Justified True Belief

(P1) For the logical version of the problem of evil to succeed, the atheist must demonstrate that the propositions "an omnipotent, omniscient, omnibenevolent God exists" and "evil exists" are strictly logically incompatible. + The intellectual problem of evil comes in several varieties, and the logical version is only one of them. Other forms (probabilistic, external, emotional) raise different challenges and require their own treatments. This argument addresses only the strongest possible form of the atheistic case, the claim that theism is internally self-contradictory, like the claim that something is both a triangle and a square. (1) The logical version makes a contradiction claim, not a probability claim. The logical version of the problem of evil is not the modest claim that evil makes God's existence seem unlikely, or hard to accept, or emotionally jarring. It is the much stronger claim that the existence of God and the existence of evil are flatly contradictory. If true, this would mean the Christian holds two beliefs that cannot both be true at once, just as no one can consistently believe that some shape is both a triangle and a square. Since evil obviously exists, God would obviously not exist, and theism would be irrational. (2) Strict logical incompatibility requires that no possible world contain both. Philosophers call two claims strictly logically incompatible when there is no possible world in which both could be true, that is, when their conjunction entails a contradiction. Think of it this way: there is no possible world, no matter how strange, in which a single object is both a triangle (three-sided) and a square (four-sided). The concepts rule each other out by definition. The atheist pressing the logical version must show that "God exists" and "evil exists" rule each other out in exactly this strong way. Anything weaker, such as "God's existence is improbable given the evil we observe," is a probabilistic version of the problem and falls outside the scope of this argument. (3) The classical formulation runs from Epicurus to Mackie. The argument has been pressed since antiquity. The Greek philosopher Epicurus first posed the dilemma: if God is willing to prevent evil but unable, he is not omnipotent; if able but unwilling, he is not good. Scottish philosopher David Hume gave the dilemma its modern fame in his Dialogues Concerning Natural Religion. In the twentieth century, Australian atheist philosopher J. L. Mackie pressed the strongest analytic version in his 1955 article "Evil and Omnipotence," arguing that theism is positively irrational because its core commitments contradict each other. (4) The burden of proof rests on the atheist. The party claiming a contradiction is the party who must demonstrate it. Christian philosopher Alvin Plantinga observes that, as opposed to twenty or twenty-five years ago, most atheologians have conceded that there is in fact no inconsistency between the existence of an omnipotent, omniscient, wholly good God and the existence of the evil the world contains. Mackie himself eventually softened his position. To revive the logical version, an atheist would have to produce a successful demonstration of contradiction, something the philosophical community has overwhelmingly judged has not been done. Therefore, the logical version stands or falls on whether the atheist can establish a strict logical incompatibility between theism and the existence of evil.

(P2) To demonstrate such an incompatibility, the atheist must establish at least one necessarily true additional premise, standardly either that an omnipotent God can bring about any world he desires, or that an omnibenevolent God prefers any world without evil to any world containing evil. + (1) The bare propositions do not generate a contradiction by themselves. "An omnipotent, omniscient, omnibenevolent God exists" and "evil exists" are not explicitly contradictory in the way that "Bob is a bachelor" and "Bob is married" are explicitly contradictory. The words "bachelor" and "married" rule each other out by definition; the words "God" and "evil" do not. Consider a parallel: "Bob is wealthy" and "Bob is unhappy" are not contradictory by themselves, no matter how unusual the combination. To force a contradiction in such a case, an objector would need a hidden universal premise (such as "all wealthy people are happy") that holds in every possible world. The atheist's argument from evil is structurally like the wealthy-and-unhappy case, not the bachelor-and-married case, which is why hidden premises are needed to do the logical work. (2) The two standard hidden premises concern omnipotence and omnibenevolence. The atheist's case requires something like the following two claims. First, that necessarily, if God is omnipotent, then he can bring about any logically possible world he desires. Second, that necessarily, if God is omnibenevolent, then he prefers any world without evil to any world containing evil. These two premises, together with the existence of evil, would generate the contradiction the atheist needs. If God could bring about any world he wanted, and would want a world without evil, then a world without evil would obtain, and yet it plainly does not. (3) The atheist must show these premises are necessarily true, not merely plausible. A claim is necessarily true (a phrase philosophers use for what holds in every possible world) only if there is no possible world, however unusual, in which it fails. This is a far higher standard than being merely plausible or generally accepted. Think of the difference between "all the swans I have seen are white" and "no swan could possibly be black." The first is a generalization from experience; the second is a claim about every possible swan in every possible world. The atheist's hidden premises about omnipotence and omnibenevolence must meet the second, much stronger standard, because anything weaker yields only a probabilistic argument, not a logical one. (4) The atheist's full burden is therefore quite heavy. The atheist must produce a premise that (a) is necessarily true, (b) is universally acceptable as such, and (c) generates a strict contradiction when conjoined with theism and the existence of evil. Anything less leaves the theist's position internally consistent. As philosopher Nelson Pike noted in his article "Hume on Evil," and as Plantinga developed at length in The Nature of Necessity, the failure to meet this burden has been the recurring weakness of every logical version of the argument since the 1950s. Therefore, the logical version succeeds only if at least one of the standard hidden premises (or some equivalent) can be shown to be necessarily true.

(P3) Neither hidden premise is necessarily true: omnipotence does not include the power to do logical impossibilities, and omnibenevolence is compatible with permitting evil for morally sufficient reasons. + (1) Omnipotence does not extend to logical impossibilities. Even an omnipotent being cannot do what is logically impossible, such as make a square circle or cause someone to freely choose a particular action. The very phrase "freely choose" rules out being caused to choose, so making someone freely choose anything is a contradiction in terms, not a real task that an omnipotent being could in principle accomplish. Think of a master craftsman: he can do anything within his craft, but he cannot make a sound that is silent or a triangle with four sides, not because his power is limited but because those phrases describe nothing at all. In the same way, if God grants creatures genuine libertarian freedom (the kind of freedom in which the choice is not determined by prior causes), then he cannot guarantee how they will use it, since guaranteeing the outcome would void the freedom. (2) Counterfactuals of creaturely freedom may render sinless worlds infeasible. Suppose libertarian free will exists. Then for each free creature, in each situation, there is a truth about how that creature would freely choose. Philosophers call these truths "counterfactuals of creaturely freedom." Because they describe the free choices of creatures, God does not get to decide their truth values, only the creatures themselves do. As a consequence, there may be possible worlds, including sinless worlds, that God cannot actually bring about (which philosophers call "infeasible for God"), because the relevant counterfactuals turn out the wrong way. Think of a chess teacher who wants every student to win every game: he can do anything a teacher can do, but he cannot force a free choice and still call it free. Some outcomes are simply not in his power to produce, even though they are possible in themselves. So the assumption that an omnipotent God can bring about any world he desires fails. (3) An omnibenevolent God can have morally sufficient reasons for permitting evil. A perfectly good agent is not required to prevent every evil. A perfectly good agent is required not to permit evil without sufficient justification. The two are different. Consider a loving parent who allows a child to experience the pain of a vaccination needle. The parent has the power to prevent the pain, and prefers the child not feel pain, yet permits the pain anyway, because the greater good of immunity outweighs the lesser evil of the brief discomfort. In the same way, an omnibenevolent God can permit evils in our world if doing so secures greater goods that could not otherwise obtain, or prevents greater evils that would otherwise arise. The atheist's assumption that omnibenevolence forces the choice of an evil-free world simply ignores this possibility. (4) The atoning death of Christ is a candidate overriding good of incommensurable value. The supreme good of Christ's incarnation and atoning death may be a good of such overwhelming magnitude that a world containing it, even with the evil that makes it possible, is a far better world than any world without it. Eternal salvation, communion with God, and the redemption of suffering are goods of incommensurable value (meaning they cannot be measured on the same scale as finite earthly pain). Philosopher Alvin Plantinga develops this proposal in his essay "Supralapsarianism, or 'O Felix Culpa,'" arguing that the Christian doctrine of the fall and redemption secures a great good that could not otherwise exist. If even one such overriding good is logically possible, then the atheist's assumption that an omnibenevolent God would prefer any world without evil collapses. (5) Natural evil also has logically possible explanations. The free will defense is sometimes thought to falter on natural evils (earthquakes, disease, animal suffering) since these are not the result of human moral choices. But the logical version requires only logical possibility, not plausibility. Plantinga points out that natural evils could possibly result from the free choices of non-human moral agents, such as fallen angels. Other possibilities (soul-making, the structure of natural law required for stable free choice, the cosmic consequences of the fall) are also logically possible. For the logical problem, the atheist needs to rule out every possible explanation, an impossible burden. Therefore, neither candidate hidden premise required by the atheist is necessarily true, and the logical incompatibility cannot be established.

(C) Therefore, the logical version of the problem of evil fails to demonstrate that God and evil are logically incompatible.