Rebecca Watson, unfortunately, reminded me of a meme from the old days of new atheism. It’s those agnostic atheism diagrams.
I have a rant in me about these diagrams. Agnosticism and atheism are political terms, and whether you identify with them has more to do with what you find useful in your social context than the literal definitions of these terms. This diagram became popular because it explains and justifies a particular choice in identification labels, but it is not an appropriate framework to understand more broadly why people do or do not identify as agnostic. Thus, as a meme, this diagram is a barrier to empathy and understanding of fellow nontheistic folks of diverse label preferences. Also it’s just kind of incoherent.
Why agnostic atheism?
Before I criticize the meme, it’s important to first understand what the meme was intended to do, and why it was effective.
When I started identifying as atheist around 2006, I pored over a variety of introductory articles on atheism. These articles were greatly preoccupied with explaining the definition of atheism and adjacent terminology. I too in my early blogging years was very preoccupied with this subject. And one of the biggest things that needed to be explained, was why you don’t need to be absolutely certain there is no god in order to be an atheist.
Today it’s taken for granted that atheism does not require certainty about god, but that’s precisely because decades atheist advocacy have been so effective at persuading people of this fact. And it’s not really a matter of “fact”, is it? To some extent the definition is whatever we say it is. When religious folks assert that atheism requires certainty that there is no god, it may be a self-serving way to give atheists as little space to exist as possible, but it’s still a correct description of what “atheism” means to them. It’s not a correct description of what atheism means to atheists though.
Not all nonreligious folks agreed on the definition of atheism. Some people agreed with the assumption that atheism required certainty about gods, and therefore identified as agnostic, not atheist. And in fairness to those folks, there were a lot of different competing labels early on, and the people who advocated the “agnostic” label weren’t obviously any less reasonable than those who advocated “bright”. Nonetheless, for those seeking to politically mobilize as a nonreligious coalition, “atheism” had become the most powerful rallying cry. Within the atheist movement, agnosticism was a flame bait topic, as people would regularly accuse agnostics of being overly meek fence-sitters.
There were a variety of ways to explain how atheists didn’t need to be absolutely certain. In introductory articles, it was common to make distinctions between positive and negative atheism, implicit and explicit atheism, weak atheism and strong atheism. Even in 2007, I had the impression that these distinctions were overly loaded and somewhat archaic, but I guess you can still find articles about it today. Another approach was Dawkins’ 7-point scale, where a 1 is a theist who knows there is a god, and a 7 is an atheist who knows there isn’t a god. Dawkins said he was a 6, and most atheists would be in that vicinity.
But I think the most effective explanation came from the simple observation that atheism and agnosticism are not, as conventionally assumed, mutually exclusive. This is easy to see because there is a clear distinction between believing there is no god, and knowing there is no god. Furthermore, it is easy to graphically illustrate with a 2D diagram. The diagram conveys the idea that atheism and agnosticism are two orthogonal concepts, and that you can be one, the other, both, or neither. It also suggests a spectrum, which is important if you want to provide space for a bunch of misfit individualists who can’t agree on definitions. As long as you don’t think about it too hard.
Let’s think too hard about this graph
The “classical” way of understanding agnosticism, is as a state that’s between theism and atheism. Dawkins’ 7-point scale serves as a good example of this framework. 1 is “strong” theism, 7 is “strong” atheism, and 4 is agnosticism. Agnostic atheism might describe the range of around 5-6. Dawkins’ scale could, hypothetically, be embedded into the 2d diagram like so:
But the 2D diagram obviously allows for so many possibilities not included in Dawkins’ scale. For example, there’s this whole upper region of the graph not covered by the scale. I could be in between atheism and theism, while being completely gnostic. But we have to ask ourselves, does this make any sense? Not only is it hard to believe that anyone would hold this position, it’s hard to believe that we would recognize it as gnosticism. If I told you that I think we can know whether there is a god, but I don’t believe one way or another, would you exclaim, “Wow, I used to think you were agnostic until now!”
The incoherence of gnostic (un?)theism puts to lie the idea that the 2D diagram is offering a superior framework to a simpler 1D scale. If we’re being honest, the diagram isn’t a 2D spectrum at all, it’s just a diagram of four distinct categories. It might as well be a Venn diagram.
Now let’s look at more examples of the diagram.
In these diagrams, agnosticism is defined as:
- “Doesn’t claim to know that (no) god exists”
- “It is not possible to be 100% certain”
- “God cannot be known”
- “Doesn’t claim knowledge”
- “can’t say for sure a god exists”/”I do not know for certain”
- “I am certain that there are no gods / at least one god exists”
Well we got 6 random diagrams from a google search, and at least 4 distinct definitions between them. Knowledge is not necessarily the same as certainty, which is probably also different from “saying for sure”. And there’s a difference between claims about your own knowledge/certainty, and claims about whether it is possible to know or be certain.
These definitions raise many questions. What does it mean to be between agnosticism and gnosticism? Is it a scale from 0 to 100% certainty? Is it a claim that it’s possible to be 50% certain? Is it a claim that it’s 50% possible to be 100% certain? Is it a claim that you only sorta know whether god exists or not?
And then there’s the question of what it means to be between atheism and theism, if it is orthogonal to certainty and knowledge. Does that mean that you don’t believe either way? How can you be certain or uncertain about a belief you don’t even have? Maybe it’s a belief that only there is only half a god? Or that there’s a 50% chance of God? How is that distinct from 50% certainty?
And some of these may be legitimately interesting questions with real answers to them, but it sure doesn’t seem like that’s the goal of the graph. The goal is, as I said, to establish four distinct categories, and just sort of hint at a spectrum without having to actually address any of the questions that raises. I have to conclude that the only reason this graph is effective is because most people are completely incurious about graphs.
As it happens, I’m somewhat of a world expert on terrible graphs of orientation. The agnostic atheism graph strongly reminds me of another terrible graph proposed by psychologist Michael Storms in 1979:
Well, it’s not that terrible, it’s just mediocre. But if you look at the other terrible graphs in my article, you might find that there are many ways to disrupt its assumptions.
The first way, that seems to come most easily to most people, is to add more axes. For example, earlier I pointed out that there were at least 4 distinct definitions of agnosticism, so why not depict those as four distinct axes? (It’s obvious why not, but humor me.) Alternatively, we could add many other identity labels to the graph, such as freethinker, skeptic, humanist, secularist, and bright.
What’s that you say? Bright is sort of a loaded term? Just because you don’t believe in the supernatural doesn’t mean you’re a “bright”? What’s next, will we make an axis that tells you whether or not you’re “godless” based on whether you are without gods? I totally agree, “bright” and “godless” are too politically loaded. Putting it on the graph would be quite a Choice.
But! Agnosticism and atheism are also too politically loaded! I’m not exaggerating, I really mean it. I, who just wrote a short history slash rant about the politics of agnostic atheism, all off the top of my head–yes, I, the very same person–happen to think that agnosticism and atheism are politically loaded terms.
And what if Alice presents a 2D graph, and Bob decides that it really ought to be a 8D radar chart full of things that Alice wasn’t even thinking about? Where does Alice fit along the 6 additional dimensions? What if Alice asserts that in fact she still prefers it with only 2 dimensions? You too can experience what Alice is feeling by trying to fit yourself into the tertiary sexual attraction spectrum. As the saying goes, you can lead a horse to an identity chart, but you can’t make the horse identify with it.
The agnostic atheism chart has a lot of makings of a chart that not everybody is going to agree with. There’s the highly dubious horizontal and vertical “spectrums”. Politically loaded terms treated as neutral ones. At least four different definitions of agnosticism treated as interchangeable when they’re clearly not. And the chart follows the naive understanding that “agnostic” and “atheist” are simply things that you are, rather than words you identify with, as demanded by circumstance. In order to determine where you fit in, the question isn’t “Where are you on this chart?” but “What’s your opinion of this chart?”
Let’s find out!
Personally, I live in a post-atheist world. I used to be involved in atheist activism, but I quit as the movement became increasingly toxic. Richard Dawkins, to whom I give credit for doing his part to destigmatize atheism, is now doing his part to reinforce stigma against trans people, and I won’t have it. In the mean time, my personal social sphere is increasingly filled with silently non-religious people (or maybe religious? who knows), and it’s kind of a non-issue. I identify as an atheist, but drilling down into adjacent labels just feels like unearthing old political arguments that most people don’t care about.
That’s just one personal narrative, and I admit, a highly unusual one. What other personal narratives are out there? My suggestion is that instead of using graphs to try to fit everyone into a single framework, we adopt a “let’s find out!” attitude. Let’s find out why people identify the way they do. Why do people identify as atheist and/or agnostic? Where do they see themselves in relation to other identities? What’s up with the “nothing in particular” group that according to surveys, is larger than agnostics and atheists combined? Aren’t you curious?
And if the answer sometimes involves terrible graphs, so be it. I love terrible graphs.