Why Does the World Exist?: An Existential Detective Story

by Jim Holt

  I didn’t pursue this potentially impious line of questioning. I had made enough demands on Swinburne’s hospitality, on his supply of tea and biscuits, and perhaps on his intellectual patience as well. The windows of his study had grown dark with the early sunset. It was time to go. I thanked him effusively, and he gave me some advice on which restaurants I might try that evening in Oxford.

  The birdsong had long since quieted down when I left Swinburne’s apartment building. Ambling back onto the main road, I noticed again the prominent Eastern Orthodox Church that loomed nearby. It seemed an odd intrusion of Byzantium into North Oxford. Swinburne had told me he was an Orthodox communicant. Did he worship there? With his sacerdotal manner and his elongated, slightly severe features, this Oxford philosopher of science and religion could almost take his place in an Eastern church mosaic, right next to the other Byzantine divines:

  O sages standing in God’s holy fire

  As in the gold mosaic of a wall . . .

  Say, was that “a great cathedral gong” I heard in the distance?

  No, it was just the bells of Oxford summoning me back toward the High Street. On reaching this destination, I went into one of the restaurants Swinburne had recommended, the Quod Brasserie. It was half full and fairly lively, provincial in an academic way that contrasted with the cosmopolitan Café de Flore back in Paris. Taking a table by myself, I ordered smoked haddock and a tomato salad, along with a split of champagne and a full bottle of Australian Shiraz, and mindlessly read that day’s issue of the Guardian as I ate and drank. By the time I left the place, it was close to midnight. Walking down the almost deserted High Street back toward my hotel, I was engulfed by a diffuse sense of contentment, and I temporarily ceased to care about the mystery of existence.

  Interlude

  The Supreme Brute Fact

  Richard Swinburne seems to have solved one mystery at the price of introducing another. He purports to explain the world’s existence by positing a God who created it. But he concedes that he can find no explanation for God himself, whose existence, compared to the stark simplicity of Nothingness, strikes Swinburne as “vastly improbable.” Is this the best that theism can do—cap off its cosmic explanation with an inexplicable being, a Supreme Brute Fact?

  Traditional philosophers of theism have not thought so. They have held that God, unlike the world, exists by his very nature. He contains within himself the principle of his own being. There are many technical terms for this. God is causa sui, the “cause of himself.” He possesses aseity, the property of being self-existent. He is the ens realissimum, the most real being, and the ens necessarium, the necessary being.

  But is there any justification for all this verbiage?

  Consider, for example, the term causa sui. It seems to suggest that God somehow caused himself to exist. But even medieval theologians refused to go that far. No being, they held, could possibly bootstrap itself into existence. Regardless of how powerful the being in question might be, it would have to exist before it could exercise its causal powers.

  To say that God is causa sui is really to say that he is uncaused. His existence needs no cause because it is necessary. Or, to put the point somewhat differently, his existence needs no explanation because it is self-explanatory.

  And how might the existence of such a self-explanatory being be demonstrated? One traditional route is the cosmological argument for the existence of God. This argument goes back to Aristotle, but the most sophisticated version of it is due to Leibniz, and it goes like this.

  The universe is contingent. It might not have existed. Given that it does exist, there must be an explanation for its existence. It must have been caused to exist by some other being. Suppose that this being, too, is contingent. Then it requires an explanation for its existence as well. And so on. Now, either the explanatory chain eventually comes to an end, or it does not. If it does come to an end, the last being in the chain must be self-explanatory. If it goes on ad infinitum, then the entire chain of beings stands in need of an explanation. It must have been caused by some being outside the chain. Then the existence of that being must be self-explanatory. In either case, the existence of a contingent world must ultimately be explained by something whose existence is self-explanatory.

  Once the existence of a self-explanatory being is deduced, just a little logical tinkering is needed to show that this being has the properties traditionally ascribed to God. (It was Samuel Clarke, an English theologian and friend of Isaac Newton, who supplied the details.) Start by observing that a self-explanatory being must exist as a matter of necessity. And if it exists necessarily, it must exist always and everywhere—that is, it must be eternal and infinite. It must also be powerful, since it caused the contingent world to come into existence. Moreover, it must be intelligent, since intelligence exists in the world and therefore must exist in its cause. And since it is also infinite, it must be infinitely powerful and infinitely intelligent. Finally, it must be morally perfect. For, being infinitely intelligent, it can never fail to apprehend the truth as to what is good; and, being infinitely powerful, it can never be prevented by any weakness of its own from acting in accordance with that truth.

  The preceding reasoning, intended to show that the necessary being deduced in the cosmological argument must be God-like, is obviously rife with fallacy. But how about the cosmological argument itself? How valid is it? In essence, Leibniz was attempting an inference from contingency to necessity: if there is a contingent world, and if everything has an explanation, then there must be a necessary being that explains the existence of this world. Leibniz’s first premise looks okay. There does seem to be a world, and it does seem contingent. The second premise, which is Leibniz’s famous Principle of Sufficient Reason, is more dubious. Even Swinburne denied that there was an explanation for absolutely everything. And without that premise, the cosmological argument collapses.

  But, valid or not, there is something peculiar about the cosmological argument. It is supposed to take us from an empirical premise—arising from our experience of the actual universe—to a necessary being. But if there is such a necessary being, why do we need this empirical premise to deduce its existence? Why can’t we infer its existence directly, through pure reason?

  There is, as it happens, a notorious bit of reasoning which attempts to do just that. It is called the ontological argument. Unlike the cosmological argument for the existence of God, the ontological argument has no need of the premise that a world exists, or of the premise that there is an explanation for everything. The ontological argument purports to establish God’s existence through logic alone. God must exist as a matter of logical necessity, it says, since he possesses all perfections, and it is more perfect to exist than not to exist.

  The ontological argument was invented in the eleventh century by Saint Anselm, an Italian monk who eventually became the archbishop of Canterbury. The gist of it seems to have come to this monk one day during his morning prayers. God, reasoned Anselm, is by definition the greatest and most perfect thing that can be conceived. Now, suppose God were merely an object of thought—something, that is, which existed only in our imagination. Then it would be possible to conceive of another being exactly like God except that this being existed in reality too. And since it is greater to exist in reality than merely in the imagination, this being would be greater than God—which is absurd. Therefore, God’s nonexistence is a logical impossibility. “So truly, therefore, dost thou exist, O Lord, My God, that thou canst not be conceived not to exist,” concluded the prayer in which Anselm expressed his argument.

  Could the ontological argument possibly be valid? Even those who believe in God may feel it is too good to be true. Aquinas did not accept it. Descartes did, although he put it into a somewhat different form. Leibniz felt it needed an extra premise, namely, that God is a possible being—which premise Leibniz easily supplied by showing that God’s various perfections were all compatible with one another. Schopenhauer dismissed the onto
logical argument as “a charming joke.” Bertrand Russell, by contrast, describes in his autobiography how as a young man he was struck by its seeming truth:

  I remember the precise moment, one day in 1894, as I was walking along Trinity Lane, when I saw in a flash (or thought I saw) that the ontological argument is valid. I had gone out to buy a tin of tobacco; on my way back, I suddenly threw it up in the air, and exclaimed as I caught it: “Great Scott, the ontological argument is sound.”

  Later in his philosophical career, Russell decided that the ontological argument was not sound after all. Still, he observed, “it is easier to feel convinced that it must be fallacious than it is to find out precisely where the fallacy lies.”

  Russell’s observation has been borne out by contemporary anti-theists, whose critique of the ontological argument often boils down to mere mockery. For instance, Richard Dawkins, in The God Delusion, dismissed the ontological argument as “infantile,” a bit of “logomachist trickery,” but he did not take the trouble to identify the defect in its logic. The very idea that “a grand truth about the cosmos should follow from a mere word game” struck Dawkins as simply ridiculous, and that was the end of the matter as far as he was concerned.

  But what exactly is wrong with the ontological argument? Anselm’s reasoning, succinctly put, runs like this:

  1. God is the greatest imaginable being.

  2. A being that exists is greater than one that is merely imaginary.

  Therefore:

  3. God exists.

  Premise (1) can scarcely be disputed, since it embodies the very definition of God. Premise (2), though, looks a little funny. Just how much greater is it to exist in reality than to exist merely in the imagination? Am I, by dint of my reality, greater than the imaginary Emperor of Ice Cream?

  Think for a moment about the phrase “exists merely in the imagination.” While it’s a familiar enough locution, it has distinctly odd implications if taken literally. It suggests that the being in question is real, yet somehow confined to a tiny piece of territory—our heads. And clearly such a cerebrally confined being is less great than one that is free to manifest itself in the cosmos at large. But that can’t be right. What is in our heads is not the thing itself, but the idea of the thing. And the idea is nothing like the thing. (You can ride a unicorn, for example, but you cannot ride the idea of a unicorn.) To say that a being “exists merely in the imagination” is really a façon de parler. It does not entail that the being in question exists in some limited way. Rather, it asserts that we have a certain idea/concept/image in our minds, but that no being corresponds to this idea/concept/image. An idea of God is not a kind of God, albeit less perfect, any more than a painting of a piece of fruit is a kind of fruit, albeit less nutritious.

  Suppose, however, we forget about “imaginary existence” and simply concede that it is more perfect to exist than not to exist. Then God, possessing all perfections, must exist, mustn’t he? So what is wrong with Anselm’s reasoning?

  The most celebrated objection to the ontological argument was brought by Kant. Existence, Kant claimed, is not a real predicate. In other words, being existent is not an ordinary property of things, like being red or being intelligent. This objection is routinely cited by all who would dismiss the ontological argument—Dawkins, for example. If existence is not a property of any sort, then it can’t very well be a perfection.

  Is Kant’s dictum—existence is not a predicate—valid? Existence certainly seems a peculiar property in one sense: it is universal. Unlike the properties of redness or intelligence, absolutely everything has it. Just try to name something that does not exist. Santa Claus? To say “Santa Claus does not exist” is not to attribute nonexistence to some entity; it is merely to say that nothing satisfies the description jolly fat man who lives with elves at the North Pole and who distributes toys to children the world over on Christmas Eve. Even to say, “There is something that does not exist,” is self-contradictory, since the “there is” part asserts the very existence that the “does not exist” part denies.

  It is not obvious why the mere fact that existence is universally possessed should disqualify it from the honor of being a property. But Kant evidently had something different in mind when he said “existence is not a real predicate.” His point seemed to be that existence adds nothing to the content of a concept. “A hundred real dollars do not contain the least coin more than a hundred possible dollars,” he wrote, adding, “My financial position is, however, affected very differently by a hundred real dollars than it is by the mere concept of them.”

  And here Kant is certainly right. Suppose I take a concept like current member of the United States Senate. There are precisely one hundred individuals of whom this concept is true. Now suppose I add existence to the concept, getting current existing members of the United States Senate. Lo and behold, this new concept is true of the same one hundred individuals that the old one was!

  So adding existence to a concept doesn’t give it any extra heft. Nor does it fortify the existential chances of the would-be object defined. Otherwise, we could bring into being all sorts of wonderful things merely by defining them in the right way. This point was made by Saint Anselm’s earliest critic, a fellow eleventh-century monk called Gaunilo of Marmoutier. By Anselm’s logic, Gaunilo observed, we could demonstrate that somewhere on the ocean there must be an ideally pleasant “lost island,” since actual existence necessarily numbers among this island’s perfections.

  What happens, from a logical point of view, when we deny God’s existence? Well, suppose we define God in the same theologically orthodox way that Saint Anselm did, as an infinitely perfect being. And, just to give Anselm’s side the advantage, let’s explicitly build existence into his definition:

  x is God if and only if x is infinitely perfect and x exists.

  Then to say, “There is no God,” is to say:

  There is no x such that x is infinitely perfect and x exists.

  But this is equivalent to:

  For every x, either x is not infinitely perfect or x does not exist.

  And there is nothing inherently self-contradictory about that proposition. Indeed, it would be true of a world in which every entity fell short of infinite perfection—which is precisely the sort of world that atheists claim we live in.

  Still, there is a reason why Anselm felt it was self-contradictory to deny the existence of God. That is because we use “God” not just as shorthand for a description—infinitely perfect being—but also as a name. If God is infinitely perfect and therefore existent, how could he fail to exist?

  To see what’s wrong with this way of thinking, consider a description that is formally similar: oldest living man. Suppose we decide to call the oldest living man (whoever he might be) “Methuselah.” Now ask the question, Is Methuselah alive? Well, of course he is. By definition, he’s the oldest living man. How could he fail to be alive? But if Methuselah cannot fail to be alive, then he can’t possibly be dead. He must be immortal! Such are the logical perils of sticking a name on a definition.

  So the ontological argument, in its classic Anselmian version, is unsuccessful. Even if existence is built into the very definition of God, it does not follow that there is a being that satisfies this definition. Is that the end of the matter?

  As it happens, no. In recent decades, the ontological argument has been resurrected in an apparently more powerful form. The new version relies on a kind of logic undreamt of by Saint Anselm: modal logic. Modal logic outstrips the resources of ordinary logic. Whereas ordinary logic concerns itself with what is and is not the case, modal logic deals with what must be the case, what might be the case, and what could not possibly be the case—a far stronger set of notions.

  Modal logic was developed by some of the greatest twentieth-century logicians, including Kurt Gödel and Saul Kripke. It was Gödel, the author of the notorious “incompleteness theorems,” who saw in modal logic a way of reviving the ontological argument in a strength
ened form. The idea seems to have come to him in the early 1940s, but he did not divulge it until a few years before his death (by self-starvation) in 1978. Whether Gödel was convinced by his own version of the ontological argument is unclear. But he was certainly open to the existence of God, maintaining that it might be possible “purely rationally” to reconcile the theistic worldview “with all known facts.”

  Gödel is not the only one to have noticed the theological uses of modal logic. Independently of him, several philosophers have come up with similar modalized updates of Anselm’s reasoning. The most prominent of them is Alvin Plantinga, a professor at the University of Notre Dame. Plantinga’s efforts to secure the existence of God by logic alone have even attracted the attention of Time magazine, which hailed his “tough-minded intellectualism” and called him “the leading philosopher of God.”

  The modal ontological argument for God’s existence can look dauntingly technical. Gödel expressed the argument in a series of formal axioms and theorems, and Plantinga took the better part of his treatise The Nature of Necessity to lay out all the details. Still, the nub of it can be put in a fairly simple form.

  A truly great being, the argument begins, is one whose greatness is robust in the face of chance. Such a being not only is great, but it would have been great even if events had turned out differently from the way they actually did. By this criterion, for example, Napoléon was not truly great, since he might have died of the flu as a child in Corsica instead of growing up to conquer Europe. Indeed, if his parents had arranged their schedule of sexual congress differently, Napoléon might not have existed at all.