Last updated on 18 Sep 2017
Or is that Grauniaded? I always get it wrong.
Andrew Brown, who writes, among other things, a blog at the Gurainad newspaper, linked to my comment that “religion” was a mess of pottage and not a single unitary phenomenon. He made some comments that we have often rehearsed on this blog, and similar comments were made also on his post. One in particular strikes me as underpinning a serious mistake about agnostics:
JSWilkins: I self-identify as an agnostic, and although I am definitely atheist about a good many gods, I refuse to assert that no gods exist, as that requires that I know things I do not know.
Except of course that isn’t what atheists assert. Atheists, at least this one, maintain that there is no evidence for the existence of god(s), in the same way that there is no evidence for the existence of fairies.
Are you also agnostic as to the existence of fairies?
Absent any evidence, there is no difference between fairies and gods. Of course, if you do have some evidence you can point us toward, then that would be a different matter entirely.
This seems to me to be a basic error of epistemology. I have already said why I think that atheism is a denial of the existence of a god, and agnosticism is the refusal to say one way or another that a position can be taken about the existence or otherwise of a god. Here the issue is whether as an agnostic I must be agnostic about all potentially supermaterial beings.
Every claim to know X, whether X is a claim there is a god, or there is no god, or no evidence for god, is relative to the particular god being proposed. Likewise X can be about other supernatural beings, such as fairies.
Each supernatural being (SB) has some posited properties. If an SB has properties that localise it in a place it is not observed, or involve it, and only it, being the cause of some phenomenon that we can fully explain naturally (like thunder or volcanoes) then we are fully justified in claiming that X is not existent. In short, if your SB means that you should find it in the bottom of the garden, or making thunder, and you don’t, then you have disproven that entity.
But for suitably empirically innoculated SBs, in other words, ones that either cannot make an empirical difference in their posited properties or have not yet been shown to make an empirical difference, you ought to remain undecided. The claims for each SB are hence indexed to that SB. If I decide that Thor is not real, I haven’t thereby decided that, say, Ahura Mazda is unreal. That is a different SB.
Now you might try to argue that if an SB exists only in the gaps in our empirical (i.e., scientific) knowledge, it is a pretty uninteresting sort of posited entity, and I would agree: gods of the gaps are weak and uninteresting (because there are an infinite number of states of SBs that can be consonant with gaps in our knowledge). But if you think that because you have shown they are uninteresting that they can be said with any degree of confidence not to exist, that is an error of inference.
In any case, the sort of gods that are not sensitive to empirical evidence cannot ever be shown not to exist, so long as their descriptions are not self-refuting (as, I suspect, all-knowing, all-powerful, all-benevolent deities probably are). For example, Epicurus’ gods, which are so far above us and so uninterested in human affairs, and which merely made the atoms in the void exist, are beyond disproof. Likewise, Spinoza’s deity is beyond disproof. So, too, is a suitably “abstracted” Thor, where his role is not to make thunder, but (say) to make us stand in moral awe of the natural world. These sorts of deities are the kind one must be agnostic about.
So the counterargument, often made to me by atheists, that if I am an agnostic I must be agnostic about all gods and SBs is a failure to index claims to specifics. I am atheist enough for most atheists – I certainly do not think that the Christian, Jewish or Islamic deities are existent, because to believe that involves either believing contrary-to-fact claims about the universe (and us) or involves reinterpreting them in Spinozan fashion to not be what they traditionally (and doctrinally) are. But I remain an agnostic in general terms, and so far as I can see, there is no way that could ever rationally change. There will always be conceptually possible SBs that are beyond empirical, and hence scientific, debunking and disproof.
Incidentally, as to the nature of “gods”: I apply the Greek Pantheon Test, of my own devising: If an entity would count as a god in the Greek pantheon, then it’s a god. Sprites and fairies count as gods. In the Greek religious traditions, which are anyway rather complicated, what counts as a god depends on what gets its own cult of ritual worship, which is why the Titans are not gods. Otherwise, gods are SBs simpliciter.