On my Twitter timeline, I saw the following:

“Atheism does not require certainty. But we can be as certain the Christian god does not exist as Christians are that Thor does not exist.”

If I were to reword the tweet, albeit in a way that is too long for twitter, I would have offered something like this:

“Atheism, when defined as the belief that God does not exist, does not require certainty. Rather, like any other belief, people can hold this belief with varyingdegrees of belief. The atheist’s degree of belief that the Christian god does not exist is the same or virtually the same as the atheist’s degree of belief that Thor does not exist.”

Okay, I said it was wordy.

What to make of this proposed rewording? Three thoughts.

1. On the assumption that atheism is true, it doesn’t follow that Thor and the Christian god have the same (im)probability. It’s easier to illustrate this point using metaphysical naturalism, which entails that no supernatural beings exist. Let’s say the naturalist’s degree of belief that naturalism is true is 0.95. It follows, then, that the maximum degree of belief a naturalist could assign to any supernatural being, including Thor and the Christian god, is 0.05. (This is because both Thor and the Christian god entail that supernaturalism is true; thus neither Thor nor Yahweh can be more probable than supernaturalism.) But it doesn’t follow that the naturalist’s degree of belief in the existence of Thor or the Christian god would actually be 0.05; it could be (and probably is) less than that amount. Why? Because Thor and the Christian god make many more specific claims than ‘mere’ supernaturalism.

2. If one adopts Draper’s theory of intrinsic probability–which says that a proposition’s intrinsic probability is determined by modesty, coherence, and nothing else–then the naturalist (who, by definition, is an atheist) would probably say that the intrinsic probability of Thor and Yahweh are roughly equal if not exactly equal. (I have to confess I haven’t studied Norse mythology, so I don’t know if that would be correct or not. It wouldn’t surprise me in the least if it were true, however.)

3. Turning to the final probability of Thor and Yahweh, final probability is determined by intrinsic probability and the likelihood of the evidence. In order to defend the claim that Thor and Yahweh have the same final probabilities, one would need to evaluate the evidence for both. I don’t have much to say about this other than one point: atheists can believe that both Yahweh and Thor have low, even ridiculously low, final probabilities without believing they are equal.

For example, it’s possible that a naturalist who has studied both Norse mythology and Christian apologetics might reach the following conclusions:

(a) Both Thor and Yahweh have very low (but not hopelessly low) intrinsic probabilities.

(b) The likelihood of the evidence relevant to Thor is 0.01.

(c) The likelihood of the evidence relevant to Yahweh is 0.04.

(d) Therefore, while the evidence for Yahweh is four times as strong as the evidence for Thor, the evidence for both is pathetically weak and hence both have extremely low final probabilities.

Given the low intrinsic probabilities of both Thor and Yahweh, however, the naturalist might justifiably believe an investigation of the Thor-specific and Yahweh-specific evidence isn’t worth his or her time. This might especially be the case if the naturalist believes he or she has strong evidence against mere supernaturalism.

This point leads to another rewording of the tweet.

“Atheism does not require certainty. But highly specific supernatural claims, like “Thor exists” and “Yahweh exists,” are so intrinsically improbable that an investigation of the evidence isn’t worth the effort.”