May 20, 2007

The Difference Between Possible and Feasible Worlds

Hello, Dr. Craig,

I recently listened to your debate with Dayton: “Does Evil Disprove God?” While I have a clear understanding of why God cannot do that which is intrinsically (or logically) impossible (i.e., create a world in which He forced people to freely choose good over evil), I am unclear as to what you meant by saying that it might not be feasible for God to create a world in which evil did not exist, even when it could be logically possible for Him to do so. As I try to conceive of such situations, they always seem to reduce to intrinsic impossibilities. Please explain and illustrate what you mean by God allowing that which is evil and preventable and yet unfeasible for God to prevent?


The distinction between possible worlds and feasible worlds is one that lies at the heart of the doctrine of middle knowledge and may have very important theological implications, such as the one that you note. The terminological distinction was first drawn by the philosopher Thomas Flint, but the conceptual distinction is inherent in Luis Molina’s theory of middle knowledge formulated in the sixteenth century.

According to Molina, logically prior to the divine decree to create a world, God possesses not only knowledge of everything that could happen (His natural knowledge) but also everything that would happen contingently in any appropriately specified set of circumstances (His middle knowledge). God’s natural knowledge is His knowledge of all necessary truths. By means of it God knows what is the full range of possible worlds, or as you put it, worlds that are intrinsically possible. He knows, for example, that in some possible world Peter freely denies Christ three times and that in another possible world Peter freely affirms Christ under identical circumstances, for both are possible.

God’s middle knowledge is His knowledge of all contingently true conditional propositions in the subjunctive mood, including propositions about creaturely free actions. For example, logically prior to His creative decree, God knew that if Peter were in circumstances C, he would freely deny Christ three times. Such subjunctive conditionals are often called counterfactuals. These counterfactuals serve to delimit the range of possible worlds to worlds which are feasible for God to actualize. For example, there is an intrinsically possible world in which Peter freely affirms Christ in precisely the same circumstances in which he in fact denied him; but given the counterfactual truth that if Peter were in precisely those circumstances he would freely deny Christ, then the possible world in which Peter freely affirms Christ in those circumstances is not feasible for God. God could force Peter to affirm Christ in those circumstances, but then his confession would not be free. By means of His middle knowledge, God knows what is the proper subset of possible worlds which are feasible for Him, given the counterfactuals that are true.

God then decrees to create certain free creatures in certain circumstances and, thus, on the basis of his middle knowledge and His knowledge of His own decree, God has foreknowledge of everything that will happen (His free knowledge). In that way, He knows, simply on the basis of His own internal states and without any need of any sort of perception of the external world, that Peter will freely deny Christ three times.

Thus on the Molinist scheme, we have the following logical order:

So there are worlds which are intrinsically possible but which God, given the counterfactuals that happen to be true, is not capable of actualizing and which are therefore, in Flint’s terminology, infeasible for God. Notice that because counterfactuals of creaturely freedom are contingently true, which worlds are feasible for God and which are infeasible is also a contingent matter. It all depends on how creatures would freely behave in various circumstances, which is beyond God’s control.

Alvin Plantinga was the first contemporary philosopher to apply this scheme to the problem of evil. In response to J. L. Mackie’s claim that since a world in which everyone always chooses to do the morally right thing is intrinsically possible, an omnipotent God should be able to create it, Plantinga pointed out that for all we know such a world may not be feasible for God. Indeed, for all we know, all the worlds which are feasible for God and which involve as much good as the actual world also involve as much evil. Hence, although a world with as much good as the actual world but with less or no evil in it may be intrinsically possible, it may not be within God’s power to create such a world. Hence, God cannot be indicted for not having created such a world. The atheist who pushes the problem of evil would have to show that worlds with as much good but less evil are feasible for God, which is beyond anyone’s power to prove; it is sheer speculation. Thus, the atheist has failed to bear his burden of proof.

In my own work I’ve tried to exploit the distinction between possible and feasible worlds in dealing with such questions as perseverance of the saints, biblical inspiration, and Christian particularism (see “Scholarly Articles: Omniscience; Christian Particularism”).