1. The aim of this paper

Two of the most interesting developments in recent years in the debate about the ‘problem of evil’ have been sceptical theism and Peter van Inwagen’s invocation of the Fall as a way of justifying God’s permitting suffering and other evils. My objective in this paper is to link these two together in a more explicit way than van Inwagen does by focusing the scepticism more exactly than is typical; and also to draw out what I take to be the main thrust of van Inwagen’s argument in a way that avoids the principle lines of criticism to which it has been subjected.

I begin by focusing the intuition behind sceptical theism, by claiming more specifically what it could be in God’s plan that we might reasonably think we were in no position to adjudicate. In a sense, this is filling out or interpreting God’s reply to Job. The path I follow in doing this is very close to van Inwagen’s in his splendid Gifford Lectures. The attempt is to ‘justify’ God’s substantive withdrawal from providential interference in the light of our chosen fallen state, and why we are in no position to judge that this withdrawal could have been done in a way that had less terrible consequences.

2. The ‘problem of evil’, the sceptical response, and why that response is problematic

John Mackie argued that the existence of evil and suffering in the world is formally incompatible with the existence of the God of traditional theism. A variety of responses concerning the role of human free-will, and of adversity in promoting virtue, are taken as having shown that there is no formal inconsistency, but a revised version of the argument claims that the extent and intensity of gratuitous suffering is such that there is an inconsistency between these so-called ‘horrendous’ or ‘pointless’ evils and the existence of God.

Rowe’s challenge to theism along these lines is expressed (by Dougherty SEP) as follows:

Theological Premise: Necessarily, if there is a God, there are no pointless evils.

Empirical Premise: There are pointless evils.

Conclusion: There is no God.

There are in fact, I think, two versions of this line of argument. One of them is in terms of pointless evils and the other in terms of horrendous evils. If I follow the jargon, a pointless evil is one that does not lead to some later greater or compensating good. Such an evil need not be terribly bad, it just fails to serve an adequate purpose. A horrendous evil is one so terrible that nothing could reasonably be held to compensate for it, at least morally.

One response to this version of the argument from evil has been to say that we are not in a position to judge whether God has good reason for allowing these things—we suffer from, as it were, incorrigible ignorance (in this life at least) on such matters. This is the position known as sceptical theism. Sceptical theists accuse proponents of Rowe’s argument of following the Noseeum principle; that argument, used by the atheologist, is that, because we cannot see a reason for these evils, then it is reasonable to conclude that there is no reason. The sceptical theist points out that such an inference, from ignorance to a negative conclusion, is not a valid inference. The atheist’s strategy in this is similar to that labeled by David Armstrong ‘the “headless woman” argument’ – because you can’t see her head, she must be headless (Armstrong, Reference Armstrong1968). The sceptical theist argues that it is more rational to adopt some such principle as the Condition Of ReasoNable Epistemic Access, abbreviated as. CORNEA:

So, if H claims ‘it appears that this suffering is pointless’, this would be reasonable only if she thought that, if it were not pointless, she would be able to recognize this fact. Why she might think that she could not do this is explained by Wykstra’s principle

Obscurity

If this latter is true, and we are in a position to see that it is probably true, then we have reason to think that we could not follow God’s reasons, and so should be sceptical about Rowe’s argument.

Sceptical theism has itself been met with a good deal of scepticism. It seems to give the theist too much of a blanket excuse, and, anyway, is there not some rational obligation to base at least a cautious, probabilistic judgement on such evidence as one can have – provided, at least, that evidence has at least a certain prima facie force – as, in the case of horrendous evils it must strike us as having? Furthermore, it has been argued the scepticism can undermine our common-sense morality: if God can, for example, allow the horrendous death of many innocents, perhaps we are not doing good by intervening to prevent suffering, as far as we can, for we may be frustrating God’s inscrutable purposes. This claim against sceptical theism is controversial, but, if accepted, it would demonstrate the destructive consequences of ‘blank cheque’ sceptical theism.Footnote ¹ Any scepticism must be more specifically focused.

On the other hand, no traditional theist – perhaps especially Christians – would deny that ‘God’s ways are not our ways’ and that there is much about Him that ‘passes all understanding’. Is there any way to reconcile the natural suspicion that one tends to feel towards sceptical theism that it is too easy (perhaps somewhat analogous to the ‘mysterion’ explanation of our supposed inability to solve the mind-body problem) and the acknowledged and legitimate limitation of our understanding of God’s ways and His unique position in relation to creation?

Accepting a ‘blank cheque’ sceptical theism on ‘obscurity’ grounds would allow God to get away with almost anything on the grounds that we cannot know what He is up to. Perhaps, even if there were nothing but misery in the world, it might still be for the best, even though it is totally unimaginable how. In fact, sceptical theists, in so far as they are trying to respond to the problem of actual horrendous evils, are not committed to this view, which suggests that there could be ways that misery might be brought about that we can see would be inconsistent with the existence of a good God. I want to contrast any ‘blank cheque’ version to what I shall call ‘issue-specific sceptical theism’. The specific issue in question is not pointless or horrendous evil as such, but certain principles that I shall claim can be invoked to explain these evils and the reasonableness of which I shall argue we are unable to pronounce on, hence the sceptical element.

3. ‘Focused’ scepticism

So the current situation concerning God and evil seems to be the following.

The debate over the problem of evil has tended to focus on whether we are in a position to judge whether there are certain evils in the world of which we can say that they are so totally pointless that a good God would not have allowed them. Consequently, there is no good God. Sceptical theists reply by claiming that we are not in a position to judge whether God might have a good reason that we cannot perceive for these apparently pointless evils. Any orthodox theist will allow that the ways of God will be, in some ways, a mystery to us and the sceptical theist applies this principle to our understanding of evil. The opponents of sceptical theism tend to regard the argument as too much of a blank cheque, or a matter of theft over honest toil.

One striking thing about the debate on God and evil is that it seems to consider particular evil events (or perhaps particular kinds of evils) and assume that, if they are pointless then a good God would or should have prevented them, by Divine intervention if necessary. So the question is why He ‘permits’ them, as if they were one-off events. So even sceptical theists, such as Bergmann seem to accept that God must have a good reason for allowing some specific evil to occur, though we are in no position to see what this might be.

Bergmann’s argument for sceptical theism is as follows.

The crucial move is from (1) to (2). Bergmann disputes the validity of this deploying the following sceptical theses (STs).

Sterba cites a grotesque case of cruel sadism from Dostoyevski to illustrate how Bergmann’s principles would give God a blank cheque, because, on Bergmann’s terms, we could not know the real results of the calculus of benefits.

Sterba challenges this by saying:

So Sterba thinks that God could permit such things only if something ‘like informed consent typically obtains’.

There are two clear assumptions in this debate. One is that God is to be judged by what one might call an act-consequentialist standard, namely, He is held responsible for whatever happens and its consequences, in any given circumstances, whether He intervenes or does not. This is held to be reasonable because, given His power, which means that He could always intervene effectively in whatever situation, He is responsible for anything He allows to happen. The other is that God, although the creator and origin of everything, must be judged by exactly the same standards as are relevant to those of us operating within the system He created. What I hope to show in what follows is that what is in question is the propriety of certain policies (as one might call them) adopted by God, not individual consequences; and that, although God’s position as creator does not mean that we cannot draw on our intuitive moral standards when trying to understand what His goodness involves, neither can we apply those ideas in a naively ‘democratic’ way, as if His situation as creator makes Him just a ‘bigger’ version of ourselves.

I think that there is available a broader approach to the issue, and one which makes more focused use of the ‘sceptical’ idea that there are some things on which we are unable to judge with any reasonable confidence.

The trouble with the blanket dismissal of the atheist’s argument conclusion is that it does not give a positive account of why one should not be able to see God’s reasons.

Something that explains why we may not be in a position to judge, but not just in general terms, would be better.

My argument turns on two major principles that involve a form of scepticism about our capacities but that, I hope, explains why these might be things we cannot evaluate.

An argument along these lines was used by Richard Swinburne (Reference Swinburne1978), and also by van Inwagen (Reference van Inwagen1995, Reference van Inwagen2006). Swinburne’s reason for it is that, if free will is to have moral significance, we must be able to do serious harm to others. So, nature must be easily open to brutality in its normal workings: if you are to be able to cause pain by hitting someone with a hammer, the world must be such that rocks and trees can fall on people and cause pain and harm. If someone replies that God could have made a world such that the people could do harm but operations of nature could or did not, van Inwagen’s response comes into play. He deploys what he calls ‘modal scepticism’, which is not a Quinean disbelief in modal truths, but an epistemological constraint. We can in a superficial sort of way imagine that something is possible, but if you try to work it through in detail, it is not clear what the margins of serious possibility are. We know that this world works and that it can produce and sustain things like us, but we cannot say with any confidence that there could be a world nomologically intelligible to us that lacks the risks but contains the benefits of ours. This case of what I call ‘focused scepticism’ is common to both van Inwagen and the author of this article.

The natural response to (A) is that God could and should intervene to prevent at least the worst cases of natural evils. Van Inwagen’s response to this is to invoke vagueness and sorites and argue that the removal of one evil makes no significant difference to the suffering in the world, so it would make no difference, by a sorites argument, if there were none, which is ridiculous, therefore one should not try to fine tune the amount of suffering in the first place.

I think most critics find this argument unconvincing: would not it have been obviously good if one of the early plots to assassinate Hitler had succeeded and this would not have seemed strange or miraculous: assassinations do occur without provoking wonder. One needs an account which frees God from the seeming obligation to intervene for the sake of lessening suffering but allows for miracles that occur for the purpose of saving souls – although this aspect I will not be discussing here: I mention it only to indicate that a deistic fatalism, which rules out miracles, is not part of my programme.

My strategy to justify the amount of suffering we find in the world is different from van Inwagen’s. It is the following.

The core of my thesis is that if (A) and (B) were true, then Rowe’s original argument would fail, for it would not be wrong to allow the evils in question if (A) and (B) are true. It seems to me that we are in no position to argue with any confidence that either (A) or (B) is untrue. Defending this last claim is the point of the rest of this paper.

4. Preamble; real or apparent strategic differences from the usual argument

(1) The strategy of my argument somewhat reshapes the classical argument. Classical forms of the sceptical theism strategy attack the second premise of Rowe’s argument. Mine can be seen either as doing that, or denying the first premise, depending on how one defines ‘pointless evils’. If such evils are ones that have no compensating good consequences, then I am allowing that there can be such evils, at least in the most natural way of understanding that idea. On the other hand, you might say that they are not pointless, if they follow from (A) and (B) and (A) and (B) are acceptable. And if (A) and (B) are correct, then horrendous evils are not ruled out. The label ‘horrendous’ is, I think, more appropriate here than ‘pointless’, because something can be pointless, but not very terrible, and it is the horrendousness of these evils that mainly concern us.

(2) A more fundamental difference is the following. In the usual debate, it is argued that God should intervene to stop any individual evil which is pointless or horrendous: He is being held to what one might describe as an act consequentialist standard. By contrast, both (A) and (B) concern the validity of general principles; if those principles are sound, then individual bad cases will be tolerable. God is not judging case by case what to do, but stands back in general and lets nature take its course. This makes an important difference to the argument. The issue then is not whether a morally good God would intervene in any case in which greater good does not follow from allowing the evil, but whether a good God could follow the general policy of apparent withdrawal that is stated in (B), and follow only a ‘rule’ based policy, not an ‘act’ based one.

Argument for Rowe’s second premise usually takes the form of citing specific kinds of case – such as the suffering of the hind in a fire – rather than something like a law of nature.

That it cannot be shown to be unreasonable on God’s part to follow certain general procedures or rules is an essential part of my argument, even though this might allow there to be cases of individually pointless or horrendous evils; that is, they have no purpose beyond the fact that they follow from some more general principle that is morally legitimate.

(3) Although the argument over evil is usually presented as a piece of pure natural theology, based on something close to a straight inconsistency between God’s goodness and the sheer existence of pointless evil, and so involving no particular doctrinal commitments, (B) makes specific appeal to the doctrine of the fallen nature of man and the world. It could be said, therefore, to locate this response to the argument in a doctrinal context particularly appropriate, but not exclusive, to Christianity. The ‘fallen nature’ in question, however, does not necessarily rest on a literal reading of Genesis, but only on a certain reading of the human condition. Nevertheless, it is an important feature of the Genesis account, even if taken as a parable and not a historical account, that God permits the ravages of nature, expressed fundamentally by death itself, as a result of sin. It is part of what I am arguing that this claim can do work when discussing the problem of evil as a piece of natural theology.

5. Arguing for (A)

It is easy to object glibly to (A) by saying that an omnipotent God could do anything that involves no contradiction, and a system that never generated any pointless pain or suffering involves no contradiction, so God could have created it. But the details of how this might work and what consequences it might have for human nature are beyond any reasonable calculation. One could suggest that God could have created beings very different from us who might have fitted into such a radically different world, but again how that would work and whether He absolutely ought to have created such beings seem to be things that we are in no position to assess. Also, it is not clear that we have the right standing to complain if God failed to create creatures radically different from ourselves that might have suffered less. The devil is in the details and the details are wholly inaccessible to us. These issues are clearly and convincingly expounded in Peter van Inwagen (Reference van Inwagen1991).

This argument is not without its problems. One might distinguish between basic laws and their human consequences. So one could argue that we cannot see how there might or might not be laws of nature following which, e.g., no heavy object would fall on conscious creatures and damage them. But could we not preserve those laws but not allow that to lead to human physiology, which makes such an event excruciatingly painful?

Now, we know that humans who do not feel pain are usually dead before they reach 30, because they do not alter their position in bed when blood is failing to reach some limb. But it seems that there are some people whose anagnosia is more moderate, who feel little pain (and do not get anxious!) but who live more or less normal lives. Their spouse has to warn them occasionally when they don’t notice they are burning themselves on the oven, but this seems manageable. Apparently, the modification of a single gene is responsible for this, and it is not clear why this could not have been general, without, so to speak, messing around with Newton’s laws of motion, which govern the things that fall on us.

Once again, I would claim that we do not know what the consequences (antecedents) of this would be, how it would have fed back through evolution and with what possibly disastrous results.

6. Arguing for (B)

Principle (B) takes, I think, a rather different approach from van Inwagen’s. We both agree that humans in an unfallen state did not, or, if one thinks they never were in fact in that state, would not have been subject to the evils we now see. The question is how these evils could have been avoided. Van Inwagen’s answer is as follows.

My version of this claim is illustrated by the following; if Adam had slipped on the edge of a cliff before the Fall, God would have caught him; after the Fall, he would have had to live – or die – with the consequences. Van Inwagen attributes the ability to avoid evil to the capacities of the human, given their union with God; my way of expressing it stresses divine intervention, but the difference is not salient.

Van Inwagen moves on from what he takes to be the fact of the broken union with God and its consequence that we are no longer protected from misfortune, to the suggestion that God could, at least, intervene to prevent the worst of evils. He has two arguments against this. The first turns on the fact that it would involve a great many interventions and hence massive irregularity in nature, and he claims that it could be the case that

Second, he fears that there would be a sorites progression: after one set of evils had been removed why not the worst amongst those remaining, and so on?

Van Inwagen does not seem to face the question of why, given that we are fallen, God would give up the close support that we did, or would have, experienced if we had not been fallen? He seems to assume that this can be treated as a sort of natural consequence, but why should this be so? After all, He is supposedly a God of great love for His creation, especially for humankind, and is said to show great patience with our sinfulness.

I think that there is an answer to this which van Inwagen does not consider. This is that, if God’s intervention could be relied upon, we would ‘play the system’.

An illustration of similar manipulation is as follows: someone I once knew remarked of a friend of his who was a step-father that the step-child treated his friend ‘like a cheque-writing machine’. Because he needed to gain the love of someone who was not his natural child, he could never refuse a request, however exploitative. The step-child was ‘playing the system’. Satan tempts Jesus to do just this.

The rationale for (B) could be put in terms of two considerations:

1. (a) We are in no position to judge what the overall consequences of God pursuing a policy of intervention to prevent great evils would be and, hence, whether it is unreasonable of God not to permit Himself to be manipulated in this way.

If it is not unreasonable for a creator to pursue such a policy, neither could it be immoral.

This brings out the difference between the moral situations of creature and Creator: we cannot say with any exactness what the rights and obligations of a creator are under these circumstances, because God’s understanding of the situation, given His role as omniscient creator is something we cannot comprehend. It is a feature of something within a system, whose perspectives are limited, and who naturally does not have the comprehensive understanding of the whole system that its omniscient creator has, that it cannot hope to see the ‘big picture’. God’s answer to Job rests on this difference of understanding. The end of the book of Job, starting as follows:

and continuing almost to the end of the book is sometimes seen as a mere statement of the authority of power, but it is also about the command of detail. People are in no position to make an overall assessment of the situation. It is not simply that, as a subordinate we have no right to challenge the Authority, it is that it requires an overall viewpoint that he cannot possibly adopt. A human parent can make a rough distinction between those cases in which heor she should intervene and those in which the child should learn a lesson the hard way, but the situation of someone operating within a world the general nature of which is given, is essentially different from the position of a creator with absolute control who has to fix the frame within which other, lesser, rational beings will live. What constitutes the good thing to do in such a case is very difficult or impossible to assess from within.

The difference between ‘it is too complicated – there are too many factors’ and ‘the situation of the creator and the creature are not in all respects analogous’ is important.

We have a pretty good idea what evils are likely to be pointless or horrendous – if we did not, Rowe’s argument would lose its intuitive appeal, and a God whom we knew could not allow any pointless or horrendous evil could be relied upon to step in if we made the kind of mess of things that, we can guess, involve such evils, so we would surely end up ‘playing the system’ too. Unfallen human nature would not do that and would call on help only when really needed. It says in psalm 36/7 (vs 23–24):

Our steps are made firm by the Lord when He delights in our way; though we stumble, we shall not fall headlong, for the Lord holds us safe by the hand.

If taken literally, this is a utopian wish (as is much of that psalm), not strictly or usually true in the world of fallen people. It represents an eschatological perspective of a return to an unfallen state. ‘Knowledge of good and evil’, however, could give us the initiative. God, therefore, has little option but to step back in all normal circumstances and let nature take its course. (The Tower of Babel story has a similar moral; in a fallen world, the unlimited use of human ingenuity and cunning would lead to unlimited evil.)

It might be replied that this only applies to evils produced by human folly: God could refuse to write cheques but still stop people from being hit by bolts of lightning. But is there a simple line between earthquakes and diseases over which we have absolutely no control and bad things encouraged by our sin or foolishness? Could God always intervene in the former case whilst not allowing us to play the system in the latter? If we knew that God would always intervene to prevent the most terrible disasters, could that not lead us to be more careless when handling fire? Again it seems to me that the devil is in the details, and there is no way in which we could make any reasonable calculation on how this might be worked out. Anyway, the fact – and the Christian doctrine – seems to be that God withdrew his special providence as part of His normal operations in response to sin, and it is hard to see how we can put ourselves in a position to conclude with any certainty that this was an unreasonable or immoral thing to do.

A creator who made a world simply of suffering could not be similarly exculpated, because it is clearly not the case that such a thing would be a necessary condition for there being a world of such complexity that its free running would involve nothing but pain for creatures within it. What we can be legitimately sceptical about is whether a world less open to dangers than this one, could, without special intervention, permit a freedom like ours.

Establishing a nuanced framework for how a creation might work, free of any further normal context does seem relevantly unique. It is tempting to think that there is behind the rationale of the ‘problem of evil’ something close to a simple or naïve act consequentialist or act utilitarian understanding of what is morally acceptable – if you can imagine a tweaking that would move the sum slightly further to the good, then that would be morally obligatory for anyone capable of enforcing it. This is probably far too simple minded – and anthropocentric.

I am in fact making a distinction between those cases where the calculation (so to speak) is beyond us because there are so many things of which we are ignorant, and the situation where it becomes impossible to apply our moral framework because that is designed for operating within creation, not the ‘responsibility’ for creation as a whole. It is the latter that I think that the book of Job is looking towards.

Once again, it is not a blank cheque we are looking for – ‘our moral categories do not apply to God’ is not discriminating enough.

One possible response to my line of argument is that it misses the point of the current debate about evil. That debate now only claims to show that, given the kinds of evil that there are, it is more probable that (or more reasonable to believe that) there is no traditional God than that there is such, and I have not shown that my theist’s attitude to (A) and (B) is more reasonable than the atheist’s. All that would be necessary would be that, given other grounds for belief, the problem of evil is not so strong as to override them. It could be that if belief turned only on the one issue, then the atheist would have the stronger case, but, given other considerations, it is definitely not a knock-down objection.

Although some of the language and references used above is biblical, I have claimed that the basic argument is one that any theist might employ. What follows now is, however, distinctively Christian. I do not think that it is unreasonable – or ‘unphilosophical’ – to claim that the doctrines of a particular religion might serve to help solve a philosophical problem.

Someone might agree that we are not able show that (A) and (B) are in any way unreasonable, but still feel that such a policy exhibits a certain harshness, particularly for a Deity whose primary instinct is expressed in the thought that ‘though we stumble, we shall not fall headlong, for the Lord holds us safe by the hand’. Would not following my policy, even if it is the necessary one, represent what we might think of as tensions in the Godhead? Perhaps this might lead us on from a neutral philosophical theology to a specifically Christian answer, in the shape of the Incarnation and Calvary. The inevitable harshness of (A) and (B) ‘demands’ from God a very specific kind of identification with the situation that such a creation constitutes.

It might, perhaps rather surprisingly, be useful at this point to invoke a version of R. M. Hare’s notion of universalizability, as it appears in Freedom and Reason. He claims that when making a moral decision, we should take on ourselves, as if they were our own, the feelings of everyone who might be affected. As a proposal for human choice this is bizarre in a variety of ways, not least because realizing as if one’s own the psychology of all those possibly affected would be both impossible, and, if possible, totally crushing (Robinson, Reference Robinson1981). But on Calvary a man who was also God could and did take on the burden of feeling the whole cost of human sin and freedom. This was not God acknowledging any guilt, though you could say acknowledging responsibility for creating a world with suffering in it, as the only way of creating a world with free creatures like us. He always knew the cost of raising people to a higher status than Adam, and bore it with us. If you think it was not worth it, no argument can directly decide the issue, but your feeling can have no further ontological grounding.