Inquiry from Johnny-Dee on the AFR

Johnny-Dee wrote: Hey Vic, I'm currently taking a grad seminar on Descartes, and I've been grappling with the "problem of the circle." This has led me to consider whether AFR arguments have a similar circularity. For Descartes, the circle goes something like this:(1) I am certain that God exists only because I am certain of whatever I clearly and distinctly perceive.(2) I am certain of whatever I clearly and distinctly perceive only because God exists.How do AFR arguments avoid a similar problem of circularity. Of course, the AFR circle (if one exists) would be different, perhaps it would go something like this:(1*) I am justified in believing my cognitive faculties function rationally only because God exists.(2*) I am justified in believing God exists only because my cognitive faculties function rationally.With all of this stuff on Descartes, this gave me a perfect opportunity to raise this question, which I've been meaning to ask you.An excellent question, and one that I have come to terms with in my own work. I think I handle this issue better in my essay, “The Lewis-Anscombe Controversy: A Discussion of the Issues,” from Christian Scholar’s Review in 1989 than I do in my book. Or at least I shortened the discussion when I wrote my book. I cover it on pp. 58-60 of CSLDI. It has to do with the distinction between Skeptical Threat Arguments and Best Explanation Arguments. The idea is this: If we begin by raising skeptical questions about reasoning and argue from there to the conclusion that theism can refute skepticism but atheistic views cannot, then we run into trouble. If we, for example, raise skeptical questions about whether the law of noncontradiction is really a sound logical principle, and then we argue that if there is a God then God’s agency can justify the status of the principle of noncontradiction, then all we have to do is look at the fact that we are presenting an argument that is at the same time using the law on noncontradiction to see that we are caught in a circle. A Best Explanation argument takes it as a fact that we do reach truths through inferences. Only if our opponent, say an eliminative materialist, says “Well, of course, there’s no such thing as rational inference, but so what,” then we point out the disastrous epistemological consequences of denying that there are rational inferences. The AFR as I construe it does not justify our confidence in rational inference with an appeal to anti-materialist metaphysics. What it does is insist that both sides in the dispute agree that there is rational inference, but while one side is trying to explain the rational in terms of the non-rational, our side is not. So I would never answer the question “Why do you believe that modus ponens is good logic?” by saying that God has created me and therefore that belief is true. If, on the other hand, I am asked why I think I am able to recognize that modus ponens is good logic, then I might give a theistic explanation at that point. Steve Lovell's account in the linked paper on naturalism is very helpful as well. Johnny-Dee's comment is here.

A redated post. Click on the link for some comments.