1. Skeptical Theism

Evidential arguments from evil often take something like the following form:

By ‘God’ I mean a being that is omnipotent, omniscient, and supremely good. Most theists accept that God will allow an evil if there is an God-justifying reason for him to do so - e.g. if that evil is required by God to secure some compensating good or to prevent some equally bad or worse evil. A ‘gratuitous evil’, by contrast, is an evil there is no God-justifying reason to permit.

By a ‘God-justifying reason’ I mean a reason that would actually justify God in permitting that evil. Suppose I can save a child drowning in a river by reaching out to him from the bank with a piece of splintered wood. I decide against doing so because I might get a splinter from the wood. The risk of getting a splinter gives me some reason not to save the child using that piece of wood, but of course it's hardly an adequate reason. If God exists, then presumably he has not just some reason to permit the evils we observe, but adequate reason - reason that justifies him in permitting those evils.

Further, for an evil to be gratuitous, there needs to be no all-things-considered good reason for God allow it. An all-things-considered good reason is a reason that, when all factors are taken into account, justifies the relevant course of action (or inaction). Suppose I see child A is about to walk into some nettles. I have good reason to prevent her doing so: she'll get badly stung. That ’s a reason that would justify me in stopping her. Still, all-things-considered it might be better if I didn't stop child A and instead stopped child B whom I can see is about to walk in front of a car (assuming I can't do both). God may similarly allow evils he has reasons to prevent, including evils he'd be justified in preventing. What God presumably won't allow is evils he is all-things-considered justified in preventing. Henceforth, when I discuss ‘God-justifying reasons’ for doing x, I'll mean reasons that all-things-considered justify God in doing x.

Why suppose the second premise of the above argument is true? A so-called ‘noseeum’ inference has been offered in its support.Footnote ¹ It is suggested that if we cannot think of any God-justifying reason for an evil we observe, then we are justified in concluding that no such reason exists.

One obvious way to challenge this evidential argument from evil is to try actually to identify reasons why God might be justified in allowing the evils we observe, thereby showing that the evils are not, after all, gratuitous. Various attempts have been made. Some suggest that much of the evil we observe (in particular, the moral evils - the morally bad things we do of our own volition) can be explained as a result of God giving us free will. Some suggest that many natural evils - such as the natural diseases and disasters which cause great suffering to the sentient inhabitants of this planet - can be explained as a result of the operation of natural laws that are required for compensating or still greater goods (e.g. perhaps the tectonic plate movements that cause earthquakes and thus much suffering are necessary for life to emerge in the first place, say). Some suggest that many evils are divinely justified because they are for character-building or ‘soul making’ purposes. Just as a parent will permit their child repeatedly to fall off their bike and graze their knees given it's only by enduring such repeated falls that the child can gain not only the good of being able to ride their bike but also the justified sense of achievement that comes with it, so God will permit us to graze our metaphorical knees given that it's only by such means that we can become better people.

However, even many theists accept that these various explanations of why God would allow such evils are not only individually, but collectively, inadequate. I'd suggest that, for the two hundred thousand year history of human beings, the death of on average around half of every generation of children (usually in a pretty horrific way), with all the child and parental suffering and grief that that involves, is on the face of it, very difficult for theists to explain in any of the above ways, as is the hundreds of millions of years of horrific non-human suffering that occurred before we humans showed up.

The skeptical theist takes a rather different approach to the evidential argument from evil. Rather than try to identify the reasons why God is justified in allowing observed evils, the skeptical theist suggests that our inability to identify such reasons is not a sound basis for concluding that no such reasons exist.

The skeptical theist challenges the noseeum inference offered in support of premise 2. True, we are sometimes justified in inferring that there are no Fs given that there do not appear to be any Fs. I am justified in believing there are no elephants in my garage given there do not, looking in from the street, appear to be any there. But, the skeptical theist, points out, noseeum inferences aren't always sound. I am not justified in supposing there are no insects in my garage just because there do not, looking in from the street, appear to be any. Given my perceptual limitations, there could, for all I know, still be insects present. But then, suggests the skeptical theist, given our cognitive limitations, there could, for all we know, be God-justifying reasons for the evils we observe despite our inability to think of any.

We might think of those goods of which we are aware and those evils of which we are aware (and the entailment relations between them of which we are aware) as the tip of an iceberg of reasons. According to the skeptical theist, we don't know how much of this iceberg is accessible to us or how representative the tip is. But then, given our cognitive limitations, we cannot conclude from the fact that the part of the iceberg to which we have cognitive access contains no God-justifying reason to allow the evils we observe that it is probable (or even more probable then not) that there is no such reason in what remains. We are, insists the skeptical theist, simply in the dark about whether such a reason exists.

So, the skeptical theist maintains that, even if we can't identify any God-justifying reasons for the evils we observe, we are not justified in concluding that gratuitous evils exist. But then the evidential argument fails. Let's call the above skeptical theist attempt to block the noseeum inference the ‘anti-noseeum argument’.Footnote ²

Note that skeptical theism involves to claims: (i) theism, and (ii) skepticism regarding our ability to think of the reasons that might God justify God in allowing observed evils. Also note than even an atheist might embrace the skeptical part of skeptical theism. While failing to believe in God, they may nevertheless accept that, for all they know, there is a reason that would justify God, if he existed, in allowing the evils we observe.

Skeptical theism has been embraced and developed by several philosophers of religion, including Alvin Plantinga who, in response to the evidential problem of evil, says:

Michael Bergmann, a leading defender of skeptical theism, puts the objection thus:

According to Bergmann, the skeptical theist's skepticism (detached from their theism) includes as a main ingredient endorsement of such skeptical theses as:

Bergmann maintains that, given the truth of ST1-ST4, we are in the dark about whether there exist God-justifying reasons to permit the evils we observe. Thus the evidential argument from evil fails.

As McBrayer and SwensonFootnote ⁵ , two defenders of skeptical theism, point out, the skeptical theist's anti-noseeum argument applies, not just with respect to God-justifying reasons to allow or bring about evils, but with respect to God-justifying reasons to allow or bring about anything at all. If skeptical theism is true, we cannot, from the fact that we are unable to think of a God-justifying reason for God to bring about or allow so-and-so, justifiably conclude that no such reason exists.

Notice however that the skeptical theist need not - and arguably should not - be too skeptical regarding knowledge divine reasons.

Note, first of all, that skeptical theism allows theists can legitimately draw some conclusions about divine reasons given what they observe of the world. For example, they can legitimately infer from the fact that Bert's house burnt down last night, that God, if exists, had an adequate reason to permit that. Here is an inference from an observed evil to a conclusion concerning divine reasons that is permitted by the skeptical theist.

Further, note that skeptical theists can allow that we can also know at least some of God's reasons by means of some form of direct, divine revelation. Perhaps God can and indeed has directly revealed to some of us what his reasons are, and indeed what reasons he lacks. In which case, no inference - let alone a noseeum inference - at all is required in order for us to possess knowledge of both the existence and the absence of God-justifying reasons.

So skeptical theists can, and usually do, allow human knowledge of divine reasons. They're merely skeptical about our ability to think of the reasons God might have for creating or allowing the evils - and indeed the various other things - we observe. They question only the ‘noseeum’ inference from (i) we can't think of any such reasons, to (ii) no such reasons exist.

Note that skeptical theists disagree over whether the evils we observe provide some evidence against theism. Some insist the evils we observe provide no evidence at all against theism. Others allow that observed evils may provide some evidence against theism. They merely insist that - given the shaky nature of any noseeum inference from observed evils to the conclusion that no God-justifying reason for those evils exists - what evidence there is falls short of allowing us justifiably to conclude that the world contains gratuitous evil and that consequently theism is false.

2. Skeptical theism and knowledge of God's goodness

As McBrayer and Swenson acknowledgeFootnote ⁶ , skeptical theism also appears to threaten a number of arguments for the existence of the God of traditional monotheism. How are we to know that, not only is there an omnipotent and omniscient creator of the universe (a lower case ‘g’ god, if you like) but this creator is good (the upper case ‘G’ God)? According to McBrayer and Swenson, not by observing the universe and drawing conclusions about divine goodness on that basis. For if skeptical theism is true, we are as much in the dark about whether a good God would, or would not, bring about the goods we observe as we are about whether he would, or would not, bring about the evils we observe. But then observed goods are no more evidence for a good God then observed evils are evidence against.

Michael Bergmann, another proponent of skeptical theism, concurs that arguments for divine goodness based on identifying some feature of the universe as an all-considered good are undermined by skeptical theism. According to Bergmann, anyone who supposes the order we see in the natural world or the joy we witness in people's lives give us reason to think that there is a benevolent God who is the cause of such things is failing to take into account the lessons of skeptical theism.Footnote ⁷

This isn't to say that skeptical theism has the consequence that we should be skeptical about the existence of a good God. As Bergmann points out: ‘We needn't conclude … that the skeptical theist's skepticism is inconsistent with every way of arguing for the existence of a good God’Footnote ⁸ . Alternative ways by which we might come justifiably to believe in the existence of God might perhaps involve other forms of inference invulnerable to skeptical theism (e.g. an ontological or moral argument), or divine revelation.

3. The Pandora's Box Problem

One leading response to skeptical theism is to show that it opens up a skeptical Pandora's Box: it entails forms of skepticism that even the theists finds implausible and unacceptable. In particular, skeptical theism appears to require we also embrace skepticism about the external world and the past.

Consider the following familiar example of an undercutting defeater. I am watching, through a window, a series of widgets pass by on an assembly line. The widgets clearly look red. I come to believe the widgets are red on that basis. Presumably, given the widgets appear perceptually red to me, then it is ceteris paribus, reasonable for me to believe the widgets are red. However, suppose I am then told, by someone who has previously proved to be a reliable source of information, that the widgets are lit by a special defect-revealing light, a light that makes even non-red things look red. Is it still reasonable for me to believe the widgets are red?

Intuitively not. Why not? Because I now have good reason to think that the method by which I acquired the original belief is, in the circumstances in which I formed it, not to be trusted.

What, exactly, is ‘defeated’ in such cases? That's arguable. Some maintain that knowledge is lost in such cases. Even if the widgets are red (and I've been misled about that defect-revealing light), I don't know they are red. Others are inclined to think that knowledge need not be lost in such cases, but that at the very least reasonable belief is lost. Lasonen AarnioFootnote ⁹ , for example, argues that in such cases knowledge may be retained (e.g. if knowledge belief is acquired by means of a safe method, and the method employed - visual perception in this case - is indeed safe) but that reasonable belief is lost. According to Lasonen Aarnio, the reason reasonable belief is lost is that someone who continues to maintain their belief that the widgets are red even after having received the new information about the defect-revealing light, has embarked upon a belief-forming strategy that is not knowledge-conducive. Lasonen Aarnio suggests reasonableness

Someone who continues to believe even after acquiring such new evidence about the unreliability of the method by which they formed their belief will likely end up believing many falsehoods. I shall assume that Lasonen Aarnio is correct: whether or not knowledge is necessarily ‘defeated’ in such cases, reasonable belief is. Call such defeaters rationality defeaters (leaving it open whether knowledge is also lost).

But then doesn't skeptical theism generate a rationality defeater for beliefs regarding the external world and past? Given that it appears to me both that I ate toast for breakfast this morning and that there is an orange on the table in front of me, it is presumably reasonable for me to believe I ate toast for breakfast and that there is an orange before me. But if I now learn that, (i) God exists, and (ii) for all I know, there is a God-justifying reason for God to deceive me about these things, then, runs the objection, I can no longer reasonably believe I had toast for breakfast or that there is an orange there. For skeptical theism blocks any attempt to justify the belief that there are unlikely to be such God-justifying reasons by means of a noseeum inference: ‘I can't think of a good reason why God would deceive me in that way, therefore there probably is no such reason.’ But then skeptical theism would seem to have the consequence that, for all I know, God does indeed have a good reason to deceive me in this way and is deceiving me for that reason. Just as learning about that defect-revealing light provides me with an rationality defeater for my belief that the widgets are red - I should be skeptical about whether or not the widgets are red - so learning that (i) and (ii) generates a rationality defeater for my beliefs about the external world and past - I should be skeptical about the external world and past.

Of course, most theists reject the view that we should be skeptical about the external world and past. They believe we can reasonably hold beliefs about both. But if their skeptical theism requires that they embrace such a broad skeptical position, then it appears they must either embrace that broad skeptical position, or else abandon their skeptical theism, thereby resurrecting the evidential argument from evil.

Note that other beliefs also appear to be thrown into doubt by skeptical theism. Take a theist's belief that their religion - Christianity, let's say - is true. Skeptical theism appears to entail that, for all they know, there is a reason that justifies God, if he exists, in deceiving them about Christianity (maybe the truth of Christianity is something about which God wishes to trick us in order to achieve some, to us unknown, greater good). But then it seems skeptical theism provides our Christian with rationality defeater for their Christian beliefs. They should, it seems, be skeptical about Christianity, just as they should be skeptical about the external world and past.

Note that, even if disbelievers (those who believe there is no God) do accept the skeptical part of skeptical theism (they endorse the thought that they are in no position to know whether there's a reason that justifies God, if he exists, in deceiving them), they don't end up falling into the same skeptical swamp. For, on their view, there exists no such God, and thus no such deceiver.

Commonsensism

In response to the Pandora's Box Objection, some insist that, yes, we cannot by means of a noseeum inference, conclude God lacks a reason to deceive us about the external world and past - i.e. we cannot think of a reason why God would deceive us about the external world and past; therefore there probably is no such reason. However, while that way of establishing that God lacks a reason to so deceive us is blocked, other ways of knowing that he lacks such a reason may remain. Perhaps, given there are these other ways of knowing about the external world and the past (ways that don't rely on any noseeum inference regarding God's reasons), skeptical theism constitutes no threat to such knowledge.

For example, Michael Bergmann, in response to the Pandora's Box objection, appeals to what he calls commonsensism.

Bergmann then considers Sally, a hypothetical agnostic who believes ST1-ST4 but who also signs up to commonsensism. According to Bergmann, given Sally's commonsensism, especially clause (b),

Note the intriguing move made here: from the fact that we do know (other than by means of a noseeum inference) about the external world and past, we can infer that God, if he exists, has no God-justifying reason to deceive us about the external world and past. A similar move is made by BeaudoinFootnote ¹³ who, in response the thought that skeptical theism entails that, for all we know, God actualised s: an old-looking universe that is in truth just five minutes old (this being the universe we inhabit), suggests that while we cannot infer God lacks a reason to so deceive us about the age of the universe from the fact that we cannot think of any such reason, insists that nevertheless we can infer God lacks such a reason from the fact that we do, in fact, know the universe is older than that. Beaudoin draws the following analogy:

Similarly, then, says Beaudoin, we may yet know the universe is old, not by way of a noseeum inference to a conclusion about God's lacking reason to deceive us concerning its age, but in some other way. Perhaps, says Beaudoin,

I call this the Bergmann/Beaudoin response to the Pandora's Box objection to skeptical theism.

I don't believe the Bergmann/Beaudoin response succeeds in disarming The Pandora's Box objection. Consider another putative example of rationality defeat, which I call Olly's Orange Footnote ¹⁶ .

Can we know God is no deceiver?

Another response to the Pandora's Box objection is to argue that we can, independently, know God is no deceiver because we can know that God is good and a good God is no deceiver. In his Third Meditation, Descartes offers an argument for this claim. He says God ‘cannot be a deceiver, since it is a dictate of the natural light that all fraud and deception spring from some defect’, and God is without defect. However, MaitzenFootnote ¹⁷ (2009) points out that while fraud and deception flow from some defective situation (a terrorist about to explode a bomb who can only be thwarted by deception, for example) it does not follow that ‘fraud and deception are defective responses to that situation’.Footnote ¹⁸ Hobbes similalrly points out, in response to Descartes, that it

Where a reason sufficient to justify us in engaging in deceive exists, our engaging in such deception does not require there be any defect in us. So why does God's similarly engaging in such deception require there be some defect in him? The New Testament also contains passages suggesting God engages in deliberate deception. St. Paul describes God as sending some people ‘a powerful delusion, leading them to believe what is false.’ (2^nd Thessalonians 2:11). So the Cartesian thought that God is no deceiver is Biblically challenged, too.

Is being ‘in the dark’ about whether God has reason to deceive us sufficient to justify skepticism?

Here's another response to the Pandora's Box problem.

Suppose Paul tells me, with seeming sincerity, that he had an apple for breakfast. I have only just met Paul and don't know anything about him. Nevertheless, I believe him. Is it reasonable for me to believe him?

Well, am I not in the dark about whether Paul has reason to deceive me about his breakfast this morning? Paul is a complete stranger to me. I know nothing about his background or his motives. So, for all I know, Paul has some all-things-considered good reason to deceive me about his breakfast. Should I not then withhold judgement about - be skeptical - about whether Paul had an apple for breakfast?

Skepticsm in this case seems absurd. Surely, despite the fact that I am in the dark about Paul's motives and the reasons he might have to deceive me, it's reasonable for me to just take Paul's word for it about his having an apple for breakfast.

Note that principle P2 explains why it's reasonable for me to take Paul's word for it about the apple: ceteris paribus it is reasonable for me to take testimony at face value; hence it's reasonable in this case. Notice that if I were sceptical in this case, then consistency would require I be sceptical about a great deal since much of what I believe is based on the testimony of folk not well known to me - folk who, for all I know, have reason to deceive me.

Hence, a defender of skeptical theism may insist, the mere fact that I am in the dark about whether God has good reason to deceive me - the fact that for all I know he has reason to deceive me - does not entail that I cannot reasonable believe God's testimony, or indeed my senses and memory.

To assess this response to the Pandora's Box Objection, we need to get clearer about what being ‘in the dark’ and ‘for all I know’ mean here. When sceptical theists say we are ‘in the dark’ about whether there exist God-justifying reasons for God to allow the evils we observe - that ‘for all we know’ such reasons exist - they re-articulate this thought in a variety of ways. Some speak of probability. They say that the probability of God having such a reason is inscrutable to us, by which they mean that we cannot reasonably assign any probability to God's having such a reason: neither high, nor low, nor middling. Others speak of probability but say only that we cannot assign a low probability to God's having such a reason.

If we now turn to the case of Paul, I think it is pretty clear that while I might, correctly, say ‘for all I know’ Paul has a reason to deceive me - that I am ‘in the dark’ about whether Paul has reason to deceive me - the sense with which these expressions are being used is not that with which they are used by the skeptical theist.

It's actually very reasonable for me to believe that Paul lacks a reason to deceive me because, after all, Paul is a human being, and I know a great deal about human beings generally, including the kind of reasons that lead them to deceive others, the extent to which they can survey the range of reasons that would motivate them to deceive others, and so on. Given all this information about human beings and their reasons to deceive, it's reasonable for me to believe that, while Paul might have reason to deceive me, the probability he actually has such a reason is low. But then that low probability doesn't give me a rationality defeater for my belief that Paul had an apple for breakfast given only that he told me so. True, I am, in a sense, ‘in the dark’ about whether Paul has such a reason - ‘for all I know’ Paul has such a reason - but only in the very weak sense that it's possible that he has such a reason - I can't entirely rule out his possessing such a reason. I can still reasonably assign a low probability to his having such a reason.

When we turn to a skeptical theist's claim that we are ‘in the dark’ about whether there are God-justifying reasons for observed evils, on the other hand, the claim is that the probability God has such a reason is inscrutable and/or is at least not low. If we could reasonably suppose the probability of there being such a reason was low, then perhaps we might still reasonably believe there's no such reason, and thus reasonably believe that the evils we observe are gratuitous. So, if their response to the evidential problem of evil is to succeed, the skeptical theist's sense of our being ‘in the dark’ re the existence of certain God justifying reasons needs to be a very different sense to that which applies in the case of Paul's potential reasons to deceive me.

But then if it's in this stronger sense that we are supposedly ‘in the dark’ regarding the existence of God-justifying reasons of observed evils, then it will also be in this stronger sense that we are ‘in the dark’ regarding the existence of God-justifying reasons for God to deceive us about the external world and past. But then the analogy drawn between our ‘being in the dark’ about God's having good reason to deceive us and my being ‘in the dark’ about Paul's having good reason to deceive me fails. Even if it is reasonable for me to trust Paul, notwithstanding my being ‘in the dark’ about his having good reason to deceive me, it does not follow that it's reasonable for me to trust God, notwithstanding my being ‘in the dark’ about God's having good reason to deceive me. It seems I really do have reason to distrust God since I cannot - as I can in the case of Paul - reasonably assign a low probability to God's having good reason to deceive me.

Note, by the way, that in Olly's Orange, for my analogy to appropriate, I must be ‘in the dark’ in the strong sense about whether Olly has turned his projector on. That's to say, I cannot reasonably assign a low probability to Olly's having turned his projector on. Under those circumstances, it appears I do have a rationality defeater for my belief there's an orange before me.