Can you simply be Agnostic?

Well, I’ve argued elsewhere that Catholics get to self-define. So, I reckon, do atheists and agnostics.

The words “atheism” and “atheist” only entered the English language in the second half of the sixteenth century. Initially the concept of “atheism” embraced three things:

- Denial of the existence of god(s)

- Lack of any belief in the existence of god(s)

- “Practical” atheism - i.e. disregard of duties to god(s), regardless of belief.

At first, the sense of practical atheism seems to have predominated. Pretty soon, however, that meaning came to be mostly expressed by other words, e.g. godlessness, impiety. So “atheism” came to signify the two philosophical senses.

It’s obvious that those who deny the existence of god(s) are a subset of those who lack belief in the existence of god(s).

“Agnostic” and “agnosticism” don’t turn up for another three hundred years or so. In fact, we know who coined the words; it was Thomas Huxley, in about 1869. Huxley intended the words for those who (like himself) not only didn’t know whether god(s) existed or not, but argued that we couldn’t know this. He believed, essentially, that the existence of anything beyond or behind material phenomena was unknown and unknowable.

It seems that, to Huxley, “denial” atheism was a faith-based position just as much as theism was. In his view denial atheists were asserting as true something which could not be known to be true. (Huxley, perhaps, would have said that a denial atheist who completed the current census by ticking “other religion” and then writing “Atheist” would be correct.)

Very quickly, a large number of people who might previously have been described as atheists began to self-describe as agnostics. This included not only people who shared Huxley’s conviction that we cannot know if god(s) exist, but also those who simply felt that they themselves were uncertain about the existence of god(s) as well as, in time, a group who simply weren’t greatly interested in the question of whether god(s) exist. “Atheist” soon came to refer mainly to the hard-core denialists.

I suspect that what may have driven this fairly rapid change in meaning was the fact that up to that time “atheist” was mostly used in a pejorative sense. The newly-coined agnostics were seeking to affirm themselves and their position as socially and intellectually respectable by embracing this new label.

The end result of these shifting meanings is that we have two fairly clearly-defined philosophical positions for which only one label is appropriate - those who assert that god(s) do not exist, who are always “atheist”, and those who assert that we cannot know whether god(s) exist, who are always “agnostic”.

In between we have what I suspect is a larger group of people who we can characterise only in negative terms - they don’t have a belief in the existence of god(s), they don’t assert that god(s) do not or cannot exist, they don’t assert anything about the limits to human knowledge in this respect. They can be described (or they can self-describe) as either “atheist” or “agnostic” according to preference, or fashion, or to which of the other two groups they prefer to tog out with. Arguments about whether they are more accurately described as “atheist” or “agnostic” are not much use (except, perhaps, that they tend to cast light on the position of the person making the argument). (FWIW, my entirely subjective and unreliable impression is that, in recent years, people who lack any belief in god(s) have become more prone than formerly to self-identify as “atheist”.)

I think this does have some implications for discussions about how the non-religious ought to complete the census, or about what options the census form ought to offer them. Because the boundary between atheism and agnosticism is movable, getting people to classify themselves as atheist or agnostic is of limited value. People with identical positions might classify themselves differently, and changes from census to census in the numbers classifying themselves as one or the other might point to shifting positions, but they could equally point to unchanged positions which are simply being differently described or labelled, and we would have no way of knowing which phenomenon was at work. I suspect that information about the aggregate number identifying as non-religious is probably more meaningful than any attempt to analyse the non-religious between atheism and agnosticism.