I could do a 50/50 split over 250!
This seems like more of a nitpick than an important objection to the C&R definition. After all, if instead of C&R to determine atheism you use instead, "Capable of considering", well that is just a sort of blanket statement covering all religions. How can you (meaningfully) say that a person is capable of considering without some way to see if they have considered? So it seems to me that the "C&R" definition extrapolates from one or some religions to all religions, whereas the "Capable of considering" definition interpolates from all religions to apply to any specific religion. Both are logically valid, but both are extremely general. A person who is capable of considering sun worship* is not necessarily capable of considering Orthodox Judaism.
*To invoke the ghost of Iacchus.
Great Ghost of Dionysus Himself! You have indeed (perhaps accidentally) stumbled upon a circularity problem in what I was asking of Claus... I need to clear it up, and simultaneously clear up your extrapolation question, which is indeed very germane.
"Capacity for considering", of course, is the circularity I am speaking of. I asked Claus to demonstrate "capacity", not merely "potential", for understanding in infants. Quite obviously, neither of these can be demonstrated except circularly, inferred from the actual behavior of the infants. To say that an infant has the "capacity for considering" simply means that this infant has been observed considering (this is a hypothetical, of course--what we typically call "considering" includes the public or private verbal comparison of two or more options). So, instead of asking for evidence that infants have the capacity to consider, I should have simply asked for evidence that infants do consider.
Of course, the problem is considerably worse when we look for the "potential for belief" in a developing infant. The circularity is the same--we know that something has the potential for becoming X only by observing it eventually become X--but the inference is considerably delayed, and the "potential" becomes an article of faith. Consider the discussion over the infants "potential to become literate": certainly, not all adults are literate; how, then, are we to know that a given infant has the potential to become literate? One could make the argument that even the illiterate adults still have the "potential to become literate"; of course, this is also a statement of faith, rather than a useful term. The only way to know that they have that potential is to actually observe them becoming literate...by which time, the "potential" is irrelevant. (Yes, Tricky, it is exactly like Iacchus's "path of least resistance" that was always inferred after the fact, and claimed as the only possible path. Useless.)
Now... on to your extrapolation bit. You are quite right--I am using this "capacity of considering" in a blanket manner--like Aristotle's "at the time when by nature it would". I would still argue (and, in fact, I think I will) that this is logically more tenable than the C&R--all the privative needs is the observation of the characteristic in similar others. (You may now quibble on precisely how similar different people can be.) To have privative atheists, all we need are believers (I suppose all we need is one--of any religion at all--to define the term. logically, if not practically.). We don't actually need to have a demonstrated "capacity", or generalization to other gods, or anything of the sort; it is a very simple definition.
Your "a person who is capable of considering sun worship..." bit is nice, but it is a problem only for C&R, not for privative atheism. The simple fact that somebody else is a sun-worshipper is enough. Practically speaking, the privative sun-worship-atheist need not have considered sun worship at all, but simply be part of a population with at least one sun-worshipper.
But as nitpicks go, I think we can give you this one. Yes, I was using a blanket statement. (feel better?) It's just that the "blanket statement" problem is a real problem for C&R, since "considered" is part of the definition, and essentially irrelevant for the privative definition, since if we did not have the observed characteristic in some member of the population, there would be no need for the privative in the first place.
