While I have already written an assessment of Stephen Law’s evil god challenge, after listening to Law engage in an informal debate on the topic with Glenn Peoples on Unbelievable, I have a few more observations to make.
Law seems to take as his starting point the idea that people reject the existence of an evil God based on the empirical evidence: there is simply too much good in the world for an evil god to exist. Then he reasons that if the presence of good in the world makes the existence of an evil God absurd, people should also recognize that the presence of evil in the world makes the existence of a good God equally absurd. The success of his argument depends on three assumptions:
(1) Empirical evidence is valid for determining the moral nature of God
(2) Empirical evidence is valid for determining whether God exists
(3) The empirical evidence is symmetrical with respect to good and evil
Empirical evidence is valid for determining the moral nature of God
With respect to (1), while it may be true as a matter of fact that many people who reject the existence of an evil god do so on the basis of empirical data, and that empirical data would equally rule out the existence of a good God if applied consistently, Law’s symmetrical argument falls apart the moment one rejects empirical evidence as a legitimate or sufficient basis for determining the nature of God. Arguably, basing one’s view of God’s moral nature on the amount of evil or good in the world is simply bad reasoning. Even Law admits this, while still maintaining that the empirical data is informative about what God is not like. Ironically, then, Law’s argument for the absurdity of the existence of a good God based on the absurdity of the existence of an evil god can only be made if one accepts a faulty method for determining the nature of God. At best Law’s symmetry argument will show theists why they should not make judgments about God’s moral nature based on empirical data. It does nothing to demonstrate that the existence of a good God is absurd. His argument is rhetorically effective only for those who falsely think empirical data is a good way to determine God’s moral nature.
Empirical evidence is valid for determining whether God exists
With regard to (2), the only reason to think the existence/amount of evil makes the existence of a good God absurd or the existence/amount of good makes the existence of an evil god absurd is if there is no theodicy that can explain the tension between the moral qualities we find in the world and God’s nature. But Law implicitly admits that there is a theodicy that can account for the tension when he applies the theist’s “morally sufficient reason” (MSR) theodicy to his evil god, arguing that the same theodicy could be used to support the existence of an evil: an evil god could have a morally sufficient reason for permitting all of the good in the world in the same way that a good God could have a morally sufficient reason for permitting all of the evil in the world. While he does not think there is an evil god, he obviously thinks the MSR theodicy successfully shows that such a god could be compatible with a world suffused with moral qualities that are contrary to his nature. The fact that a successful theodicy can be offered to explain why a god of a particular nature would allow a world to exist that contains contrary moral qualities demonstrates that our intuitions about the improbability of the existence of a good or evil god based on empirical evidence alone are misguided.
The empirical evidence is symmetrical with respect to good and evil
As for (3), even if empirical data was legitimate and sufficient for determining the moral nature of God, one could argue that the evidence is not symmetrical because there is more good in the world than evil, and the amount of goodness in the world can only be explained by the existence of a good God.
The fact remains that empirical data is not sufficient to determine the existence and/or nature of God. While we might appeal to all the good in the world in support of the existence of a good God, we have independent reasons to think God exists and that He is good. For example, our conscience reveals both the existence of a moral law giver as well as His nature. Moral laws are unlike other laws in that they come equipped with an internal motivation to keep them. That motivation works in a one-sided direction toward good and away from evil. This aspect of conscience is informative, telling us something about the moral nature of the moral law giver: He is good.
Law claims that empirical evidence cannot be used to establish what kind of God does exist, but it is can be useful for eliminating certain conceptions of god as impossibilities. Just as we intuitively recognize that an evil god cannot exist based on all of the good in the world, we should also recognize that a good God cannot exist based on all of the evil in the world. But to even say that we can use empirical observations to rule out certain conceptions of God requires that one believe empirical observations are informative about the nature of God. If they are informative about the nature of God, then it seems arbitrary to say empirical evidence can inform us about what God is not like, but cannot inform us of what He is like.
Law claims that since theodicies can be used to support an evil and a good God, it shows they are useless. But clearly this is an incorrect assessment. While theodicies are not useful for telling you whether God is good or evil, they are obviously useful for demonstrating that there is no logical incompatibility between a God of a certain moral nature and a world suffused with examples of a contrary moral character.
It should also be pointed out that even if the empirical evidence counts against both the existence of a good God and an evil god, it does rule out the existence of God altogether because God’s nature could be amoral.