A brief epistemological detour

With respect to justification, we can divide epistemologists into two camps: internalists and externalists.Footnote ⁴ Internalists hold that S is justified in her belief B if and only if S is, by reflection alone, (at least) potentially aware of that which contributes to B being justified. Externalists, conversely, deny this. Externalists hold that S may be justified in her belief B in cases in which she isn't aware of that which contributes to B being justified. For example, an externalist about justification will hold that B will be justified for S if it was produced in the right way (e.g. reliably or by way of proper function), and S need not be (even potentially) aware that B was produced in the right way for her to be justified in holding it. In what follows, I'm going to assume externalism, and in particular proper functionalism. According to proper functionalism – an externalist theory of justification – S's belief B is justified if and only if (i) B was produced by properly functioning cognitive faculties that have been successfully aimed at truth, (ii) B was produced in an appropriate epistemic environment, and (iii) S has no believed defeaters for B.Footnote ⁵ While I focus on proper functionalism, this is only for the sake of convenience – my argument below can be run in terms of other externalist views (e.g. reliabilism), and it arguably can be run in an internalist-friendly way (see endnote 10). The reason that I single out proper functionalism – as opposed to staying at a general level – is because it's easier to illustrate my point by using a concrete theory of justification. If the reader is attracted to a different theory of justification, she can easily switch out my uses of ‘proper function’ for whatever her preferred theory of justification dictates.Footnote ⁶

Next, let's say that a belief is basic if and only if it isn't based on another belief. For example, I believe that I'm sitting in a chair, and this isn't based on any other beliefs; it's basic. Non-basic beliefs are, unsurprisingly, beliefs held on the basis of other beliefs. For example, suppose my friend tells me that she'll attend the department party only if she gets off work early, and that I believe her. And suppose later that I see her at the party and form the belief that she got off work early. This belief is non-basic since it's based on my other belief that she will show up to the party only if she gets off work early. We may say that a belief is properly basic if and only if it is basic and justified.Footnote ⁷ Reformed epistemologists hold that belief in God can be properly basic – it can be justified in the absence of argument.Footnote ⁸ And, more generally, they think that the following proposition is true:

Why think this? That is, why should we think that belief in God is probably justified if theism is true? Below, I will provide two arguments for (J). My purpose isn't to provide a full-fledged defence of these arguments – that has been done elsewhere. Instead, my purpose is to show that there are plausible and (apparently) widely accepted arguments for (J).

Plantinga's argument

So, why should we endorse (J)? Alvin Plantinga provides the following argument:

Basically, Plantinga argues that God – if he exists – would intend to produce creatures that know him and can have a relationship with him, and that this would involve producing creatures with cognitive faculties that form true beliefs about him. In essence, Plantinga argues that the following proposition is true:

Clearly – since we're assuming proper functionalism – (P) entails (J).Footnote ¹⁰ And therefore, Plantinga's argument gives us reason to affirm (J).

The argument from divine hiddenness

In Hendricks (Reference Hendricks2021), I showed that the argument from divine hiddenness (explained below) gives us reason to endorse (P) and therefore reason to endorse (J). This is because, according to the argument from divine hiddenness, God – due to his perfect love – would (at least) probably be open to being in a conscious, reciprocal, and positively meaningful relationship with any (capable) creature who is not resistant to being in such a relationship with himFootnote ¹¹ – being in such a relationship, after all, is these creatures’ greatest good. However, believing God exists – so the argument goes – is a prerequisite for being in such a relationship. And this means that God would (at least) probably ensure that this prerequisite is satisfied – God would probably ensure that his (non-resistant) creatures believe he exists, since that is necessary for them to achieve their greatest good. However, there are some non-resistant creatures that don't believe he exists, and this counts as evidence against theism.

In other words, the proponent of this argument claims that God would (at least) probably ensure that there are no non-resistant non-believers: since God is perfectly loving, he is open to a relationship with his (capable) creatures. And this means that he would (at least) probably remove any barriers that would prevent this relationship from obtaining, and this means that he would (at least) probably ensure that those who are non-resistant to a relationship with him believe that he exists. However – so the argument goes – there are at least some persons who aren't resistant to being in such a relationship with God who nevertheless don't believe God exists. And this means that we have at least some evidence against theism: since God's existence makes non-resistant non-believers unlikely and we know that there are non-resistant non-believers, we have some evidence against theism.Footnote ¹² However, this line of reasoning commits the proponent of the argument from divine hiddenness to the following proposition:Footnote ¹³

To see why those who endorse the argument from divine hiddenness are committed to (H), suppose (H) were false. If that were the case, then the argument from divine hiddenness couldn't get off the ground: if it's false that God would probably provide the non-resistant with an ability to form beliefs about him, then non-resistant non-belief isn't surprising at all given theism. Indeed, it's expected: if (H) is false, then the non-resistant can't form beliefs about God, since they lack an ability to do so. And hence, proponents of the argument from divine hiddenness must assume (H). However, (H) entails (P): the background assumptions at work in the argument from divine hiddenness entail that belief in God is produced by properly functioning cognitive faculties. This is because

And this means that, if (H) holds, God probably has provided us with a cognitive faculty aimed at producing true beliefs about him. So, basically, if God provides us with an ability to form beliefs about him that serves as the basis for a relationship with him, this ability will be designed to produce true beliefs about him and will be properly functioning when doing so – if beliefs about God are produced by way of malfunction, they won't be able to sustain having a (conscious, reciprocal, and positively meaningful) relationship with God. And so if (H) holds, our beliefs about God will be produced by a properly functioning cognitive faculty aimed at truth, and this means that those who endorse the argument from divine hiddenness are committed to (P) and therefore they are committed to (J). And hence the argument from divine hiddenness provides us with reason to endorse (J).

The evil-god challenge answered

In the prior two sections, it was shown that there are (at least) two avenues of support for (P) and therefore (at least) two avenues of support for (J). That is, both Plantinga's argument and the argument from divine hiddenness give us reason to think that if theism is true, belief in God is probably justified.Footnote ¹⁵ The upshot of this is significant: (J) makes theism (significantly) more reasonable than maltheism. This is because maltheists can't mirror (J). Indeed, maltheists must reject the following proposition:

Why must maltheists reject (J*)? Because we have good reason to think (J*) is false. That is, we have good reason to think that it's unlikely that evil-god would ensure that his creatures’ beliefs about him are produced by properly functioning cognitive faculties. This is because it's worse to hold unjustified beliefs than it is to hold justified beliefs, and this means that evil-god has reason to ensure that his creatures’ beliefs about him aren't formed by way of proper function. In other words, since evil–god is concerned with increasing the badness in the world and unjustified beliefs are worse than justified beliefs, he has reason to ensure that his creatures' beliefs about him aren't justified. Therefore, we have an(other) answer to the evil-god challenge: theism is (significantly) more reasonable than maltheism because theistic belief is probably justified if true whereas maltheism probably isn't justified even if true.Footnote ¹⁶ Put differently, whether or not evil-god exists, belief in evil-god is probably unjustified. But this isn't the case for theism, and this means theistic belief is more reasonable than maltheistic belief.

Avenues forward for the evil-god challenge

There are (at least) two responses that proponents of the evil-god challenge could make to my argument in the previous section. First, they could argue both that Plantinga's argument fails and that the argument from divine hiddenness either doesn't support (J) or fails. And this means that theists can't – or need different reasons to – affirm (J), which means theism doesn't have a leg-up on maltheism. And second, they could argue that evil-god has reason to ensure that his creatures’ beliefs about him are produced by properly functioning cognitive faculties, that is, they could argue in favour of (J*) or a proposition that entails (J*). And this means that theism no longer has an advantage over maltheism – if (J*) and (J) are true, then theists are no better off than maltheists with respect to justification.Footnote ¹⁷

Neither response strikes me as particularly promising. However, it seems to me that the latter route has more promise than the former. Importantly, if this second avenue is pursued by the proponent of the evil-god challenge, she needs to show more than the mere possibility that evil-god would ensure his creatures’ beliefs about him were produced by properly functioning cognitive faculties: she must provide a plausible argument for thinking evil-god probably would do so – to establish (J*), one has to show that evil-god would probably ensure our beliefs about him are justified, not merely that he would possibly do so.

For ease of reading, let's call the proposition that evil-god would ensure our beliefs about him are justified (EJ). To show that the probability of (EJ) is high is to show that (J*) is true.Footnote ¹⁸ One way for the maltheist to show that the probability of (EJ) is high – and therefore (J*) is true – is to mirror Plantinga's argument above. That is, the maltheist could hold that evil-god would create us in his image.Footnote ¹⁹ And since evil-god is omniscient, this would mean that – since his creatures are made in his image – his creatures would have (at least some) knowledge, and therefore (at least some) of his creatures’ beliefs would be justified.Footnote ²⁰ Is this – being made in evil-god's image – sufficient to show that the probability of (EJ) is high? In brief, No. To show that the probability of (EJ) is high, it would need to be shown that evil-god would intend for us to have justified beliefs about him. And that doesn't follow from evil-god creating us in his image – perhaps evil-god would ensure that we have some justified beliefs (for whatever reason), but it doesn't follow from that that our beliefs about evil-god would be justified. What needs to be shown, therefore, is that evil-god has good reason to ensure that his creatures' beliefs about him are justified – good reason that overrides his reason for ensuring his creatures’ beliefs about him are unjustified, mentioned above. One way to do this would be to show that there are certain evils that evil-god would intend to bring about that would require him (evil-god) to make it such that our beliefs about him are justified. In other words, one could show that some evils that evil-god would intend to bring about require his creatures to have justified beliefs about him. If this could be shown it would, of course, mean that the probability of (EJ) is high and therefore that (J*) is true. Unfortunately, it doesn't appear that such evils are forthcoming; I know of no evils that are plausible candidates for rendering (EJ) high. And so this avenue shows little promise.

Now, perhaps the maltheist would argue that the fact that we know of no evils that render (EJ) high doesn't entail that there are none – after all, we've learned from sceptical theism that our grasp of goods, evils, and their entailment relations is quite poor (e.g. Michael Bergmann Reference Bergmann2001; Perry Hendricks Reference Hendricks2020). And this – the maltheist says – means that the probability that there is such an evil is inscrutable, not low. This move, however, won't be of use to the maltheist. This is because it's equivalent to holding that the probability of (EJ) is inscrutable – if the probability that there's an evil that overrides evil-god's reason to ensure his creatures beliefs about him aren't justified is inscrutable, then the probability of (EJ) is inscrutable. However, this provides maltheists with an undercutting defeater for their beliefs about evil-god. This is because if the probability that evil-god would ensure his creatures beliefs about him are justified (i.e. (EJ)) is inscrutable, then we should withhold belief about whether maltheistic belief is justified.Footnote ²¹ And if maltheists should withhold belief about whether their belief in evil-god is justified, then they should withhold belief in evil-god. This follows from the more general fact that if one withholds belief about the status of one's belief p's justification, that one has an undercutting defeater for p (Bergmann, Reference Bergmann2005).Footnote ²² And so objections that claim that the probability of there being an evil that entails (EJ) is inscrutable won't help the maltheist.

Due to the problems raised above, I don't see much promise for the maltheist to provide support for (J*) and revive the evil-god challenge. But perhaps a more developed ‘maltheology’ can answer this problem. I leave it to proponents of the evil-god challenge to develop this route of response more fully. Evil-god knows I can't.