that, by proselytising, she, unlike God, does not risk foreclosing the very good for which resistant non-belief is tolerated. For if God were to overwhelm the resistant non-believer with a powerful and unmistakable display of his existence, this might compromise the non-believer’s autonomy and dignity (and thus the greater good). If, however, the theist softens the resister’s heart over time through good pastoral care and persuades her of her error, perhaps by presenting a reasonable yet not wholly compelling case for God’s existence, this need not strip her of autonomy and dignity (and thus would not compromise the greater good). Thus, even without pointless non-resistant non-belief, so long as she knows there are resistant non-believers who can be converted, the theist would know that she is not always foreclosing a greater good by successfully proselytising. Hence, she would know that she could make the world better. So much, then, for Hendricks relying on (RO¹).¹²

Next, consider (RO²). (RO²) does not work for Hendricks’s argument, for it does nothing to establish that God must allow non-resistant non-belief. It would therefore severely limit the scope of Hendricks’s response, since, arguably, non-resistant non-believers are a main impetus for hiddenness arguments. On (RO²), we would expect only resistant non-believers in the world. But since Hendricks’s argument aims to defend the compatibility of God and pointless non-resistant non-belief, he cannot rely on (RO²).

That leaves Hendricks with (RO), which holds that only non-resistant non-belief is required for proselytisation. If Hendricks or his supporters were to respond that resistant non-believers can also be converted, and so theists have the responsibility to proselytise them, then we are back to (RO¹) and its attendant problems. Thus, in putting forward (RO), Hendricks must assume that resistant non-believers cannot be proselytised.¹³

Let’s assume with Hendricks, then, that theists cannot successfully proselytise resistant non-believers. In other words, suppose that the world is such that theists can convert only non-resistant non-believers. Moreover, we must assume not merely that Hendricks alone thinks this is true, but that most theists believe that only non-resistant non-believers can be converted. Otherwise, we are back to the problems engendered by accepting (RO¹). For if most theists believe that resistant non-believers can be converted, then their motivation would not be undermined in the way Hendricks envisages. As we established earlier, when the theist prevents resistant non-belief in God, she knows that she does not thereby foreclose the greater good for which that resistant non-belief is permitted. So, even if the theist were to know that there were no instances of pointless non-resistant non-belief, she would know that she could make the world better by proselytising, namely by preventing resistant non-belief.¹⁴

Keeping this in mind, consider the following claim, which I’ll call No Eternal Disadvantage (NED):

NED: Non-resistant non-believers are not placed at a post-mortem disadvantage because of their ante-mortem non-belief in God; that is, their eternal welfare is not at stake by virtue of their ante-mortem non-belief.

(NED) is plausible. Many theists accept it (Baker-Hytch 2016, 376–377; Braddock 2018, 194; Mawson 2012, 195), and I am not aware of anyone in the literature who explicitly rejects it. This acceptance is also for good reason: rejecting (NED) would greatly intensify the problem of non-resistant non-belief. It would mean that non-resistant non-believers not only miss out on a conscious ante-mortem relationship with God, but also risk being deprived of eternal bliss, or even condemned to eternal torment, by virtue of their non-resistant non-belief. Of course, resistant non-believers might be at a post-mortem disadvantage because of their ante-mortem non-belief; their eternal welfare might be at stake by virtue of their non-belief in God. After all, their non-belief in God is the result of a freely willed opposition towards God (Schellenberg 1993, 27–28, 2015, 54–55). But surely non-resistant non-believers are not placed at an unfair disadvantage due to their ante-mortem non-belief. Their ante-mortem non-belief is not the result of resistance towards God.¹⁵

Perhaps some theists would reject (NED). Indeed, one reviewer of this paper writes that many theists would; as far as they can tell, ‘the entire Calvinist tradition would deny (NED)’. Suppose this is right. That these theists would deny (NED) is irrelevant to my argument’s success; it misdiagnoses its dialectical purpose. My purpose is not to convince those already in fundamental disagreement with me (or presumably Hendricks) about the nature of God’s attributes or how those attributes play out in the world. It is to convince Hendricks (or someone like Hendricks, with similar presuppositions) that his own reasoning is faulty in certain ways. I suspect that Hendricks and many readers of this paper will accept (NED). Whether the staunch Calvinist would reject our assumption here is not important. At most, this shows that the success of my argument is person-relative; but whose arguments, throughout history, have not been?

More importantly, I would question the claim that these theists would, upon sufficient reflection, reject (NED). At the very least, doesn’t the five-point Calvinist deny that there are any non-resistant non-believers? Doesn’t this kind of Calvinist assume that all non-believers are to some extent resistant? Similarly, do other Protestants really think that, if there are non-resistant non-believers, they are subject to an unjust disadvantage because of their ante-mortem non-belief? Or do they rather think that individuals are judged according to the information they have, in virtue of what God has given them, and that, if there are such non-resistant individuals, they will not be disadvantaged for lacking information they had not been given? Of course, theists need not believe that there are non-resistant individuals. But this is consistent with accepting the conditional claim that, if there are non-resistant individuals, they are not unfairly disadvantaged by virtue of their ante-mortem non-belief. So I will proceed as if (NED) is true; it is, after all, plausible and likely to be accepted by Hendricks himself.

Granted, if Hendricks were to reject (NED), he could avoid my argument in this section. But it is worth noting that, to reject (NED), he would need a good reason, since such a rejection appears to conflict with our conception of how a perfectly loving and just God would govern the world. After all, how could God punish an individual with eternal torment or deprive them of eternal bliss on account of a factor for which God, rather than the non-believer’s resistance, is ultimately responsible?¹⁶ Moreover, if the good of proselytisation were to come at the cost of, say, eternal damnation for the non-resistant non-believer, would it not be highly surprising on theism if (RO) were true? How could God see the permission of non-resistant non-belief as worth the cost? Yet Hendricks merely assumes (RO) for his argument. Could that assumption really be justified in that case?

Suppose, then, that (NED) is true. (NED), combined with (RO), would entail that no non-believer who can be proselytised has their eternal welfare at stake by virtue of their ante-mortem non-belief. Resistant non-believers might have their eternal welfare at stake, but, as we are assuming with Hendricks, (RO) would entail that resisters cannot be successfully proselytised. Thus, no non-believer who can be converted would risk losing out post-mortem on account of their ante-mortem non-belief (this might strike some theists as false about the actual world, but, if so, it is only because Hendricks must assume something false of the actual world for his argument to succeed).

Moreover, we can safely assume that, if God exists and (NED) is true, theists would know (or reasonably believe) (NED). After all, Hendricks assumes that one of God’s chief aims is to bring humans to knowledge of his existence and his loving, just nature. Since (NED) follows fairly straightforwardly from that nature, if God brings us to knowledge of his existence, he brings us to knowledge of (NED), or at least brings theists to knowledge of (NED). Furthermore, knowledge bearing on eternal destiny is hardly trivial, so on theism it stands to reason that theists would be sufficiently informed about the conditions for attaining or avoiding such outcomes – or, at minimum, would know something as modest as (NED). If Hendricks were to object that God would prevent theists from knowing or inferring (NED) when such knowledge would undermine their motivation to proselytise, the same move would undercut his own premise (1): God could likewise prevent theists from knowing or inferring that he prevents all pointless non-resistant non-belief. So, at the very least, Hendricks seems committed to the idea that theists would know (NED).

However, theists’ knowing (NED) and (RO) would spell trouble for Hendricks’s argument. For in the actual world, many theists place such importance on proselytising because they perceive the eternal welfare of at least some non-believers to be at stake and believe they can help change this. The following proposition is therefore true of many theists, which I’ll call Motivation for Proselytising (MFP):

MFP: A significant part of what leads many theists to attribute great importance to proselytising is the perception that being an ante-mortem non-believer disadvantages some non-believers eternally, post-mortem, and that they can help to change this.

But if (MFP) is true and theists were to know that no one they could convert was at risk of an eternal post-mortem disadvantage (that is, if they knew (RO) and (NED)), many would lose much of their motivation to proselytise. After all, their decision to proselytise would have no bearing on preventing any such disadvantage for any potential convert. Since many theists are motivated, in large part, by the belief that at least some non-believers’ eternal destinies are at stake and that they can help to change this, the realisation that no one they could convert is at such risk would remove one of their strongest motivations to proselytise.

Note that it doesn’t matter that theists might have other potential reasons to proselytise, such as putative commands from God in their scriptures. We’re discussing Hendricks’s motivation-formulation, not his rationality-formulation. So we’re not discussing what reasons theists could have to proselytise, but what actually motivates a good number of theists to do so. My claim is that, if that motivation were removed, a good portion of theists would be significantly demotivated. Moreover, if one were to press this point, it would cut equally against Hendricks’s reasoning and mine. Hendricks does not claim that there are no alternative reasons available to theists to proselytise beyond the reason he states (i.e., that proselytising ‘makes the world better’); rather, he assumes a primary motive that actually drives theists (i.e., the belief that proselytising ‘makes the world better’) and argues that this motivation would be removed if theists were to know that God prevents all pointless non-resistant non-belief. So that move is not available to Hendricks or his supporters.

So, in summary: if many theists in the actual world are motivated to proselytise in order to prevent a post-mortem disadvantage for some non-believers (i.e., (MFP)) but were to know that only non-resisters can be converted (i.e., (RO)) and that those non-resisters are not at such a disadvantage (i.e., (NED)), then it seems God would be bringing about a situation in which many theists’ motivation to proselytise would be significantly undermined. Those whose proselytising is largely driven by preventing a post-mortem disadvantage would no longer be so motivated. Thus, premise (3) would be false. And if we reach this conclusion without rejecting any of Hendricks’s assumptions and while adding only a few further plausible ones of our own, something has gone wrong with his reasoning. After all, the point of his argument is to show that God must permit pointless non-resistant non-belief to preserve theists’ motivation to proselytise; yet if his own assumptions, together with a few other plausible ones to which he also seems committed, yield a scenario in which God would be undermining many theists’ motivation to proselytise, we arrive at a reductio of his position.

I therefore submit that Hendricks’s motivation-formulation fails by its own lights. Or, more cautiously: given Hendricks’s own assumptions, together with a few plausible ones to which he also seems committed, there is at least as much reason to think (CA) is false as to think it is true.

The rationality-formulation stated

Let’s turn to the second way Hendricks might formulate his argument, namely his rationality-formulation. In this version, Hendricks’s central claim seems to be that it would not be rational for theists to proselytise if God were to prevent all pointless non-resistant non-belief. I say ‘seems to be’ because Hendricks does not provide a clear-cut rationality-formulation for non-resistant non-belief. However, he does cite (something like) Hasker’s rationality-formulation concerning suffering and thereby suggests how a parallel argument could proceed for non-resistant non-belief.¹⁷ In what follows, I modify Hendricks’s recounting of Hasker’s argument to refer to ‘non-resistant non-belief’, to show how the argument might go:

Assume we are irrational if we act in a way that goes against what we have most reason to do. And assume that every instance of [non-resistant non-belief] leads to a greater good. If that’s true, we lose one – and perhaps our strongest – reason for preventing [non-resistant non-belief]. That is, one of our strongest reasons for preventing [non-resistant non-belief] is that the world is made worse off on account of the [non-resistant non-belief]. But if that’s not true – if all [non-resistant non-belief] result[s] in a greater good – that’s no longer a reason for us to prevent [non-resistant non-belief], and it may be irrational for us to do so – we would be making the world worse by doing so, after all! Insofar as rationality is valuable, God has reason to ensure that his creatures act rationally when preventing [non-resistant non-belief]. (Hendricks 2025, 592)

Hendricks writes that ‘God has reason to ensure that his creatures act rationally when preventing’ non-resistant non-belief (Hendricks 2025, 592; emphasis added). But this still leaves ambiguity about God’s aim, given other things Hendricks says, and thus about how we should formulate his argument. Does Hendricks mean that God’s aim is to ensure that theists have (i) much more reason to proselytise than not, (ii) more reason to proselytise than not, or (iii) at least as much reason to proselytise as not? I suggest the most charitable reading is that God has reason to ensure that theists have more reason to proselytise than not. Framing Hendricks’s argument with this aim also makes my task harder than if we adopted the weaker idea that God’s aim is merely that theists have at least as much reason to proselytise as not, while avoiding commitment to the very strong claim that God would ensure they have much more reason to proselytise than not, a claim unwarranted by Hendricks’s own minimal language, which contrasts ‘rational’ with ‘irrational’ acts. So I suggest we formalise Hendricks’s rationality-formulation roughly as follows:

(1*) If God wants theists to be rational in proselytising, then they must have more reason to proselytise than to refrain from proselytising.

(2*) If God prevents all pointless non-resistant non-belief, then theists do not have more reason to proselytise than to refrain from proselytising.

(3*) God wants theists to be rational in proselytising.

(4*) Theists have more reason to proselytise than to refrain from proselytising only if God permits pointless non-resistant non-belief.

(5*) Therefore, God must permit pointless non-resistant non-belief.

In what follows, I challenge premise (2*). But let’s first examine Hendricks’s support for this premise, so we can see where it goes wrong. The support for (2*) seems to mirror his support for premise (2) of the motivation-formulation. However, instead of focusing on theists’ motivations to proselytise, we are now talking about the reasons they have to proselytise. Hendricks’s support for (2*) therefore seems to be this: in the actual world, many theists’ best reason (or one of their best reasons) to proselytise is that preventing non-belief in God makes the world better (Hendricks 2025, 592). But if God were to prevent all pointless non-resistant non-belief, theists would know this and would infer that they could not make the world better by proselytising (Hendricks 2025, 593). Therefore, if theists were to know that God prevents all pointless non-resistant non-belief, their best reason to proselytise would be removed, and many would not have more reason to proselytise than to refrain.

Again, the exact details of why theists would believe they cannot make the world better need not concern us here.¹⁸ What matters for our purposes is the general line of thought. Hendricks’s rough idea is that, if theists were to know that God prevents all pointless non-resistant non-belief, then they would believe that, by proselytising and preventing non-resistant non-belief, they would invariably foreclose the greater good for which that non-resistant non-belief is necessary. And if they were to believe they would invariably foreclose that greater good, they would conclude that they cannot make the world better by proselytising.

The rationality-formulation assessed

Let’s bring Hendricks’s assumption behind premise (2*) to the fore of our discussion, because it is this assumption I wish to contest. I’ll call it the Foreclosure Assumption (FA):

FA: If theists prevent an instance of non-resistant non-belief, they thereby foreclose the greater good for which that instance of non-resistant non-belief is (or would have been) permitted by God.

(FA) is crucial to premise (2*). If (FA) were false, then theists could prevent an instance of non-resistant non-belief without foreclosing the greater good for which that non-resistant non-belief is (or would have been) permitted by God. And if they could prevent an instance of non-resistant non-belief without foreclosing that greater good, then they could make the world better. And if they could make the world better by proselytising and preventing non-belief, then their best reason (or one of their best reasons) to proselytise remains intact. Hence, premise (2*) would be questionable, and the conclusion of the rationality-formulation would be correspondingly undermined.

Unfortunately for Hendricks, (FA) is false. Or at least, I’m going to argue that Hendricks must accept that (FA) is false unless he rejects the idea that God has bestowed on humans contra-causal (libertarian) free will. But I take it this is not something Hendricks would wish to jettison, or even could jettison, given his acceptance of (RO). The point of (RO) is that God permits some non-resistant non-belief to secure morally significant freedom, so that theists can bear responsibility for proselytising; that is, responsibility for bringing others into relationship with God or failing to do so. If God were to bestow such freedom, then Hendricks’s argument would be in trouble. For, on the view that God prevents all pointless non-resistant non-belief, it would follow only that God could not make the world better by preventing any non-resistant non-belief, not that we could not make the world better by preventing some. Put slightly differently: on the theism Hendricks targets, there is no instance of non-resistant non-belief that God could prevent in a way that would make the world better. But it does not follow that there is no instance that theists could prevent in a way that would make the world better.¹⁹

By analogy, it is commonly thought that even if God prevents all pointless suffering and so could not make the world better by preventing any suffering, this does not rule out theists’ making the world better by preventing suffering. Suppose, for instance, that God endows human beings with contra-causal (libertarian) free will. Olaf faces the decision to slap Kristoff and freely decides to do so. This introduces −2 units of badness into the world. Assume God could not prevent Olaf from slapping Kristoff without unacceptably interfering with Olaf’s libertarian freedom; that is, assume God’s interfering here would make the world worse. Also assume that the value of Olaf’s having the choice in this situation is +3 units of goodness, and that these units are commensurable with the bad units and bear on the overall value of the world. In the actual outcome (slap), the net value is +3 − 2 = +1. But had Olaf freely chosen not to slap Kristoff, the +3 units of value from having the choice would remain, and the −2 units of badness would not be introduced; the net would therefore be +3. Thus, although God may not be able to prevent the slap without making the world worse, Olaf could have prevented the slap and thereby made the world better. The key idea is that the responsibility or free will Olaf has here can itself be a greater good that outweighs the suffering resulting from its misuse. And that explains how Olaf could make the world better by preventing the suffering even when God could not.

Similarly, that God could not improve the world by preventing non-resistant non-belief does not entail that theists could not improve the world by preventing non-resistant non-belief. This is reinforced by what Hendricks himself assumes. Hendricks assumes (RO). (RO) is not merely descriptive but evaluatively loaded; it is offered as a reply to the problem of non-resistant non-belief. Moreover, it is offered as a reply because the theist assumes that the responsibility or free will to proselytise is itself a greater good that can outweigh or offset at least some of the non-resistant non-belief resulting from its misuse. But notice: if responsibility or free will is the greater good connected to non-resistant non-belief in at least some instances, theists would not foreclose that good by proselytising and preventing those instances of non-resistant non-belief, just as Olaf would not foreclose the greater good of responsibility or free will by choosing not to slap Kristoff and thereby preventing the suffering.²⁰ Theists would foreclose a greater good only if some further greater good were attached to the non-resistant non-belief in addition to the responsibility or free will to proselytise. In the present context, however, that suggestion appears ad hoc. For if the value of the responsibility or free will to proselytise is enough to offset, if not outweigh, the disvalue of non-resistant non-belief in some cases, there is no reason to suppose that God would connect any other greater good to these instances; they are already connected to a greater good. (FA) is therefore dubious. And if (FA) is dubious, this casts doubt on (2*), and hence on Hendricks’s rationality-formulation.

There are two ways Hendricks might try to salvage his argument. First, he might accept that the responsibility or free will to proselytise is a good, but deny that it is ever great enough, by itself, to offset or outweigh any instance of non-resistant non-belief. In other words, the responsibility to proselytise is a good, but not sufficient to incline God to permit any non-resistant non-belief in the absence of other goods secured by its permission. This would commit Hendricks to a very specific cumulative-case reply to the existence of non-resistant non-belief. Instead of the responsibility to proselytise being a good that alone can justify God in permitting some non-resistant non-belief (as advocates of (RO) typically claim), Hendricks would say that this good must always be combined with one or more additional goods secured by the permission of non-resistant non-belief, the value of which, only together, is sufficient to offset or outweigh the disvalue of a given instance of non-resistant non-belief. If this were the case, and provided the good of responsibility or free will constituted only a small part of the overall value of the combined goods connected to a particular instance of non-resistant non-belief, then Hendricks’s argument could be revived. For, in preventing an instance of non-resistant non-belief, the theist would still thereby foreclose the other associated greater goods for which that non-resistant non-belief is (or would have been) permitted by God. Hence, the theist could not make the world better by proselytising.

This is certainly an option for Hendricks, but the onus is on him to show that it is defensible. His argument would then rely not only on (RO) but also on other goods allegedly secured by the existence of non-resistant non-belief. Not only does this complicate his case, by making it more complex and importing debates about further controversial goods, Alexander Bozzo (2020) has shown that a cumulative-case reply of this sort is notoriously difficult to run against arguments from divine hiddenness. Briefly, and very roughly: for virtually any greater good posited as a reason for God’s permitting non-resistant non-belief, token goods of the same type can be achieved without God’s permitting it. Hence there is little, if any, reason for God to prefer token goods that involve non-resistant non-belief over token goods that do not. Thus, Hendricks would be resting his case on yet another contentious thesis, one that has already been shown to suffer from problems.

Moreover, to claim that the responsibility or freedom to proselytise is a greater good yet is never, by itself, sufficient to offset or outweigh any instance of non-resistant non-belief runs against what many theists claim about the high value of freedom and the limited disvalue of non-resistant non-belief. Many theists hold that non-resistant non-belief is not a very bad state of affairs, or that it is not especially surprising on theism, at least if post-mortem salvation does not require ante-mortem belief in God. This sentiment is captured well by Richard Swinburne and T. J. Mawson, who write:

Agnosticism, though (if there is a God) an evil, is not an evil as awful as some of the suffering that does require God to provide a compensatory afterlife or to become incarnate. (Swinburne 2004, 272; emphasis added)

For lack of belief in God…[is] much less obviously an evil than…[phenomenological pain]; that sort of suffering is much more obviously an evil…. [L]ack of belief in God needs to be linked to something that is obviously an evil… (Mawson 2012, 190; emphasis added)

…our believing theism to be true ante-mortem isn’t as good for us as some have maintained. (Mawson 2012, 195; emphasis added)

Additionally, many theists attach great value to morally significant free will and the responsibility that comes with it. Hence, Swinburne:

And the great good of significant free choice…is a great good indeed, sufficient to outweigh – I suggest – the evil of agnosticism. (Swinburne 2004, 272; emphasis added)

So, if the badness or disvalue of non-resistant non-belief is not especially great, and the goodness or value of having the responsibility to proselytise is great, then it seems mistaken to think that the responsibility to proselytise is never, by itself, great enough to offset or outweigh any instance of non-resistant non-belief.²¹ From personal conversation with Hendricks, I happen to know that he shares the general sentiment of these aforementioned theists (i.e., he holds that non-resistant non-belief is not especially bad). Of course, he need not share it. He, like others, is free to think that an instance of non-resistant non-belief is a very bad state of affairs, one that is never offset or outweighed by the good of the responsibility to proselytise. But in that case, our earlier criticisms about relying on a cumulative-case reply return with added force. The worse non-resistant non-belief is, the greater and/or more numerous the goods would need to be for God to permit it. And these would be particularly hard to find, especially in light of Bozzo’s criticism.

The second way Hendricks might try to salvage his argument is as follows. He could accept that the responsibility or free will to proselytise is, by itself, sometimes great enough to offset or outweigh instances of non-resistant non-belief, but maintain that such cases are rare. That is, in most cases the responsibility or free will to proselytise is not, by itself, sufficient to offset or outweigh an instance of non-resistant non-belief. For most instances, God would need to secure additional goods to help offset or outweigh the disvalue of non-resistant non-belief. If this were correct, then provided that in those cumulative-goods cases the good of responsibility or free will constitutes only a small share of the overall value of the combined goods, Hendricks’s argument could be revived. For, all else being equal, in most cases where an instance of non-resistant non-belief was prevented, the theist would still thereby foreclose the other associated greater goods for which that non-resistant non-belief is (or would have been) permitted by God. Hence, other things being equal, in most cases the theist could not make the world better.

But again, whether this route succeeds depends in part on one’s views about the value of free will and responsibility, and about the disvalue of non-resistant non-belief. The onus is therefore on Hendricks to show that this route is consistent with reasonable positions on these matters. He must also show that it is consistent with his other assumptions, including (KC). Recall that (KC) claims that one of God’s important aims is for theists to be responsible for helping non-believers come to know him; it must be important, since, on this view, God is willing to permit a pointless bad state of affairs to secure it. Whether (KC) can be retained under this route is questionable. If the responsibility to proselytise carries negligible value in most cases, why would one of God’s important aims be to preserve such a lacklustre good at the substantial cost of pointless non-resistant non-belief? If Hendricks adopts this route, he owes us an explanation of why this would matter so much to God. Furthermore, if it is not an important aim of God’s, would God really be willing to permit a pointless bad state of affairs to achieve it?

To summarise, Hendricks predicates his argument on a mistaken assumption (i.e., (FA)) about the relation between bad states of affairs and the greater goods to which they may be connected – more specifically, about the relation between non-resistant non-belief and its alleged greater goods. His mistake is to think that, by preventing non-resistant non-belief, theists invariably foreclose the greater good for which such non-belief is (or would be) permitted by God. Yet theists do not invariably foreclose that good. And Hendricks’s own assumptions, including (RO), together with common views about the high value of freedom and the limited disvalue of non-resistant non-belief, reinforce this.

I therefore submit that Hendricks’s rationality-formulation fails. More cautiously, Hendricks gives us no more reason to think (CA) is true than to think it is false. Perhaps there are ways to resolve the problems raised here, but I, for one, cannot see what they would be.

Conclusion

In this paper, I have examined Hendricks’s conditional argument for the claim that pointless non-resistant non-belief is compatible with, and even expected on, theism. Whether we consider his motivation-formulation or his rationality-formulation, I have argued that accepting his (RO) assumption (that God allows some non-resistant non-belief because it’s required for the responsibility of proselytisation) along with some other plausible assumptions to which Hendricks also seems committed leads us to the doubtfulness of his conditional claim. I have not argued that these problems cannot be overcome, but only that, as it currently stands, Hendricks has not yet established his intended conclusion. Further work is therefore required to make good on his conditional claim, and I leave it to Hendricks or others to clarify and resolve these difficulties.