Here is a thought I’ve had about altering the standard structure of Reformed Epistemology while remaining true to the view’s original spirit. Suppose that religious belief is never basic, and a fortiori, never properly basic. That is, suppose that religious beliefs are always and everywhere inferred from other beliefs. To employ some of Plantinga’s examples, suppose I read the Bible. Instead of simply acquiring the spontaneous conviction that God is speaking to me, suppose I go from my belief that I am reading the Bible (and maybe some other beliefs about how personally pertinent or emotionally powerful my reading is) to the conclusion that God is speaking to me: not via some suppressed premiss along the lines of ‘If I am reading the Bible and…, then God is speaking to me’, but just by a direct intellectual movement from one the belief to the other (much as one does not complete a modus ponens inference via the suppressed premiss that modus ponens is a valid argument schema). So too, I observe the stars, and instead of spontaneously thinking that the stars were made, I infer that the stars have a maker from the evidence that there are stars.
Of course, a religious sceptic will think that these are terrible inferences, much as she will think that basic religious beliefs are unwarranted. But supply some substantive religious assumptions, much as Plantinga supplies his Aquinas/Calvin model, and these inferences turn out to be pretty good. To undermine the inferences, the religious sceptic will have to undermine the assumptions – just as she would to undermine Plantinga’s putatively basic beliefs. So we have considerable parity between the original Reformed Epistemology and my modification.
The point of all this, I suppose, is to suggest that Reformed Epistemology took a wrong turn in getting all hung up on the question of whether religious beliefs require evidence. There are other ways to upset the sceptic’s epistemological assumptions than by invoking proper basicality.
A more schematic way to think about. We have the set of all the theist’s beliefs, S. Hold S fixed. Let’s think of it in terms of a processing system involving input, operation, and output. The inputs are the properly/justified basic beliefs, the operation is the totality of justification-preserving inferences, and the outputs are the justified beliefs. The Reformed Epistemologist’s central point is that what the outputs are depends upon what the world is like (holding S fixed). Plantinga gets to this conclusion by arguing that what the inputs are depends upon what the world is like. I am noting that there is an alternative route to that conclusion, namely, to argue that what the operation is depends upon what the world is like. Of course, to do so you would have to say a lot about what a justification-preserving inference is, just as Plantinga says a lot about what a properly basic belief is.