I've been watching a bit of Christopher Hitchens lately. I was slow to warm to him because up until a few days ago I'd only seen a few sound bites--most of them from QandA sessions or open-form debates--and his overly aggressive rhetoric turned me off a bit. In my experience the closer someone strays to the ad hominem, the less likely they are to be useful. However, I came across an extended lecture followed by a non-threatening set of questions, and I was won over a bit.
Pardon the strange file format.
I watched it a few days ago, and I'm trying to recollect some of my impressions. If I'm off a bit, I apologize.
First I'll say that a large part of what drew me in was his disciplinary background. Unlike Dawkins--who too can be a bit more abrasive than is helpful (that is, the strength of the position speaks for itself, and those who fail to grasp it won't be helped along by abuse)--who is a scientist, and who I believe teaches something like science education; Hitchens' grounding is in philosophy. He was at Oxford or Cambridge or some such during 68, and belonged to a Marxist group, and knows a fair thing or two about Hegel's and Marx's critiques of knowledge and religion.
So his approach to critiquing religion has some similarities to my own, and as such it is easy for me to like him. The thing coming to mind now that bothered me a bit about his discussion is the question of Agnosticism. Like many so-called militant atheists, Hitchens sees agnosticism as giving too many concessions to religion. Atheists, again if I'm piecing this together correctly, see that there is no serious evidence pointing to the existence of the divine, and conclude that lacking evidence the most reasonable thing to presume is non-existence. There is something safe about this (and I suspect something both too easy and too timid), and exploring it by analogy lends it support, if of an only weakly logical sort: there is no evidence that lettuce sings and dances when unobserved, and so we are inclined to believe it does not.
However (and here is where I would unfetter myself from Hitchens' fundamentally humanist moorings), each analogic comparison elides an important distinction when discussing the divine, and that is divinity. The divine as it is often defined occupies a different category from the formal and material. (Though we do need to take care: we are talking about the divine as it is defined. When we say "divinity" we have to acknowledge it as always already not divinity itself, but a more or less reified and entrenched construction. But we will leave this problematic aside for now.) Popular synonyms include the infinite (Kiekegaard, particularly) and the supernatural.
So we can say we know that lettuce does not sing and dance while we're away because when we do see it it is constantly and consistently vegetative. Any other conclusion leads to paralyzing absolute empiricism (and backgammon). But the divine, as supernatural, is not something we observe most of the time and not others, but something we observe not at all. As a category it is something for which we have no sense-able experience. either positive or negative.
Indeed, where we to encounter the divine in an empirical fashion, were the divine to become quantifiable or qualifiable, the breach of category would dissolve divinity into materiality. The supernatural would be merely natural, and ostensibly science could take care of the rest. Divinity, or the infinite in this loose sense of the word, is not only illusively unvarifiable (like the singing lettuce), but structurally so (which gets into some serious problems regarding omnipotence).
If the atheistic thesis is correct, then the unvarifiability of the divine, which is a categorical necessity, demonstrates its non-existence (or at least impels us to claim non-existence). But knowing non-existence violates unvarifiability, which in turn disintegrates the category of the divine (this is the equivalent of saying that science claims as its domain not only all knowledge, but everything; those categories which operate outside this particular economy are not merely problematic but invalid). In effect, those things which elude empirical verification are removed from the equation not merely on the ontological level--which is problematic to consider already, given the conflict between, to use our current example, the divine and formed matter--but categorically: we cannot explore the existence of the divine because we are denied the condition of possibility.
But agnosticism of the sort I practice takes this double-bind to heart. We have, for better or for worse, a concept of the divine. Because the logic of divinity operates outside that of scientific empiricism, it can be correct neither to claim a set of religious beliefs nor to assert the non-existence of God. Simply put, if we are going to continue to use the category of the divine (and saying "atheist" does just this) we cannot too say that the divine does not exist due to lack of empirical evidence. This is a category error, just as surely as we are in error to look upon the beauty of nature or the cosmos and proceed from there to wax religious.