Here are three thoughts which occurred to me recently as I was rethinking my arguments in The Paradox of Hell and Justice. The first is that, contra the Molinist apologetic offered by William Lane Craig, the atheist (or other critic of Christianity) could argue that since the pool of logically possible worlds includes many more worlds where not every single counterfactual conditional about the libertarian free choices of those in hell is such that every person in hell incurs, by their free uncoerced will, indefinitely more damnation upon themselves, than it includes worlds which do, the Molinist account is unlikely to succeed. In other words, worlds where the counterfactuals allow for at least one person in hell to reach the end of their deserved punishment are more plenteous among the set of logically possible worlds than worlds in which the set of infinitely many subjunctive counterfactuals do not allow even a single person in hell to reach the end of their justly deserved punishment. This is an argument from the incredible improbability of the Molinist account – what the Molinist proposes is so unlikely to be true that if it were the only way in which hell could possibly be justified then the likelihood that hell is not justified would be so incredibly great it couldn’t be reasonably denied. This would not be a case for the logical incompatibility of God’s being just and the existence of hell, but rather an evidential or probabilistic argument against their being coincident.
The second thought is this. First, I had thought that the apologetic Pruss proposed could be made to work on either theory of time (the A-theory or the B-theory), but I am now thinking that there may be a subtle problem on the A-theory. Clearly, on either theory, there is no point in time at which the subjective suffering of the damned is ‘complete.’ The trouble, though, is that on the A-theory the amount of suffering demanded by God’s justice is not and will never be satisfied. Although the point at which such suffering would be ‘complete’ is removed an infinite distance away from any point in time (on either theory of time), the difference is that on the A-theory that point acts as the limit towards which something endlessly approaches and never actualizes, whereas on the B-theory the whole infinite set of times is already actual. So, if some measure of suffering is both justified and is demanded by justice itself, then this second apologetic account of hell, at least on the A-theory, does not allow for justice to ever be satisfied in actuality, or, to be more precise, it doesn’t allow for justice to be satisfied in actuality (we remove tensed indexicals). Therefore, if this second apologetic account of hell were correct, it would provide us with good reason to adopt a B-theory and reject the A-theory. On the other hand, one could always resort to arguing in the reverse direction, suggesting that since the A-theory is true this apologetic model of hell must be wrong.
This leads me to my final thought for the day; namely that the problem here may be stickier than the A-theorist imagines. It’s not just that this particular model of hell proposed by Pruss poses a problem for the A-theorist. Now that I think of it, it seems as though this would provide a serious problem for any apologetic account of hell on the A-theory, for it is strictly not logically possible that i) the A-theory be true, that ii) the classical doctrine of hell be true, and that iii) God’s justice is (possibly) satisfied in actuality.