Reply to the Basingers on Divine Omnipotence

by Alvin Plantinga

Alvin Plantinga is Professor of Philosophy at Calvin College, Grand Rapids, Michigan.

The following article appeared in Process Studies, pp. 25-29, Vol. 11, Number 1, Spring, 1981. Process Studies is published quarterly by the Center for Process Studies, 1325 N. College Ave., Claremont, CA 91711. Used by permission. This material was prepared for Religion Online by Ted and Winnie Brock.


The author discusses some of the criticism put forth by David & Randall Basinger. He discusses I-omnipotence, free will and evil.

Most of what the Basinger brothers say seems to me both sound and sensible. They are quite right, for example, in pointing out that I enthusiastically reject what David Griffin calls "I-omnipotence." According to Griffin, if God has I-omnipotence, then he "can unilaterally effect any state of affairs, if that state of affairs is intrinsically possible." Perhaps another way to put this is to say that God has I-omnipotence if and only if every state of affairs possible in the broadly logical sense is such that it is within God’s power to cause it to be actual. But then, surely, God does not have I-omnipotence. For while the state of affairs consisting in Eve’s freely taking the apple is possible, it is not within God’s power to cause it to obtain; if he causes it to obtain, then he causes Eve to take the apple, in which case she does not take it freely. Eve’s freely taking the apple is a possible state of affairs, but it is not possible that God cause it to be actual. Hence it is not within God’s power to cause it to be actual, and God is therefore not I-omnipotent. So if, as the Basingers say, I am a classical theist, then, as they also say, a classical theist need not (ought not, I would add) accept I-omnipotence.

Still, I do have a couple of caveats. In the first place (contrary to what they say) I do not argue that "it is possible that all creatures (creaturely essences) are such that they would go wrong with respect to at least one action in any world in which they were free with respect to morally significant actions." I have no doubt that you and I could have been significantly free but morally impeccable; there is a possible world in which we are free to do wrong but always do what is right. Indeed, for any significantly free creature there is a possible world in which that creature is significantly free but always does what is right. That this is so, furthermore, is a necessary truth; so I do not think it is possible that there be free creatures some of whom go wrong with respect to at least one action in every world in which they are significantly free. What I do think is this: there are many possible worlds God could not have actualized; and it is possible (I know of no reason to think it is true) that among these worlds are all the worlds in which there are free creatures who always do only what is right. There are plenty of possible worlds where free creatures do no wrong, but it could be that God might not have actualized any of those possible worlds (NN 168-84).

Secondly, I do not believe that "the possible world containing Hitler’s actions -- i.e., the actual world -- contains the greatest net amount of good over evil of any possible world containing free moral agents which God was free to actualize." In the first place, there is not just one possible world in which Hitler exists and perpetrates his abominations; there are any number of such worlds. More importantly, I do not believe the actual world "contains the greatest net amount of good over evil of any possible world containing free agents" God could have actualized; I see no reason to think there is any such world. Perhaps for every world God could have actualized, there is another, containing an even better balance of good over evil (1:9).

Third and most significant: I fear the Basingers perpetuate a confusion Griffin perpetrates about defenses. The Basingers quote Griffin as follows:

Of course, one can extend the free-will defense to the subhuman realm, without positing any inherent power of self determination to its entities, by pointing to the irrefutable possibility that all evils in this realm are due to Satan and his cohorts. But such a suggestion only returns to the previous point about the general illumination that theism needs to provide to render itself plausible in our day.

They then make the following comment:

It is true, of course, that by appealing to the freedom of Satan and his cohorts to explain natural evil, Plantinga himself, has adopted a defensive, seemingly ad hoc manner of preserving the consistency of his position. But it must be recognized that one who adopts a Plantingan free-will position need not necessarily respond in this manner.

Both these comments, I believe, conceal confusion. The Free Will Defense is not a theodicy, and it is not an attempt to explain the existence of evil; it is a defense. In particular, it is a defense against the charge of inconsistency or contradiction. Numberless hordes of atheologians have claimed that a proposition most theists believe -- God is omnipotent, omniscient, and wholly good -- is logically inconsistent with another proposition they believe, namely, that there is evil (or that there is some specific kind of evil such as e.g., nonmoral evil). The Free Will Defense is an effort to show that these two propositions are jointly consistent by finding a proposition that is consistent with the first and such that its conjunction with the first entails the second. It therefore appeals to the following truth of modal logic:

It is possible that (p & r) & ((p & r) -- q) -- it is possible that (p &q).

In the particular case at hand, p is God is omnipotent, omniscient, and wholly good and q is there is evil. Now what sorts of conditions must r meet, if the argument is to be successful? Clearly it need be neither true, nor probable, nor plausible, nor believed by most theists, nor anything else of that sort. All it needs to be is consistent with p and such that in conjunction with p it entails q; and if there is such a proposition r, then p and q are consistent. The fact that a particular proffered r is implausible, or not congenial to "modern man," or a poor explanation of q, or whatever, is utterly beside the point.

Since this point has proven unduly unappreciated, let me belabor it a bit. Suppose you have a bright but impetuous student who has been reading epistemology and become enamored of various "High Accessibility principles. In particular, he embraces this claim: if a person is rationally justified at a time t in believing a proposition p, then he knows, at t, that he is rationally justified in believing p. You remonstrate with him as follows: first, you observe that

P In 1879, W. K. Clifford was justified in believing that ship owners should not send their ships to sea without checking their seaworthiness

is consistent with

R In 1879, W. K. Clifford had never thought about epistemology and had not acquired the concept of rational justification, so that he didn’t believe that he was rationally justified in believing that ship owners ought not to send their ships to sea without checking their seaworthiness.

Next, you point out that P and R together entail

Q In 1879, W. K. Clifford did not know that he was rationally justified in believing that ship owners ought not to send their ships to sea without checking their seaworthiness.

"If so," you conclude (perhaps a bit pedantically) "P and Q are consistent, so that your principle isn’t true." Now suppose your student responds as follows: "Look," he says, "by 1879 Clifford had been lecturing and writing about epistemology for years. In fact in 1879 he published his Lectures and Essays, containing that famous piece ‘The Ethics of Belief’. How could he have done that, if he had never thought about epistemology and hadn’t so much as acquired the concept of rational justifiability? Your R is utterly implausible. No informed person could believe it." This response, obviously enough, is entirely beside the point. R doesn’t have to be plausible to do its job; all it has to be is consistent with P and such that in conjunction with P it entails Q.

But the same holds in the case of the Free Will Defense. Griffin apparently thinks hypotheses about Satan and his cohorts implausible. Plausibility, of course, is in the ear of the hearer, and no doubt many people do find such hypotheses implausible. Their plausibility or lack thereof, however, has nothing whatever to do with their role in the Free Will Defense. That defense is aimed at establishing just one thing: that the relevant P and Q are jointly consistent. Any proposition, plausible or not, that is consistent with P and together with P entails Q will do the trick. Some people seem to think that if you employ an implausible H, then somehow you are committed to it: they seem to think that your claim -- that P an are jointly consistent -- is no more plausible than the R you use to establish it. But of course that is a confusion; it confuses an argument that P and Q are consistent with the very different enterprise of explaining Q, given the truth of P.

Now of course someone might set out on this different enterprise; he might try to explain the existence of evil, or of nonmoral evil, from the theistic perspective; he might try to explain why it is that God permits the various sorts of evil we do in fact find. Here questions of plausibility are indeed relevant; a good explanation will not be unduly implausible. Whether you find a hypothesis implausible, of course, depends on what else you believe. One who does not believe in God, for example, may find the existence of free, nonhuman, immaterial persons such as Satan quite implausible; one who already believes in the existence of at least one such free, nonhuman, immaterial person -- i.e., God -- may find it much less implausible. But at any rate plausibility is relevant to this enterprise.

Now of course it may turn out that the theist cannot think of a plausible explanation for the evil we find. He may believe in God, believe that God is both good and powerful, and believe that God has a reason for permitting evil -- a reason for each specific evil; but he may have nothing but the most general idea as to why God permits these evils. And this can constitute a problem for him. Perplexed and disturbed about a horrifying evil in his own life or the life of some one close to him, the believer may find it hard to trust God, may come to question God’s goodness or his concern, may even come to rebel against God. His perplexity about God’s reasons for permitting evil can thus precipitate a spiritual crisis. But as an intellectual or theoretical problem it does not come to much. If God is good and powerful as the theist believes, then he will indeed have a good reason for permitting evil; but why suppose the theist must be in a position to figure out what it is? And now suppose the best the atheologian can do by way of an antitheistic argument from evil is to point out that theists do not have an explanation for evil: then theism has nothing to fear from him.

One final matter. The Basingers believe "that most influential classical theists -- e.g., Augustine, Aquinas, Luther, Calvin -- have affirmed I-omnipotence"; they go on to say that "unfortunately, Plantinga, himself, has not explicitly acknowledged the fact that his analysis of the relation between divine sovereignty and human freedom is basically an attack upon, not a defense of, the view of omnipotence that most classical theists seem to hold." But I very much doubt that there is any one view of omnipotence clearly accepted by most classical theists. A being x has I-omnipotence, you recall, if for every possible state of affairs 5, it is within x’s power to cause S to be actual. And some classical theists -- Leibniz, for example -- seem fairly clearly to affirm that God is I-omnipotent. Many others, however, do not, or do not clearly. Take the four the Basingers mention: Augustine, Aquinas, Luther, and Calvin. Augustine’s views on these matters are not clear, but in De Libero Arbitrio he offers one of the first versions of the Free Will Defense and in some passages seems to suggest that it is not within the power of God to cause it to be the case that someone freely performs an action. Aquinas discusses the matter of omnipotence in some detail and at any rate pays lip service to the claim that God has I-omnipotence. In Summa Theologiae I, Q.25, a.4, however, Thomas discusses the question whether it is possible for God to restore a fallen woman to virginity. Consider Miss X, who is now no longer a virgin: is it now within God’s power to bring it about that she is a virgin? The question is not: is it now within God’s power to bring it about that Miss X is both now a virgin and formerly a fallen woman; that state of affairs is impossible in the broadly logical sense. The question, instead is just this: can God now bring it about that Miss X is a virgin -- i.e., that she is not now and never has been a fallen woman? That state of affairs -- Miss X’s being now a virgin -- is indeed possible, but Aquinas concludes that it is not within God’s power to cause it to be actual. So Aquinas seems to affirm I-omnipotence, but also to deny it.

Calvin sometimes speaks of omnipotence, but I know of no passage in which he discusses the question whether omnipotence consists in being able to actualize just any possible state of affairs. Furthermore, I know of no passage in which he says something from which we could reasonably infer that he thinks of omnipotence in that way; so far as I know, he doesn’t raise that issue. As for Luther, I do not know his thought on this matter. Some of the medievals, however, apparently thought that every state of affairs, whether possible or not, is such that it is within God s power to cause it to be actual. Descartes, shared this view; he clearly teaches that God’s power is absolutely unlimited, so that it is within his power to cause to be actual any state of affairs whatever (DGHN 92-146). My guess is that Luther is closer to Descartes, here, than to Aquinas or Leibniz. In any event, it should be clear that there is not any one conception of God’s omnipotence common to classical theists.



DGHN -- Alvin Plantinga. Does God Have a Nature? Milwaukee: Marquette University Press, 1980.

NN -- Alvin Plantinga The Nature of Necessity. Oxford: Clarendon Press, 1974.

1. Alvin Plantinga. "The Probabilistic Argument from Evil," Philosophical Studies (1979).