I’ve been thinking about the question recently of whether it’s actually so bad the bite the bullet in ‘Super Awesome Evidence’ cases. Wanted to share some of those thoughts, though I’m going to present the case as strongly as I can which I think is probably more strongly than I think is reasonable atm:

There’s a sense in which ‘Super Awesome Evidence’ cases induce an ambiguous intuition. There’s no dispute, really, about the coherence of pagan gods, or of aliens, or other supernatural entities. Often, when people say ‘theism’ they mean something like perfect being theism. People are often highly triumphalist about the demise of logical arguments from evil, or logical arguments in general about theism. I think, even if you agree these arguments don’t work, you should be somewhat skeptical of the degree to which people emphasize their failure. In addition there are arguments against specific faiths, like the logical problem of the Trinity, as well as logical foreknowledge arguments (ironically I think foreknowledge arguments are terrible, way way worse than Logical POEs, which is probably the inverse of the academic consensus). If any of these arguments are right, there’s a sense in which you should think Super Awesome Evidence (SAE) cases don’t provide evidence. The same way nothing I see could convince me of the existence of square circles, nothing I see should convince me of the existence of a contradictory being. If the LPT goes through, the correct response to seeing Jesus of Nazareth rise from the tomb and proclaim his divinity is “You have to be confused, because that is impossible.”

So, as long as you think no empirical evidence should update you in favor of things which are impossible (this is obviously a trivial result of Bayes theorem in one sense, but you might have things to say about how you handle higher order evidence and if you should ever assign a prior of 0 to something) then so long as logical anti-Theistic arguments have a chance of going through, you shouldn’t think biting the bullet on SAE cases is that crazy.

I mentioned an ambiguity though, what did I mean by that? Well, some people mean something broader by theism. They mean something like “a being worthy of worship that created the universe exists.” Obviously, some of the beings covered by this hypothesis are immune to most or all of the logical arguments against Perfect Being Theism (henceforth, I’ll use PBT to refer to the sort of theism defined in the previous paragraph and Theism simpliciter to refer to this much thinner notion of Theism). Dialectically though, a debate over fine tuning may look like this: Theist (T) asserts that God is very simple, because the divine properties are just all properties maxed out, and predicts fine tuning data very well. Atheist (A) presents the problem of divine psychology. T responds with the SAE reductio. The atheist may feel an intuition pulling in the reductio’s favor because of the permissiveness of Theism as a descriptor, but notice, may beings described by Theism will not be simple at all. They will not enjoy the standard simplicity motivation for theistic hypotheses. On the other hand, as we saw, if T means perfect being theism then biting the bullet on SAE arguments may not be so bad. If you an atheist about PBT it’s not at all implausible that you take the hypothesis to be deeply defective in a way that evidence couldn’t support it in principle. Theism is of course far more modest than PBT, but it’s not as clear that it will as strongly favor the fine tuning argument as PBT would.

So if you’re an atheist, maybe biting the bullet on Super Awesome Evidence Arguments just isn’t that bad.

(As I say at the opening, this argument is presenting the case more strongly than I really believe, there are some places I’d want to push back here, but curious what thoughts others have so I wanted to give the boldest statement I could).