Robert Oerter has written an interesting post on his blog outlining what he calls a fine-tuning argument for naturalism. It’s important to keep in mind that Oerter doesn’t actually believe that this argument is a good argument for naturalism. Rather, he thinks it’s useful for showing what’s wrong with the fine-tuning argument for theism.
Rather than try to summarize his argument, I invite readers to simply read it for themselves.
What follows is a comment I left at Oerter’s site.
Remember that God is, by hypothesis, omnipotent. That means that God could have caused life to arise by miraculous means, even in a universe that was not fine-tuned.
Let us define ML as the hypothesis that God miraculously allows life to arise in a universe that is not fine-tuned.
Both ML and its denial (~ML) are logically compatible with theism. So I think the best way to evaluate the evidential significance of ML is to treat ML as an auxiliary hypothesis and apply the theorem of total probability.
Pr(FTU | T & K & L) = Pr(ML | T) x Pr(FTU | ML & T & K & L) + Pr(~ML | T) x Pr(FTU | ~ML & T & K & L)
What that formula shows is that, in order for the fine-tuning argument for naturalism to work, Pr(ML | T) must be greater than Pr(~ML | T). But it is far from obvious that that is the case. So what reason is there to suppose that Pr(ML | T) > Pr(~ML | T)?
ETA: And if Pr(ML | T) < Pr(~ML | T), then it’s no longer clear how this argument is supposed to show what’s wrong with the theistic fine-tuning argument.