This will be the final post in my series on the Andrews-Schieber debate on Christian theism. In this final post, I want to comment on just one statement made by Max Andrews in his closing statement.

Remember that Schieber’s soteriological argument from evil is as follows:

(21′) If the Christian God exists, he is essentially morally perfect, omnipotent, omniscient.

(22′) If the Christian God exists, he chose to create Hell and send the vast majority of people to suffer eternally within it.

(23) There is no moral justification for sending anybody to suffer eternally in Hell.

(24) A being who acts in a way that is morally unjustified cannot be essentially morally perfect.

(25′) The Christian God does not exist.

In his closing statement, Andrews says this.

Secondly, the argument against God due to hell is a demonstrable lack of understanding of divine love and justice. Likewise, this argument severely lacks a biblical and hermeneutical interaction of the biblical text and he completely side steps my philosophical arguments. This is an intra-Christian issue and should it be the case that the interpretation of an eternal hell not be true it does not negate the existence of God. This argument, at best, demonstrates a revision of hermeneutical approaches to the doctrine. This argument simply doesn’t belong in this debate and should be considered off topic.

This response fails because it contradicts the axioms of the probability calculus. To be precise, it violates the theorem of total probability. Here’s why. Let Pr(CT) be the epistemic probability that Christian theism is true. We can partition Christian theism into two versions: those that include an eternal hell (H) and those that do not (~H). (For those familiar with Venn diagrams, draw a circle to represent CT and then divide the circle into two parts, one for H and for ~H. The size of the parts does not matter at this point.)

According to the theorem of total probability, Pr(CT) is equal to an average of the probabilities of CT&H and CT&~H. It’s not a simple straight average, however, because the probability of CT conditional upon H, Pr(CT|H), may not be equal to the probability of CT conditional upon ~H, Pr(CT|~H). Rather, it’s a weighted average which takes Pr(CT|H) and Pr(CT|~H) into account. Applied to our discussion, the formula looks like this:

Pr(CT) = Pr(CT|H) x Pr(H) + Pr(CT|~H) x Pr(~H)

If Pr(CT|H) > Pr(CT|~H), then Pr(CT) will be closer to Pr(H). If, on the other hand, Pr(CT|~H) > Pr(CT|H), then Pr(CT) will be closer to Pr(H). So if we believe that Christian theism is committed to some sort of doctrine of Hell, then the theorem of total probability tells us that the Pr(H)  is relevant to Pr(CT).

Suppose you agree with me so far. You may be asking, so what? Well, suppose the doctrine of Hell is internally inconsistent. Since anything that is internally inconsistent has a logical probability of zero, we should set our corresponding epistemic probability to zero as well. So if the doctrine of Hell is internally inconsistent, then Pr(H) = 0. And the greater the value of Pr(CT|H), the closer Pr(CT) will also be to 0.

So, contrary to what Andrews claims, the argument against God due to Hell is at least relevant to a debate on God’s existence.