Yesterday I debated good friend and fine Oxfordian gentleman Amos Wollen. It was a devil’s advocate debate; I lean towards the view that there is no God, Amos thinks there is a God, yet I argued that God existed and Amos argued the opposite during the debate. Over the course of the debate, Amos mostly presented the problem of evil, yet also presented the evil god challenge. I think this is an interesting and troubling challenge, yet not totally decisive.
There are two somewhat different versions of the evil god challenge, which I’ll call EG1 and EG2.
EG1
The evil god is very unlikely.
A good God, of the type posited by most theists, isn’t much more likely than an evil god.
Therefore, a good God, of the type posited by most theists is very unlikely.
There’s also:
EG2:
The problem of good (one that claims that our world is too good to be made by a maximally evil deity) is a knockdown objection to a maximally evil deity.
The problem of evil is as convincing as the problem of good.
Therefore, the problem of evil is a knockdown objection to a maximally good deity.
Often, those presenting the evil god challenge fail to distinguish between the two. My comments will be about EG1, however, many of them will also be applicable to EG2. I will say that I find each premise in EG2 pretty plausible, however, it’s not clear that they’re so obvious to make theism implausible. EG2 is just a way of showing how ridiculously evidentially significant the problem of evil is, but that’s not something I deny.
EG1 is quite a challenge but not, I think, an insurmountable one. Let me present 12 partial answers to the evil god challenge, which together, I think, comprise a potent cumulative case for theism over the evil god.
I think it’s very intuitively obvious that there is no evil god, while not so intuitively obvious that there is no good God. Many people find the notion of a good God plausible, yet virtually no one finds the evil god that way. This diffuses the challenge in two ways. First, even if you’re not super confident in any of the next 8 solutions, you should be confident that one of them works, just as you might not be sure exactly where the proof that 1=0 goes wrong, but you’re very sure it goes wrong somewhere. Second, even if there aren’t additional reasons to doubt an evil god, the fact that it just isn’t plausible gives a reason to doubt it. The claim that suffering is good has the same going for it as the claim that pleasure is good, except that it’s just not plausible at all. So even if evil god defenders can say similar words to good God defenders, that doesn’t mean the two views are equally plausible.
Some, like my friend Dustin Crummett, have argued that that there is some sense in which, considered intrinsically, God is perfect, the best way a being intrinsically could be. He suggests that it’s not clear that maximally powerful, and all knowing, but evil is the worst way a being intrinsically could be, even if it's the worst instrumentally. You might think that perfection carves reality at its joins more than maximal horribleness. I’m a bit dubious of this, but it’s not obviously totally crazy.
The problem of good is harder because more evil aims could be accomplished through deception. A good God is limited to achieving their aims without too much deception or wrongdoing. In contrast, an evil god has more options to bring about evils, for they’re unbound by moral constraints. There are more goods than bads on the objective list, so it’s more plausible that those goods would require evil.
I think we have basic justification for trusting that we’re not super confused, on both abductive grounds and from first principles. Any theory that tells you that you might be hallucinating the world is probably wrong. But on the evil god view, it’s much more plausible that we’d be massively deceived, for the God has no scruples about such deception.
An evil god is intrinsically weirder. If, like me, you doubt that bad people deserve to suffer, then you’d have to think that the evil god wants itself to suffer, to an infinite degree. But that’s clearly crazy. A being who loves the good is much more likely to exist than one who wants himself to suffer infinitely. A being who wants to infinitely suffer might be coherent, but if so, only barely. Alternatively, you might think that people deserve bad things, but probably only to a finite degree. This is even weirder—on this account, the evil god wouldn’t like the first few bits of suffering, but it would love it as its suffering becomes more than it deserves.
An evil god would almost definitely just make a torture world. Surely this world would be worse if we were just being tortured all the time, to an unlimited degree. This response is defended in more detail here.
An evil god has an arbitrary limit on rationality, while a good God has no arbitrary limits, assuming that, like Parfit, one accepts that full rationality is incompatible with being maximally evil. This lowers the prior probability, even if it stems from a simple property, just like it would be harder to be a theist if maximal perfection entailed being shaped like a boat and having 58 arms. Even if arbitrary seeming limits fall out of a more fundamental property, they still undercut the probability of a theory.
I’ve argued in various places that theism has a high prior probability, yet most of these arguments don’t apply to an evil god. An evil god is not similarly provable, maximal in regards to being an agent, or derivable from a single property like perfection (arguably there is no such property of anti-perfection).
A maximally evil god is a worse explanation of psychophysical harmony. An evil god would be more likely to make the world disharmonious, like a bad trip, where everything is chaos and we can’t learn or grow. Similarly, while it’s not obvious that they wouldn’t make moral knowledge, it’s not obvious that they would, unlike with a perfect being. Because there are more types of goods than evils in the world, and for all the types of goods in the world except pleasure vs pain, the good is more common than its evil opposite, it’s hard to see why a perfect being would make such a diverse array of goods available, even when many don’t have an obvious foil.
It seems like to provide moral knowledge one would need some degree of moral knowledge. But if one has sufficiently detailed moral knowledge, I claim, along with Parfit, that they would inevitably do the right thing, just as a person who is equipped with full understanding of the resultant pain wouldn’t stab a fork into their eye for no reason.
You might just think love of the good is more intrinsically likely than the love of evil. Good has a certain inherently attractive quality by its very nature, unlike evil. The default is liking the good, at least, the subset of the good that one experiences directly. That’s why so few people desire their own bad for its own sake.
If you like holding false views, you might think that evil is a privation of good and so there couldn’t be unlimited evil.
I don’t think any of these are single-handedly decisive (except maybe 1). But together they can construct a powerful cumulative case that is enough to reduce the odds of an evil god to be very low.
> Therefore, the problem of evil is a knockdown objection to a maximally evil deity.
Don't you mean a maximally good deity?
An all-evil God seems to be like solipsism: psychologically impossible to believe. I think theism is committed to two principles: (1) life is generally worth living, and (2) procreation is generally good. If there is an all-evil God, we should embrace antinatalism.