Let me start with a parable:
A guy named Mike has the following beliefs:
- "The sun will rise tomorrow."
- "Crows are black"
- "I have a brain"
- "The moon will turn purple tonight."
- "Voices come from humans, parrots, or electronic devices, but not rocks."
Then, out of the blue, he hears someone say that the reason we believe the sun will rise tomorrow is because it has risen every day for as long as anyone can remember. He thinks about this explanation and decides it makes a good deal of sense.
From there he realizes that the situation with crows and his brain is similar: no one has seen every crow in existence, but every crow anybody as seen, as far as Mike knows, has been black, so we infer that all crows are black. Similarly, no one has seen Mike's brain, but every human that has been autopsied has had a brain, so we infer that all humans do and therefore Mike has one too. Mike realizes that a lot of our beliefs are like that, and thereby discovers the principle of induction.
With this principle, he can realize that his belief about the origen of voices has some basis in evidence, but his belief about what the moon will do tonight does not, and in fact induction would suggest that the moon is not going to turn a funny color. So he gives that belief up.
How close much resemblance to reality does this scenario I just sketched out bear? A fair amount, I think. Many people become quite convinced of the rising and setting of the sun before ever thinking of such principles as induction, much less verificationism, falsificationism, and Bayesian confirmation. And many people carry around quite a few false beliefs in their world of true ones, even if the real world examples aren't so extreme. As far as I can tell, such principles of inference are almost never arrived at in a vacuum. To show the problem with non-falsifiable beliefs, it helps a lot to be able to give an example or two of a patently absurd non-falsifiable hypothesis.
I would go so far as to contend that for our hypothetical Mike, the only way to discover that the moon will not turn purple tomorrow is to begin by accepting all his beliefs, including it, as properly basic. Ultimately, all beliefs may be reduced to the general reliability of perception and memory, as well as some ideas about good rational inferences, these being the only things taken as properly basic. But even once this framework is worked out, it will always be useful to refine our notions of inference against "test cases" where the correct answer seems obvious. In all this process, though one belief may be occasionally traced to another, there must always be givens, or else we fall back into the problem of regress.
With all this down, I think one can understand what is wrong with Plantinga's use of proper basicality to defend the belief in God. It seems to me that Plantinga starts in the right place, but strongly resists any attempt to move beyond square one. To cite just one example from Faith and Rationality, which Carrier includes in his bibliography on Plantinga: on pp. 74-78, he deals with what he calls "The Great Pumpkin Objection," that if belief in God is allowed as properly basic, one might as well have a properly basic belief about the Great Pumpkin returning every Haloween. Plantinga proposes that inorder to determine conditions for proper basicallity, "We must assemble examples of beliefs and conditions such that the former are obviously properly basic in the latter, and examples of beliefs and conditions suchc that the former are obviously not properly basic at the later." Thus we can arrive at criteria inductively.
So far, not bad. Then, however, Plantinga goes off and says that someone who accepts his epistemology does not necesarrily need to work out "full-fledged criterion of proper basicality." This strikes me as lazy, but perhaps prudent. If one were to draw up a list of things which are not properly basic to believe, it would include the vast majority of the gods that human beings have ever believed in. This, among other things, would tend to suggest that belief in your god of choice is a bad candidate for proper basicality. When I see things like this, I get the feeling that Plantinga is anxious to assure the Mikes of the world that they need not critically examine their beliefs, that they need not try to beyond their common prejudices.