I have said before that there are three kinds of Atheism; Agnosticism, positive Atheism, and some kind of positivism according to which the propositions involving the word ‘God’ are unintelligible. However, apparently Norman Geisler distinguishes between two kinds of Agnosticism, and this distinction is one upon which I would like to briefly voice some thoughts. He distinguishes the ordinary agnostic, as somebody who simply says they do not know whether God exists, with a kind of epistemic-agnostic (my term not his) who maintains that we cannot in principle know whether God exists.
The burden of proof obviously has to be carried by the positive atheist, and likewise the burden of proof has to be carried by this Agnostic, and it seems particularly heavy. While the positive atheist seems, at first glance to be making a stronger claim, the epistemic-agnostic has to maintain that we cannot possibly ever have a true justified belief that God exists. Thus, there is no logically possible world in which we can come to know that God exists. More weakly they could perhaps say that there is no realistic standard of evidence which could be satisfied for belief in God (though how one could maintain this is beyond me). They, it seems, would be committed to the epistemic principle that we should ordinarily presume that we cannot know that which we do not know that we can know – we do not know that we can know that God exists, therefore we should presume that we cannot know that God exists. In other words, we should assume that unless we can demonstrate that we know something we should presume that we cannot know it – and this is a principle even stronger than the sceptical principle that we cannot know what we do not know that we know. [Edit: ok, it probably isn’t stronger than the sceptical claim, since it is really a claim about methodology]
This brings up an interesting thought about absence of evidence and evidence of absence: I think it should be agreed upon that absence of evidence counts as evidence of absence if and only if the evidence which is absent is evidence would we would expect to find if the hypothesis were true – thus we can say that (this example comes from William Lane Craig’s podcast reasonable faith) we have evidence of the absence of fairies insofar as we have no evidence of fairy corpses or clothing factories or whatever we would expect to find if they were real. I think there may be something wrong with this (I’m not sure), since it seems I want to say I am justified in believing that there is not now in the room with me an invisible intangeable (to me) yet corporeal spaghetti monster for whose existence no physical/empirical evidence would be expected. I think the same problem arises in examples like the Grue paradox. Thus we should find a way of saying that there are epistemological canons which justify our believing that some things are not true (such as being a brain in a vat) just in case there is no reason whatever to suppose that those things exist.
Another interesting observation is that the epistemic-agnostic who maintains that we cannot ‘know’ such things as whether God exists could still be a Theist existentially. We have examples of such in people like the canonical empiricist John Lock, or the famous existential philosopher Søren Aabye Kierkegaard. Paradoxically then, the epistemic-agnostic could be a Theist existentially; it would essentially be Fideism.