As soon as you say in a public forum that you are agnostic, somebody will rejoin smugly, “Are you also agnostic about fairies?” as I said last post. Obviously I am not. I am also not agnostic about many claims that are undemonstrable but which are unlikely given our present state of knowledge. It is possible that there are aliens visiting earth, but given what we know about physics, astronomy, and biology, it is so unlikely as to be vanishingly improbable, and so I do not reserve judgement on that claim.
So it is clear that to be an agnostic, one has to index one’s doxastic attitude to specific claims. For example, if I am to be atheist about Thor, then I must subscript the “God” predicate to indicate this:
Believe[I](Not Thereexist(God[Thor])) [(BI(¬∃x(GThor))]
Obviously if I put it this way, my atheism regarding Thor commits me to no similar attitude regarding any other deity. I may reject all deities, some deities or no other deities, subject to logical coherence. I may be agnostic about other gods. I may even accept one or more of them. One’s doxastic commitment need not extend beyond the particular belief in question.
However, it usually does, for the reason that there are an infinite number of logically possible propositions one might have doxastic attitudes to, and we are limited in what we may consider reflectively, one by one. Instead we have to group our doxastic attitudes into classes and deal with them. This, of course, is to classify our beliefs. Many people who are atheists or agnostics have grouped beliefs into kinds, and the scope of their attitudes ranges over these kinds. How might that happen?
I introduced a simple set of icons or flags in the last post, in order to signify the various positions one might take. Here is a complete set:
The positive denial there exist gods is positive atheism, the denial that gods are to be believed is negative atheism. The statement that gods cannot be known to exist is strong agnosticism, that no knowledge statements can be made either way is weak agnosticism. Finally, a class of atheism that is basically the complement to a theist claim is what I call privative atheism. When you call yourself an atheist with respect to a claim of some kind, it pays to be clear what scope you are applying the name to.
Now, I may be a positive atheist about some gods a … x, and a weak agnostic about some other gods y … ax, and so on. What I would do in this case is to group the specific indexed claims into classes: those I am a positive or negative atheist regarding, those I am a strong or weak agnostic regarding, and perhaps I might find it worthwhile to group them in a larger class of privative atheists (infidels) for, say, political purposes. If you live in a predominantly theistic society like the US, India or Saudi Arabia, it might pay to overlook the various epistemic differences of the nonreligious in order to gain them a space within the social order.
Note: I do not regard these privative atheists as identical to “secularists”, because, as I have already argued, secularism is something that the religious ought to support in their own interests as well.
Perhaps one can come up with principles for including each indexed claim into one or more of these classes. Myself, I think that if a claim of belief involves the denial of factual information, then that is a pretty good reason for thinking that it should be positively denied. But beliefs that are not vulnerable to counterfactual argument, which are empirically inoculated, are not thereby right. They may instead be something about which I cannot say and must be weakly agnostic about. Some deities (for example, Spinoza’s deity, and maybe Leibniz’ as well) are like that. I cannot say they do not exist, nor that they do, and nothing factual makes their existence more or less likely.
So any particular person my adopt several of these viewpoints, depending upon what deity existence claims are being indexed. In people’s minds, though, it gets hard to maintain these separations. I shall return to the “field of contrasts” idea once more, in the next post.