Argument Against God From Evil
One thing to remember when discussing the existence or non-existence of God is that we are limited by our language. The paradox is that by even suggesting the variable "God" we are forced to assume its existence for the sake of the discussion. This paradox was first recognized by the presocratic philosopher Parmenides and partially refuted by Plato in his dialogue, the Sophist. Medieval philosophical thinking best represented by Anselm's ontological argument for the existence of God, attempted to establish God at the top of an Aristotelian "Chain of Being." Simply put, Anselm's argument was that if we can conceive of an all-powerful being called "God" that nothing can be more powerful than, then God exists. Descartes advances this argument in his fifth Meditation as well. However, Kant argued effectively in his Critique of Pure Reason that existence cannot be considered a predicate because no matter how many characteristics we attribute to a being we do not add anything to that being by predicating existence of it. Descartes's (and Anselm's) suggestion that "God is x" confuses the grammatical construction of the copula is while remaining unclear about the is of existence. For example "the ball is red" and the "ball is" make use of two entirely different senses of the connecting copula "is." In the first case we are predicating a characteristic of the ball, while in the second case we are predicating existence of the ball. Thus, Kant accuses Descartes and Anselm of wrongly conflating the is of existence with the is of predication.
In order to clear up this grammatically confusing mess, we say today that "God exists" means "for all x, if x is a God, then x exists." Therefore, when Descartes predicates existence of God he is really saying "if God exists then God exists" which, rather than predicate actual existence to God, merely restates the problem. So this is why very few philosophers today accept the a priori ontological argument whereby existence is a perfection and God's existence can be understood just by reflecting on the divine attributes of God. It seems that we humans are a picky lot, demanding at least some empirical evidence to be presented to us before we go around deciding what exists and what does not exist.
Obviously, my argument is informed by Mackie's famous paper published in Mind (Vol. 64, No. 254: 1955) whereby theism is said to be logically (or internally) inconsistent. Am I arguing, as Mackie does, that God and the existence of evil are logically inconsistent? Before I answer that, let me review Mackie's argument and the famous Free Will Defense that Plantinga asserts in his reply to Mackie. Mackie presents these two premises:
(1) God is omnipotent, omniscient, and wholly good.
(2) Evil exists.
Mackie argues that these two premises are logically inconsistent. Now, as Rowe points out in his Philosophy of Religion (Wadsworth, 1993), the logical inconsistency in these two premises are not explicit in the same way as "all cats are grey" and "at least one cat is not grey" are inconsistent. (This is directly inconsistent because it asserts x and not-x.) Rather, Mackie's argument is like saying "all cats are grey" and "at least one cat is orange." These two premises do not explicitly contradict each other, but if we derive a third premise "at least one cat is not grey" from them, we implicitly reach a logically inconsistent set of premises. There's a catch to this. (There's always a catch eh?) The third premise must be necessarily true in order for the entire set of premises to be considered logically inconsistent. In this case, we can see that given the proposition that all cats are grey and at least one is orange, then it is necessarily true that not all cats are grey. So, while this set of premises is not explicitly contradictory, it is found to be implicitly contradictory. This is Mackie's strategy in his argument. Obviously, (1) and (2) are not explicitly contradictory since they do not assert x and not-x. However, Mackie feels that there are two additional premises that can be derived from (1) and (2)--premises that implicitly show that theism is logically inconsistent. These two additional premises are:
(3) Good is opposed to evil in such a way that a good being always eliminates evil as far as it can.
(4) There are no limits to what an omnipotent being can do.
Let's get one thing straight. We're dealing with logical limits here. No one wants to become embroiled in the odd sophistry of whether or not God can create round squares, rocks too heavy to lift, or married bachelors. For the sake of this discussion, there are no logically possible limits to God's omnipotence. God can do any logically possible thing.
If Mackie is correct, the set of premises (1) through (4) is logically inconsistent and thus, theism is internally logically inconsistent. However, let's not begin writing God's eulogy just yet! Remember that (3) and (4) must be necessarily true before we can say that theism is logically inconsistent. Plantinga rejects Mackie's argument in a series of famous papers collected in his The Nature of Necessity (Clarendon Press, 1974). In one paper, reprinted in The Problem of Evil (Oxford University Press, 1990) entitled "God Evil, and the Metaphysics of Freedom" Plantinga argues convincingly that it is possible that God cannot create every logically-possible world, thus, contra (3) God cannot always eliminate evil.
Plantinga's full argument is very technical and involves those tricky beasts called counterfactuals. If your idea of a good time is cuddling up next to an old book of logic (perhaps Cohen and Nagel's ancient tome?) then my summary may not satisfy your appetite. However, for the rest of us mere mortals I'll summarize Plantinga's argument. Plantinga argues that God's omniscience in (3) is not contradictory with the inability to actualize just any logically-possible world. This is because humans possess freedom of the will and a human being must be free to choose morally significant actions. Some of us spoiled the party by freely choosing the evil rather than the good and these choices are the source of moral evil. Mackie seems to argue that God could have just chosen to create that one logically-possible world in which everyone who is created chooses only the good. If this is so, why didn't a wholly good God create that world? It seems then that if we can make free choices, we could have done something differently in an alternate world than what we did in the actual world. I could have turned left to go into the student union, rather than right to enter a coffeeshop. These "could have beens" are called counterfactuals because they deal with possible situations contrary to actual fact. For example, if in the actual world Napoleon chose to command 20,000 troops in his victory over the Austrians, there are two logically possible worlds W where:
W1 = if Napoleon chose to command only 10,000 troops, he would still have won.
W2 = if Napoleon chose to command only 10,000 troops, he would have lost.
In both W1 and W2 the antecedent is equally true so it cannot determine the consequent in either of these worlds. (Napoleon commands 10,000 troops in both logically-possible worlds.) Yet, in the actual world both W1 and W2 are logically-inconsistent, i.e., both cannot be true. Therefore, one of them is a logically-possible world that God cannot actualize. This is not a limit on God's omnipotence, rather, a logically-necessary accompaniment of "world-making." Plantinga further argues that it is possible that all beings suffer from something called "transworld depravity," the condition whereby in all possible worlds that God creates, a being will never avoid at least some evil choices.
If Plantinga is correct then is it logically possible that God cannot prevent the occurrence of evil in the actual world. Interestingly, Robert Adams in "Middle Knowledge and the Problem of Evil," American Philosophical Quarterly 14: 1977, argues that both Mackie and Plantinga are wrong about God's omniscience. Adams feels that Mackie and Plantinga assume incorrectly that God possesses "middle knowledge"--the knowledge of all "future" (to us) contingencies in every detail. Friends who believe that God possesses middle knowledge can be quite annoying. For example, when you turn left to enter the student union, your friend will say "God knew you were going to turn left before you actually turned left." Suppose you begin walking backwards briefly in order to both annoy your friend and demonstrate that you are acting against God's determinism. Your friend will merely say that God knew that you would do that as well. Of course, this can get quite ridiculous as you begin flapping your arms and walking backwards in open defiance of God's penetrating middle knowledge. (Don't try this at home but go to the student union where no one will think it is out of the ordinary.)
Adams argues that given a counterfactual like the one about Napoleon above: "if x was the case, then y would result," the truth of y cannot be known until x is actualized and the antecedent x depends upon the consequent y for its truth value. The only way to break out of this viscous circle is to actualize the event which is to say that God must guess as to the outcome of a counterfactual. What this implies is that if history turn out swell then this can be attributed more to God's "good luck" rather than God's wisdom! In other words, you can tell your annoying friend (and anyone else at the student union) that God really didn't know you were going to flap your arms wildly and walk backward. Adams's suggestion is a serious blow to Mackie's argument for, if it is true that it is logically impossible for God to possess middle knowledge, then God cannot be held responsible for the occurrence of evil in the world. This is to say that God is along for the ride just like the rest of us.
So what does all this mean? It would seem that, barring any new necessarily true premises from God's omnipotence, omniscience, and goodness, Mackie's conclusion that theism is logically inconsistent is false. However, this does not preclude the possibility that the conclusion to the nonexistence of God is rationally justifiable. This is where my reductio argument comes in. After all, it is difficult to reconcile an omnipotent, wholly good being with the existence of an Adolf Hitler. Why didn't God quietly "call home" the infant Adolf in the same manner that God was said to call home the children at the Oklahoma City bombing's day care facility? My argument does not attempt to prove that theism is internally logically inconsistent. Based on Plantinga's and Adam's refutation above I am satisfied that the existence of God and evil are not implicitly contradictory. After all, God may be unwilling to prevent evil. However, I am arguing that, given the divine attributes of goodness and omnipotence along with the existence of evil, it is rational to hold an atheistic position with respect to such a being as God.
Premises to the Argument
(1) God exists.
(2) God is omnipotent and possesses the ability to do any logically consistent thing.
(3) God is omniscient and possesses knowledge of the past, present, and future in every detail.
(4) God is the paradigm of moral behavior and perfection; right and wrong, good and evil are decided based upon the standard of behavior that God possesses and which He fulfills perfectly.
(5) God created the universe and humans.
(6) Humans do not satisfy God and evil exists in the universe; the universe is imperfect.
These are six standard tenets of theism. Now we may begin to draw some conclusions from these premises:
(7) If God is omnipotent, omniscient, and a moral paradigm then he can only create perfection; from (2), (3), and (4).
(8) God only creates perfection; from (7).
(9) Therefore, the universe is perfect; from (5) and (8).
There seems to be an error in our conclusion since (9) directly contradicts premise (6). We had better take a look at the other statements to see what could be the problem. We could deny premise (2) and suggest that God was aware of evil but was powerless to prevent it. This is probably an unsatisfactory thing for most believers to accept however. Perhaps we could instead deny premise (3) and agree with Adams that while God was all-powerful he couldn't see what may lie ahead after he created the universe. If we deny that God is the paradigm of perfection in (4), then we would be contradicting another divine attribute. Similarly, no theist would deny the truth of (5) and dismissing (6) implies that theistic notions such as "forgiveness" and "redemption" are incoherent. We had better look at our conclusions to see where the problem lies.
Perhaps the best approach is to deny (7) and suggest as Plantinga does that God cannot avoid creating evil. But if this is so, why create anything at all? If we deny (8) and suggest that God does not create perfection then premise (4) would be untrue. (Interestingly, this is exactly what the early Gnostics did to reconcile imperfection in nature and the problem of evil coexisting with a perfect God.) Suggesting that God is not perfect will solve our dilemma but may be distasteful to some believers who have grown accustomed to their gods being paradigms of perfection. Along the same lines, denying (9) is problematic since no theist who also believes that Christ's death was salvific would readily assert the universe's perfection.
Unless we are willing to accept a less than perfect Creator this argument is riddled with problems. Given these difficulties, it is rationally justifiable to hold an atheistic position with respect to such a being. I conclude with the denial of the one premise that we have yet to discuss but seems to be giving us all of the trouble: contra (1) it is not the case that God exists.