Let’s take a break from my grand account of normativity and just do some old-fashioned conceptual analysis. I am going to try and show that God cannot be both omniscient and morally perfect.
Omniscience does not only require having all propositional knowledge (although indexicals provide some problems there), it is also requires that you have all phenomenological knowledge (again, setting the indexical problem aside). That is, you can’t just the 0-60 time of a Ferrari you must also know what it is like to smell a flower.
I think this presents a problem with God’s moral perfection. To be omniscient, God must know what it is like to murder or rape someone. And the only way that God could know what it was like to do these things is to actually do them. Clearly, a person who rapes and murders is not a morally perfect being.
Now, one might object here that God might simply be able to simulate doing these things in His mind, and thus avoid the heinous acts. I think Christians should be uncomfortable with this solution since they think the intention is as evil as the deed but set that aside. I think the problem is that God is omniscient: She knows that these are simulations. Simulations work for human beings because you can fool our mind into thinking that it is real by simulating neuro-physiological states. But God can’t be fooled; She will always know that She is engaged in a simulation and cannot have genuine phenomenological knowledge of the bad acts.
The other way you might want to go is to argue that omniscience does not require phenomenological knowledge. Alas, I don’t think this works. First, it seems to be a fairly important constituent part of our own knowledge sets. But more importantly, it is a very odd omniscient being that has total knowledge of the universe, but doesn’t have the kind of knowledge that ordinary, fallible human beings have. Very odd.
Or you could argue, similarly, that God can obtain phenomenological knowledge through some other mechanism besides becoming embodied and actually doing certain activities (or taking control of human bodies engaged in those activities). But this is logically incoherent: phenomenological knowledge just is the knowledge that we gain from being embodied and experiencing the world.
I am curious if anyone sees an obvious flaw in my argument.