The article on theistic modal realism is ingenious. (One-sentence summary: God’s options when creating should be thought of as ensembles of worlds, and most likely he’d create every world that’s worth creating, so the mere fact that ours is far from optimal isn’t strong evidence that it didn’t arise by divine creation.)
I don’t find the TMR hypothesis terribly plausible in itself—my own intuitions about what a supremely good and powerful being would do don’t match Kraay’s—but of course a proponent of TMR could always just reject my intuitions as I’d reject theirs.
However, I think the TMR hypothesis should be strongly rejected on empirical grounds.
It is notable—and this is one element of a typical instance of the Argument From Evil—that our world appears to be governed by a bunch of very strict laws, which it obeys with great precision in ways that make substantial divine intervention almost impossible. It seems that there are many many many more possible worlds in which this property fails than in which it holds, simply because the more scope there is for intervention the more ways there are for things to happen. Therefore, unless the sort of lawlikeness we observe is so extraordinarily valuable that tiny changes in it make a world far less likely to be worth creating, we should expect that “most” worlds in the TMR ensemble would be much less lawlike than ours: e.g., we might expect prayers to be commonly answered in clearly detectable ways. So how come we’re in such an atypical world?
Generalizing: I think we should expect that for most measures of goodness X, worlds with higher values of X should be dramatically more numerous in the TMR ensemble unless increasing X reduces the number/measure of possible worlds much more drastically than for most other choices of X. (Because when you increase X, you get the chance to reduce Y or Z or … a bit. More choices.) Therefore, we should expect that for measures of goodness X where “better” doesn’t imply “much more constrained” most worlds (hence, in particular, ours, with high probability) should have values of X that are close to optimal, or at least far from marginally acceptable. This doesn’t seem to be true.
It seems to me that counter-arguments to these are likely to be basically the same as counter-arguments to the original argument from evil.
The other thing about TMR is that it undermines any version of theism that expects God to behave as if he cares about us. If TMR is right then, any time God has the option of doing something to make your life better, then he forks the universe vastly many ways and tries out every possible option (including lots of ways of doing nothing, and even ways of deliberately making things worse for you) apart from ones that make the whole universe not worth while. As mentioned above, it seems to me that this should make us expect that visible divine intervention should be pretty common, but in any case it’s not terribly inspiring. A bit like having a “friend” who, any time she interacts with you, rolls dice and chooses a random way of behaving subject only to the constraint that it doesn’t cause the extinction of all human life. Similarly, you’ve got no reason to trust any alleged divine revelation unless its wrongness would be so awful as to make the world not worth creating. (These arguments are again closely parallel to ones that come up with the ordinary argument from evil, in response to responses that basically take the form of radical skepticism.)
Granted that god exists and cares about us and he can change the world, even in tiny aspects, it’s very likely god will use those small aspects as a base to create the perfect world (kind of like AI FOOM). It follows that any world where god has some kind of minimum control will converge to the perfect world. Given that we are not in the perfect world, we can assume god does not have the minimum level of control.
The article on theistic modal realism is ingenious. (One-sentence summary: God’s options when creating should be thought of as ensembles of worlds, and most likely he’d create every world that’s worth creating, so the mere fact that ours is far from optimal isn’t strong evidence that it didn’t arise by divine creation.)
I don’t find the TMR hypothesis terribly plausible in itself—my own intuitions about what a supremely good and powerful being would do don’t match Kraay’s—but of course a proponent of TMR could always just reject my intuitions as I’d reject theirs.
However, I think the TMR hypothesis should be strongly rejected on empirical grounds.
It is notable—and this is one element of a typical instance of the Argument From Evil—that our world appears to be governed by a bunch of very strict laws, which it obeys with great precision in ways that make substantial divine intervention almost impossible. It seems that there are many many many more possible worlds in which this property fails than in which it holds, simply because the more scope there is for intervention the more ways there are for things to happen. Therefore, unless the sort of lawlikeness we observe is so extraordinarily valuable that tiny changes in it make a world far less likely to be worth creating, we should expect that “most” worlds in the TMR ensemble would be much less lawlike than ours: e.g., we might expect prayers to be commonly answered in clearly detectable ways. So how come we’re in such an atypical world?
Generalizing: I think we should expect that for most measures of goodness X, worlds with higher values of X should be dramatically more numerous in the TMR ensemble unless increasing X reduces the number/measure of possible worlds much more drastically than for most other choices of X. (Because when you increase X, you get the chance to reduce Y or Z or … a bit. More choices.) Therefore, we should expect that for measures of goodness X where “better” doesn’t imply “much more constrained” most worlds (hence, in particular, ours, with high probability) should have values of X that are close to optimal, or at least far from marginally acceptable. This doesn’t seem to be true.
It seems to me that counter-arguments to these are likely to be basically the same as counter-arguments to the original argument from evil.
The other thing about TMR is that it undermines any version of theism that expects God to behave as if he cares about us. If TMR is right then, any time God has the option of doing something to make your life better, then he forks the universe vastly many ways and tries out every possible option (including lots of ways of doing nothing, and even ways of deliberately making things worse for you) apart from ones that make the whole universe not worth while. As mentioned above, it seems to me that this should make us expect that visible divine intervention should be pretty common, but in any case it’s not terribly inspiring. A bit like having a “friend” who, any time she interacts with you, rolls dice and chooses a random way of behaving subject only to the constraint that it doesn’t cause the extinction of all human life. Similarly, you’ve got no reason to trust any alleged divine revelation unless its wrongness would be so awful as to make the world not worth creating. (These arguments are again closely parallel to ones that come up with the ordinary argument from evil, in response to responses that basically take the form of radical skepticism.)
Granted that god exists and cares about us and he can change the world, even in tiny aspects, it’s very likely god will use those small aspects as a base to create the perfect world (kind of like AI FOOM). It follows that any world where god has some kind of minimum control will converge to the perfect world. Given that we are not in the perfect world, we can assume god does not have the minimum level of control.