A contradictory answer for the problem of evil.

Since at least Ancient Greece, the problem of evil has been regarded as one of—if not the—strongest arguments against theism. Discussion of this problem, from Leibniz to Plantinga and Mackie, has occupied a central place in the philosophy of religion. Despite deep disagreements, some assumptions have remained relatively stable throughout this debate. One of them is that reality is consistent.

In this post, I explore what happens when we abandon that assumption. More precisely, I investigate how the problem of evil changes once we stop requiring consistency. I argue that, given some plausible assumptions about divine omnipotence, it is possible to formulate a response to the problem of evil that has been little explored so far—and which appears promising.¹

Without further ado.

Contradictions in God

It is not immediately clear how to begin an inquiry into the impact of contradictions on the problem of evil. Any proposal in this direction must at least be compatible with the reality we experience, and must also respect certain minimal criteria we wish to preserve. A trivial move would be to claim that God exists and does not exist. It is hard, however, to imagine any theist adopting this position, at least as a final strategy.

Fortunately, there are more interesting paths. Divine omnipotence has always been one of the most difficult attributes to define—and it is not hard to see why. Naively, omnipotence is said to mean the power to do everything.

But “everything” is a lot. Can God do what is logically impossible? What is metaphysically impossible? Many are tempted to answer negatively, holding that divine power is somehow limited by necessary truths.

I want to suggest that this is not the best route. Instead, we should take the naïve conception of omnipotence seriously. Since this is not the main focus of the post, I will only briefly present four reasons in favor of this view.

First, it seems plausible to treat absolute sovereignty as a perfection. An agent has absolute sovereignty over something if and only if that thing exists at a time t solely in virtue of the agent’s will at t: its continued existence depends entirely on the agent’s willing it to exist, and it would cease to exist were the agent to will otherwise. Note that it is plausible to say that the reason why a set to which I belong exists is that I exist; yet I cannot decide whether that set continues to exist. Thus, I do not have absolute sovereignty. If God has absolute sovereignty, it follows that nothing—not even necessary truths—lies beyond His power.

Second, ceteris paribus, it seems we should attribute to God the maximal degree of power. If it is coherent to grant God total control over everything, that would be the best theoretical outcome. There is therefore a prima facie reason to adopt a maximalist conception of omnipotence.

Third, on certain analyses of conceivability or epistemic possibility—such as those developed by Priest—any proposition is conceivable. Thus, if we claim that there is something God cannot do, it will seem conceivable that there exists a being more powerful than God, which conflicts with the very idea of divinity.²

Finally, suppose we say that God cannot do x and offer an explanation for this limitation. It always seems possible to conceive of a being very similar to God who can do everything except some trivial thing, together with an analogous explanation. In that case, however, it is clear that such a being would not be omnipotent.

These arguments are quick, but quite strong. For that reason, I conclude that, at least for now, we are justified in accepting a naïve conception of omnipotence. This, of course, generates contradictions. To see this, consider the famous paradox of the stone.

Given the naïve conception of omnipotence, God can create a stone so heavy that He cannot lift it. At the same time, given that same conception, God can lift any stone. It follows, therefore, that God can and cannot lift that stone—a contradiction.

Contradictions and inferences

Why does this matter? Because contradictions act as filters on inferences that, in classical contexts, would seem obvious. The simplest example is the well-known disjunctive syllogism:

1. (A V B)

2. ¬(A)

3. Therefore, (B)

In a consistent setting, the inference is valid. If contradictions are admitted, however, (A) may be true and false while (B) is just false, implying that (1) and (2) are true and (3) is false. The contradiction blocks the inference.

Something analogous happens in the case of omnipotence.

Definition (naïve):
(x) is omnipotent if, for any action (a), (x) can perform (a).

From this, it seems immediate to conclude that, for any particular action (z), (x) can perform (z). Once contradictions are admitted—as in the paradox of the stone—this inference is no longer guaranteed. There are models in which a being satisfies the definition of omnipotence and yet it is false that it can perform some particular action, precisely because of the contradiction involved.

I assume the following notion of validity:

(X) validates (Y) if and only if there is no model in which (X) is true and (Y) is false.

Given this definition, and admitting contradictions, it follows that being omnipotent is compatible with not being able to perform some action. The contradiction blocks the usual inferential move from the definition to the maximalist conclusion.

Contradiction and evil

Here is the central proposal. Given the contradiction involved in omnipotence, we cannot infer

for any action, God can perform it

to

God can prevent some particular evil.

The contradiction blocks this implication. Thus, any argument that moves from divine omnipotence to the conclusion that God can perform some specific action relies on an invalid inference. Such arguments therefore fail.

Some objections

One might object that, even if we cannot infer from omnipotence that God can prevent evil, all that the argument from evil requires is that it be true that He can prevent some particular evil. Thus, it is true that

(1) God can prevent m,

where m is some actual evil.

This is correct. However, given the characterization of omnipotence above, there will be models in which it is both true and false that God can perform the action of preventing m. One such model is one in which God uses His power to cease being omnipotent.

Note that if we are interested only in a defense—in Plantinga’s sense—this is already sufficient, as far as I can see, to undermine virtually all logical arguments from evil.

Someone might insist that although such a scenario is possible, the contradiction involved is so remote that it is unlikely to be true of the actual world. But in some paraconsistent logics, the possibility of a contradiction implies its actuality.³ Thus, it follows that it is actually both true and false that God can prevent that evil.

Another objection is that even granting all this, theism would still be committed to something deeply absurd. After all, it would still be true—though also false—that for any evil, God can prevent it. Thus, the following claim would seem to be true:

(*) God could have saved our friend, who was dying; He loved him, had the power, and yet did not do so.

This sounds absurd. The best response is to accept the claim as true while remembering that it is also false. For it is equally true that God could not have saved your friend. An example may help validate the intuition.

Suppose Sally is an extremely devoted mother who loves her children above all else. Due to a strange encounter with a mysterious priest, she acquires the following property: for any action Sally performs, she both performs and does not perform that action.

One day, a robber breaks into Sally’s house and threatens to kill her children if she does not reveal where the money is hidden. Sally tells him—and does not tell him. The robber shoots.

We then have:

(**) Sally loved her child, could have prevented what happened, and yet allowed him to be shot.

This seems absurd. And it would be, if all we could say were that Sally could have prevented it. But in this case, she both could and could not have prevented it. Given these circumstances, (**) expresses only a trivial implication of a deeply contradictory situation, not a genuine moral failure on Sally’s part.

Conclusion

I hope you enjoyed this post. This is my first one, so I am somewhat apprehensive about the result. God bless you all.

Citations

Martin, B. (2015). Dialetheism and the Impossibility of the World. Australasian Journal of Philosophy, 93(1), 61–75. https://doi.org/10.1080/00048402.2014.956768

Priest, Graham (2016). Thinking the impossible. Philosophical Studies 173 (10):2649-2662.

Beall, Jc (2023). God, gluts and evil. Analysis 83 (4):643-652.

Weber, Zach (2019). Atheism and Dialetheism; or, ‘Why I Am Not a (Paraconsistent) Christian’. Australasian Journal of Philosophy 97 (2):401-407.

Gordon, Noah (2024). God, Über-God, and Unter-God. Religious Studies 60 (4):564 - 579.

1. The discussion is strongly influenced by Beall, and the core of the solution is mentioned by Jeff Speaks; however, I add some details of my own. Therefore, there is no claim that this is his view, since it includes modifications.

2. This and the next argument are known in the literature as the Über and Unter God argument. For a more sophisticated version, see Gordon.

3. Indeed, this is the situation in what is probably the most well-known paraconsistent logic, LP, as Martin (2015) show.