BEATRICE: I'd like to pull you up on something before we go further. You've given a very scant account of what evil is. You've pointed to certain bad acts (like murder) and certain bad states of affairs (like natural disasters), but you haven't really filled in more than that. Don't you think you need to?

ADAM: It is important to find the right complete account of right and wrong. But it isn't important for the present argument. We are often more confident on what is evil than why things are evil. Murder might be evil because of the suffering it caused, the categorical imperatives it broke, or the vices it exhibited, but no one is in doubt that it is evil. So long as you agree with me that the examples I point to really are evil (even if you don't agree with me why they are), that's all I require for present purposes.

BEATRICE: But you do need to say there really are evil things, that in fact there is evil. Yet that's controversial, at least among atheists: many think all talk of good and evil is just emotion, or expression of cultural norms, or false fictions, or something else. Yet if they are right – if there are no objective moral values – then this argument is meaningless. So if you can't defend the belief that there are objective moral values from your point of view, can you really present the problem of evil to challenge my belief in god?

ADAM: I'm unsure whether there are objective moral values, but I'm granting them for the sake of argument. If there are no objective moral values, then there cannot be a morally perfect being – at best it would be ‘morally perfect by this set of norms’, or whatever. So either there are objective moral values, in which case a believer should respond to this argument; or there aren't, in which case God cannot exist.

BEATRICE: Alright. I accept that it appears the factual premise is probably true (and so God probably doesn't exist), at least at first glance. I'd like to take another stab at explaining why these appearances are misleading – to show that, on further reflection, these evils are necessary after all. In short, I want to try a theodicy.

I want to offer a couple of explanations at once (much like the problem, these have been going around since antiquity); theodicy need not be a single sheet, but rather a patchwork quilt. I hope that, combined, these explanations can provide good explanatory ‘coverage’ over all those evils that at first appear unnecessary.

The first is the free will defence. I think our acts would lose all value if we were automata governed by the laws of nature. Instead, we are creatures who have freedom to act. One of the costs of this freedom is that people sometimes do awful things. However, the prize of this moral significance is worth the cost.

ADAM: Isn't this position on free will really controversial? Loads of people think morally-significant freedom and being determined are compatible. If that is true, the free will defence doesn't get off the ground: God could have arranged the world so that we always do the right thing and we'd still be free.

BEATRICE: Right, free will is controversial, and I don't have a knock down argument against ‘compatibilism’ between determinism and moral significance. But if I can show a free will defence works, depending on a particular position on free will, we've made progress. It means the answer to ‘are these evils unnecessary?’ hinges on ‘is this take on free will right?’ As free will is controversial, we shouldn't be that sure this take on free will is false – and so we shouldn't be sure these evils are unnecessary.

Second is soul making. Evil can often be fertile ground for greater goods to spring up. Many who have suffered appalling things nonetheless are thankful for what happened to them, and ‘wouldn't have things any other way’. These people think that the evils that happened are outweighed by greater goods, and I think we should take their word for it. Further, evil allows the exercise of all sorts of moral virtue: in helping those who suffer, bringing wrongdoers to justice, and so on.

ADAM: I think there are two big problems with the enterprise of theodicy. The first is one of breadth. It seems pretty easy to find whole swathes of evil events which evade your explanations. Consider Sam Puttick dying from meningitis: no one was responsible for it (so the free will defence is out), and given his death and the suicide of his parents, the idea that this was soul making is implausible – it seems more likely it was soul-destroying.

BEATRICE: Fair enough, the explanations I have given do not cover everything. But there are many more I'd want to add. I'd want to say that there are important goods realized by having uniform laws of nature, and that justifies bad natural events like meningitis even the absence of soul making.

ADAM: This is the second big problem – that all these explanations just aren't plausible, even for what they are supposed to ‘cover’. I cannot see any value in uniform laws of nature over and above ‘generally uniform but imperceptibly tweaked for our benefit’ laws of nature, and certainly no benefit worth evils like the Putticks'.

Now, there are further things you could say here, perhaps ‘God tweaking the laws would reveal his presence, and that isn't on’. But that reply doesn't get very far either: for God could have stopped Sam getting meningitis and none of us would be any the wiser – it wouldn't have been evidence of God is someone didn't contract meningitis aged five. (Besides, perhaps we should expect God to make his presence abundantly clear to us.)

There are similar questions we could raise against the other defences. We could ask why God couldn't just stop us once we've (freely willed) a given evil, instead of letting us carry it out. Or why such horrendous evils are needed for soul making, given many people develop extraordinarily virtuous characters without suffering significantly. I'm not saying these questions cannot be answered, but the answers themselves to give rise to equally problematic questions. I don't think the collection of all the different branches of replies and counter-replies makes theodicy more convincing – just more contrived.

Also these concerns can be flipped any way you please. It seems we could use these theodicies to defend a perfectly malevolent God. This God needs free will for moral significance, and so he puts up with free acts of good as the price for significant evil. Likewise, he allows good things to be done only to provide further opportunities of soul-ruining (I'd say the Putticks' would be an ideal case for such a soul-ruining anti-theodicy). That an explanation can be ‘flipped’ – that it can be used to defend the inverted hypothesis as well as the hypothesis – is a poor explanation.

Finally, these concerns are alien to our moral decision making. Consider this:

An evil genius cackles maniacally: ‘My decades of work have finally come to fruition! I can wipe out HIV at a stroke! I will deprive all sorts of opportunities for doctors and research scientists to exercise virtue! I will deny millions of people the soul-making opportunities that come from a progressive life limiting condition! I will deny people the significant freedoms of whether to inflict discrimination! My name will be immortalized in infamy!’

This seems implausible. Although the evil genius is right in that a world without HIV would lack these goods, everyone things stopping HIV is still on balance a really good thing. The reasons appealed to in theodicy just don't seem powerful enough to permit any ‘real life’ evil.

BEATRICE: I think there's another response which can defuse this argument. I'll accept (for argument's sake) that there are many evils that appear unnecessary, even after careful reflection and the best theodicy has to offer. However, this provides no support at all for the claim there really are unnecessary evils. Instead of trying to rebut appearances, I am going to undercut them.

ADAM: How does that work?

BEATRICE: Some appearances are misleading. Suppose I'm up in an aircraft and a look down upon a field. It appears there are no insects on the field. Yet it would be mistaken for me to conclude from this that there are no insects on the field (or me to use this appearance as evidence for there being no insects on the field). From high up, a field will appear to have no insects on it whether there are insects on it or not, so I shouldn't trust these appearances. If God exists, his insight into the moral world far exceeds our own. There will be all sorts of reasons for evils which will lie beyond our ability to see, but not his. We should no more expect to be able to see the necessity for evils than a chess novice to understand why a grandmaster made a certain move, or a primary school student to follow a physics professor's proof. As such, there will appear to be unnecessary evils whether there are any or not. So we shouldn't trust these appearances either.

ADAM: Might this scepticism go too far? If we can't take the appearance of unnecessary evil as evidence for there being an unnecessary evil, then it seems we cannot trust any moral decision me make. It might appear to me that punching you in the face is evil, and unnecessary for any sort of greater good, but what do I know? There are all sorts of goods beyond our ken, so I should say that although it appears wrong for me to punch you, I have no idea whether it really is wrong or not. We are morally paralysed: we have no idea whether anything we do really is good or evil, however they appear. This seems crazy, so if sceptical theism leads to this, we should reject sceptical theism.

BEATRICE: Well, I definitely don't want to be morally paralysed. Particular sorts of Theists can draw upon other sources of moral knowledge: perhaps I know the right thing to do from understanding the Bible, or from the inspiration of the holy spirit, for example.

ADAM: I think this will just push the problem one step back. How can you trust these sources? Perhaps God is using these to mislead you – for reasons beyond your ken.

BEATRICE: Yes, that seems problematic. I'd want to say there are some reasons for believing God would be honest with us, but short of an analytic proof, whatever supporting evidence I can find could be part of God's elaborate ploy to trick us (for reasons beyond our ken).

It would be nice to have a sceptical principle with selective toxicity: one that rules the appearances of unnecessary evil untrustworthy, but keeps other appearances unscathed. God is in a different position to us: some things might be permissible for him that aren't for us (and vice versa). Could we use this to be sceptical about judging God's actions, whilst still judging our own?

ADAM: I'm not sure. There are differences between us and God, and these differences might give rise to important moral concerns – concerns we can't comprehend. The problem is there isn't much more we can say than that. So whatever God's reasons are for permitting the Putticks' deaths, we don't know whether they apply to our circumstances or not: even if it was right for God to have permitted what happened, that doesn't help us answer whether it would have been right for us to. I struggle to see why we should believe there are all these big moral concerns beyond our ken – but they only apply to excuse God's apparently unnecessary evils. It begins to sound like: ‘Feel free to use your moral faculties, except when they lead to adverse consequences for these beliefs I want to keep.’

There's another problem with sceptical Theism. I accept that our moral faculties aren't that great: many times we won't be able to work out the answer, and sometimes we will get things wrong. But you are asserting something far stronger than fallibility: that the moral world is obscure to us – so obscure that our moral faculties are no use at all. That's implausible.

Let's borrow the chess analogy to make this clearer. I'm rubbish at chess: show me some complicated mid-game scenario, and I would have no idea who is winning. Further, I have some insight into my inability: even if a given scenario naïvely appears to be favouring black or white, I'd still say I didn't know, because I know I shouldn't trust my faculties here. But that doesn't mean all chess is obscure to me. I can tell if white has black in mate, for example. And I think I have rights to be pretty sure that ‘black is winning’ in some simple situations, like when black has all his pieces but white only has his king and a pawn. So too morality: many things will be obscure to me, but some things are simple enough for us to figure out.

BEATRICE: I can only disagree. It seems to me that (but for the grace of god) we have no hope of making the right moral decisions.

ADAM: Alright, but moral obscurity isn't necessary for belief in God, and so it is an extension to the God-hypothesis: instead of ‘God exists’, we have ‘God exists, and things are morally obscure’. In sceptical theism, it is this claim of moral obscurity that is doing all the work fending off appearance of evil.

The problem is adding additional beliefs to your hypothesis to explain away troublesome data is irrational. One explanation you could give for what happened to the Putticks' is this: ‘Don't worry, God made the world last Thursday! So none of evils (like the Putticks’) before then really happened.' The addition of ‘the world was made last Thursday’ to ‘God exists’ does explain the Putticks' story and make it unproblematic, but the problem migrates elsewhere. The extended hypothesis ‘God exists, and the world was made last Thursday’ is implausible. I think the ‘moral obscurity’ extension is similarly problematic.

BEATRICE: I agree that bolting-on additional beliefs ad-hoc to explain away evidence is irrational. But moral obscurity isn't some ad hoc bolt on to God like ‘God made the world last Thursday’ is. It is instead part of the package deal of my conception of God. If you don't buy that, fair enough. But unless you can show that moral obscurity is flatly implausible, or that I'm mistaken and belief in God should lead us to deny moral obscurity, then I don't think I should change my mind.

ADAM: Let's map out where we are so far: the square brackets put premises together, the arrow shows how premises go to a conclusion, and flat arrows show objections:

To start with, I gave an argument from evil. You accepted the ‘theological’ premise, but asked me to support the factual premise – that there are unnecessary evils. I appealed to appearances: there appear to be many unnecessary evils, so it is highly likely there are at least some unnecessary evils.

You gave two objections to this. First, you offered a theodicy – that although evils appear unnecessary at first, further reflection show them to be necessary. Second, you offered sceptical theism. Appearances of unnecessary evil, like the appearance of an insect-less field when you look at it from an aircraft, are untrustworthy, and should not be counted as evidence. In turn, I counter-objected to these responses.

BEATRICE: Perhaps we can put theodicy and sceptical theism together into a sort of combined attack on the argument from evil. It would go something like this. Firstly, theodicy would serve to ‘muddy the waters’, and make us less confident that evils we identify as unnecessary really are unnecessary, and show there are fewer evils that appear unnecessary than we first thought. The sceptical theism part would lead us to lower confidence in our faculties, so that sometimes we will miss reasons why evils are necessary. If we can show the proportion of evils that appear unnecessary on reflection is about the same as the proportion of cases we can expect our faculties to miss a reason why an evil was necessary, that would defang the argument from evil.

ADAM: That seems about right. For example, if you think our ‘necessity sensor’ will fail to find a reason about 10% of the time, if there were nine evils that appeared necessary for every evil that appeared unnecessary, that wouldn't pose any problems for Theism, as you'd expect to see that proportion of appearances of unnecessary evil if there were really no unnecessary evils. Even better would be demonstrating a track record of Theists ‘filling in’ explanations for evils previously thought to be unnecessary.

But those are very big ifs. Because I don't think there's been a satisfying explanation for any evil. Less ‘tricky’ examples: a case of depression, or a violent crime, or an animal being killed strike me as cases where explanations remain inadequate – just not as inadequate as the hard cases. Nor do I see a track record for ‘filling explanations in’. The Lisbon earthquake that perplexed philosophers 250 years ago is not discussed anymore, not because it was ‘solved’, but because there are recent perplexing cases instead. Instead of the ‘problem of evil’ comprising a few hard to explain cases, it strikes me as most cases of evil being hard to explain, with a few cases which seem pretty much impossible.

BEATRICE: Perhaps so, but God might be a good explanation of other phenomena, such as the nature of morality, the beginning of the universe, and things like that.

A small selection of arguments about God.

ADAM: True. We've only looked at one argument. It could be that the argument from evil provides really good evidence against God, but other arguments (from fine-tuning, morality, etc.) provide even better evidence for God. Short of a deductive argument where we are certain of all the premises, we need to look at all the arguments, pro and con, see if any of the work, and weigh up the ones that do as best we can.

After all, you could flip the argument from evil around:

1. 1) God exists

2. 2) If God exists, unnecessary evil does not exist

3. 3) Unnecessary evil does not exist.

If your reasons for believing God exists are better than your reasons for believing there is unnecessary evil, you should deny there are unnecessary evils. And, of course, vice-versa.

BEATRICE: In some strange sense, you either believe in evil, or believe in God.