Dominican Friar. Let us begin with certain assumptions. God’s choice to actualize this world—the world you and I live in—was a free choice. He could have chosen otherwise; He could have actualized a different order of things. This implies alternatives, which raises the question: What is the existential status of these non-actualized possibilities? If we deny them any status at all, then God’s creative decision risks appearing baseless, as though He chose irrationally out of sheer nothingness.
Disciple of Maimonides. But in weighing the merit of each possible order, would not God be contemplating objects that are, in some sense, nonexistent? And how could there be knowledge of what is not? A nonexistent has no properties—not even the property of nonexistence. If true knowledge is correspondence between idea and object, then knowledge of nonexistents is not knowledge at all. Moreover, if a nonexistent were later actualized, would this not imply a change in the knower, as though God’s knowledge increased with creation?
Dominican Friar. To this I reply: God can indeed know what does not exist, for all such possibles correspond to ideas within His intellect. They are known as aspects of His own essence—the fullness of being from which all creatures might flow. Thus God’s knowledge of nonexistents is simply His perfect self-knowledge. In Him, nothing can be added or diminished. Therefore, to know possible creatures that never come to be is not to know nothing, but to know Himself as the source from which they could be.
Disciple of Maimonides. Yet I must press the point further. If knowledge is, as both our traditions affirm, a correspondence between intellect and reality, then this talk of God’s knowing nonexistents falters. For what is there in them to correspond to? By definition, a nonexistent is not real. It has no actuality, no fact, no determinate property by which truth could be measured. And if there is no fact of the matter, then there can be no truth of the matter; and where there is no truth, there is nothing to know. Knowledge, after all, is not knowledge of shadows or fictions, but of truth. Tell me then, what is there for God—or for any mind—to know in what does not exist?
Disciple of Abelard. If I may, I should like to untangle this knot. The trouble, it seems to me, is purely linguistic. When you say “nonexistent,” you are already speaking as if of some thing. But what is it you speak of? A word, perhaps; an intention of the mind, perhaps—but not a res, not a thing. To grant the nonexistent standing is merely to mistake the grammar of our speech for the structure of reality.
Dominican Friar. Permit me to refine the distinction. Not all nonexistents are of one kind. Some are impossibilities, like a square circle—mere words, carrying contradiction within themselves. These are no possibilities at all, for what contradicts itself could never be.
Others are fictions, like the centaur or golden mountain. These can be imagined, for they are cobbled together from elements of reality; yet they lack any true ordination to existence. God has not planted in them a real aptitude for being; they live only in the fancy of our mind.
But beyond these stand genuine possibilities. These are free of contradiction and capable of existence, were God to will them. Their potency to be is not their own—for as nonexistents they are nothing—but arises from God, whose power could bring them forth. Thus they differ from fictions: fictions are mere figments, but genuine possibilia are grounded in the divine intellect as reflections of His inexhaustible plenitude. And so, when God knows them, He knows not shadows but Himself as the source of all that could be.
Disciple of Leibniz. There is here an indissoluble link between logic and metaphysics. The realm of possibles is simply the whole space of noncontradiction. Every possible world is a maximal set of consistent propositions, and so there are infinitely many. Yet it does not follow that two possible worlds can be blended together; each world is self-contained, consistent in itself but not necessarily across worlds.
Now, in actualizing the best world, God must, as it were, solve infinitely many equations in infinitely many unknowns. This begins to show why contingency differs from necessity: necessary truths admit of finite demonstration, but contingent truths are infinitely analytic—their resolution proceeding without end and converging not merely on God’s intellect, but upon His free will.
Thus possible worlds are as real as anything else—but they are not contingent (i.e., actual), like our world; they exist necessarily, etched like logical blueprints in the divine intellect. By contrast, the actual world, the one in which we live, is contingent, because it issues not from God’s intellect alone but, additionally, from His free will.
And here the matter reaches its climax. Since the realm of possible worlds is as real as anything else (though never contingent/actual), and since it cannot be situated anywhere except within the divine intellect, it follows that if there were no God, then not only would nothing be actual—nothing would even be possible. Thus we must go beyond Berkeley and declare: to be possible is already to be an idea in the mind of God.
Dominican Friar. Friend, I find in your discourse no rupture with our tradition, but a noble elaboration of it. What you have called possible worlds, we have long called divine ideas; what you describe as the space of noncontradiction, we have known as the breadth of God’s power. You speak in the idiom of analysis and equation, yet the heart of your claim is one with ours: all possibility is grounded in God, and without Him not only would nothing be actual, but nothing could even be possible. This is no foreign doctrine, but the very truth the Church has confessed through the ages—that in God we live and move and have our being, and that in Him even the might-have-beens find their order and their measure.
World-Bound Metaphysician. Yet the Leibnizian account raises a pressing question. If there are infinitely many possible worlds in the divine intellect, what are we to say about individuals who appear in more than one of them? Assume individuals x1 and x2 share all properties except world membership: x1 belongs to world w1 and x2 to world w2. By Leibniz’s Identity of Indiscernibles, x1 = x2 if and only if they have precisely the same properties. But they do not, since x1 has the property of belonging to w1 while x2 has the property of belonging to w2. Therefore, x1 ≠ x2. No identity across worlds!
Trans-World Metaphysician. Not so. World-membership is not an essential property but a derivative one. If individuals and worlds alike are contingent, then one and the same x may be considered under different possible arrangements of properties. Were this not so, we could not speak of genuine alternatives for x at all!
Pragmatist. Let us be plain. No one has ever encountered a possible world; we inhabit only the actual. When we speak of x in w1, w2, or wn, we stipulate by assumption that it is the same x under different suppositions. This is not an empirical discovery but a convention of thought. Trans-world identity, so far as human reasoning is concerned, is trivial: we posit it in order to consider counterfactuals, but we never observe it.
Modern Semanticist. If we are inquiring about the actual world—the world you and I live in—it makes perfect sense to ask whether two incompletely described individuals are, in fact, identical. Frege’s example is well known: the “morning star” and the “evening star.” At first, the descriptions differ; yet they pick out the same object, Venus. In general, individuals are always more descriptively complex than we know, so a partial description is enough to anchor our reference. We may ask, for instance, whether “the beagle that lives in that house” and “my dog” refer to the same animal. But nothing like this obtains with possible worlds, where objects have only those descriptions we stipulate for them.
Pragmatist. Exactly so. That is my point. The problem of trans-world identity is trivial, settled not by discovery but by stipulation.
Disciple of Leibniz. And here we return to Leibniz’s own criterion of the actual. Non-actual objects have only those properties we stipulate; but an actual object exhibits never-ending, ever-unfolding descriptive detail. Thus the distinction between the actual (contingent) and the possible is precisely this: the actual bears inexhaustible descriptive richness, while the merely possible remains bound to the limits of assumption.