God has had a lot of bad press recently. The four horsemen of atheism, Richard Dawkins, Daniel Dennett, Sam Harris, and Christopher Hitchens, have all published books sharply critical of belief in God: respectively, The God Delusion, Breaking the Spell, The End of Faith, and God Is Not Great. Dawkins, Harris, and Hitchens pile on the greatest amount of scorn, while Dennett takes the role of good cop. But despite differences of tone and detail, they all agree that belief in God is a kind of superstition. As Harris puts it, religion “is the denial—at once full of hope and full of fear—of the vastitude of human ignorance.”
The question of God’s existence is one of those few matters of general interest on which philosophers might pretend to expertise—Dennett is a professional philosopher, and Harris has a B.A. in the subject. Still, of the four, it is Dawkins who wades the furthest into philosophy. So what can philosophy contribute? In particular, have philosophers come to a verdict on the traditional arguments for God’s existence?
Although it would be too much to expect complete consensus, it is fair to say that the arguments have left the philosophical community underwhelmed. The classic contemporary work is J. L. Mackie’s The Miracle of Theism, whose ironic title summarizes Mackie’s conclusion: the persistence of belief in God is a kind of miracle because it is so unsupported by reason and evidence. The failure of arguments for God’s existence need not lead straight to atheism, but philosophers often seem to find this route tempting. In his contribution to Philosophers Without Gods, a collection of atheistic essays by twenty prominent philosophers, Stewart Shapiro observes that “among contemporary philosophers, the seriously religious are a small minority.” Dean Zimmerman, a notable member of the minority, has ruefully remarked that “although numerous outspoken Christians are highly respected in analytic circles, many of our colleagues still regard the persistence of religious belief among otherwise intelligent philosophers as a strange aberration, a pocket of irrationality.”
The world was very different when a distinguished philosopher could say, as St. Thomas Aquinas did, “the existence of God can be proved in five ways.” Contemporary Christian philosophers often content themselves with pulling up the drawbridge and manning the barricades, rather than crusading against the infidel. Alvin Plantinga, perhaps the most eminent living philosopher of religion, devotes the five hundred pages of his Warranted Christian Belief to fending off objections to either the truth or rationality of belief in traditional Christian doctrines. He does not argue for the existence of God, and still less for the truth of Christianity; rather, his main question is whether a reasonable person who finds herself with firm religious convictions should change her mind. Plantinga is not trying to persuade Dawkins and company to change their minds.
The traditional arguments for God’s existence are very much worth our attention, though, for at least three reasons: they are of great intrinsic interest; popular discussions of them often fail to pin down their defects; and one argument, the “design argument,” has had a new lease on life as the intellectual underpinning of the intelligent design movement.
Before turning to some of the arguments, who or what is God supposed to be? Zeus, Thor, Ganesh? Alternatively, the depersonalized Deus sive natura (God or nature) that got Spinoza excommunicated from Amsterdam’s Jewish congregation? The philosophical literature focuses on the God of the Abrahamic tradition: a person who is all-knowing, all-good, and all-powerful. Is there any reason to think that God, so conceived, exists?
Arguments for the existence of God are usually divided into those whose premises may be known from the armchair, and those whose premises are the result of experiment and observation. The best-known armchair argument is called (following Kant’s unhelpful terminology) the “ontological argument,” while the design argument (also called the “teleological argument”) is the main representative of empirical arguments. Let us start from the armchair.
The ontological argument was first developed by the eleventh-century monk St. Anselm, who spent his formative years at Bec Abbey in Normandy and later became Archbishop of Canterbury. Anselm was a central figure in early scholasticism, which brought the logical and metaphysical apparatus of Aristotelianism to bear on the interpretation of Christian texts.
In chapter two of his Proslogion (“Address”), Anselm considers the Fool of Psalm 14, who “hath said in his heart: There is no God.” Anselm argues that the Fool’s position is self-undermining: the very act of denying that God exists shows that God does exist. It is as if the Fool were to say, very foolishly, “I am not speaking.”
God, Anselm says, is a perfect being, “a being than which nothing greater can be conceived.” We may assume that any ignorance or malice or feebleness detracts from greatness, so Anselm’s God is all-knowing, all-good, and all-powerful. Just for simplicity, let us also assume that there could be at most one perfect being, so Anselm’s God is unique. Anselm then draws a distinction between “existing in the mind” (or “in thought,” or “in the understanding”) and “existing in reality.” When a painter intends to paint a picture of, say, a dragon, the picture, and the dragon, exist in his mind but not in reality. When he has finished putting paint on canvas, the picture, but not the dragon, also exists in reality. Dragons—as opposed to pictures of dragons, or the word “dragon”—exist only in the mind. Conversely, there are many things that exist only in reality: a certain rock at the bottom of the Pacific, say, which no one has ever seen.
Having explained this distinction, Anselm observes that the Fool must admit that God exists in his mind, just as the Fool must admit that a dragon exists in the painter’s mind. Dragons, of course, exist only in the mind. The Fool will say the same of God. Anselm thinks the Fool can be hoisted by his own petard.
Here we come to the crucial step in Anselm’s argument. An entity that exists only in the mind, he thinks, is not as great—not so perfect—as one that exists in reality. I imagine a dry martini: unfortunately it exists only in my mind. You imagine a martini, shake the gin and vermouth, and add the olive: happily for you, the martini exists both in your mind and in reality. According to Anselm, the martini that exists only in the mind is less perfect than the martini that also exists in reality—and after a long day at the office, this can sound quite convincing. Similarly, a being that only exists in the Fool’s mind is not as perfect as one that also exists in reality. So if God exists only in the Fool’s mind, the Fool is not thinking of a perfect being, because a perfect being also exists in reality. Equivalently: if the Fool is thinking of a perfect being, then God exists in reality. The very existence of atheists, Anselm concludes, shows that “something than which greater cannot be conceived undoubtedly exists both in the mind and in reality.”
Should we agree with Dawkins that something has gone badly wrong, on the grounds that Anselm’s argument reaches “such a significant conclusion without feeding in a single piece of data from the real world”? As a general reason for suspicion, this is not very persuasive. In 300 BC Euclid proved that infinitely many prime numbers exist. He needed no empirical data, and surely his conclusion—infinitely many—is pretty significant.
A better complaint is that sound philosophical arguments with significant conclusions are as rare as atheists in foxholes: the track record of philosophical “proofs” is not exactly impressive, unlike the mathematical variety.
Still, the ontological argument may be an exception to the rule. A more urgent cause for concern was given by Gaunilo, an elderly monk at an abbey a few days ride from Bec. In his In Behalf of the Fool, Gaunilo considers an island than which no greater island can be conceived, “abundantly filled with inestimable riches.” (Dennett alludes obliquely to Gaunilo when he asks his reader to consider “the most perfect ice-cream sundae.”) Presumably an island that exists only in the mind is not as great as a similar island that also exists in reality. But then Anselm’s reasoning proceeds just as well, and we can conclude that a perfect island exists, which is absurd. We know a great deal about islands, and although some of them are undoubtedly very agreeable, improvement is always possible.
Gaunilo’s objection is that the argument proves too much; something must be wrong, but Gaunilo doesn’t tell us what. So what is wrong with it?
The first thing to note is that Anselm’s talk of “existing in reality” and “existing in the mind” is misleading. Possums exist in Australia and New Zealand, but not in Antarctica. If “existing in reality” were like “existing in Australia,” then there might be some other realm distinct from reality where things exist. But that’s wrong: if something exists anywhere at all, it exists “in reality,” because to exist in reality is simply to exist, period. Similarly, if “existing in the mind” were like “existing in New Zealand,” then if dragons exist in the mind then they must exist. But there are no such creatures—dragons do not exist. The observation that dragons exist in the mind but not in reality is, then, better stated as follows: people think of dragons, but dragons do not exist.
Let us return now to the assumption that Anselm tries to reduce to absurdity: that a perfect being exists only in the Fool’s mind. Unpacked, the assumption is this: (a) the Fool is thinking of a perfect being, and (b) no perfect being exists—that is, in a complete inventory of reality, we will not find a being than which nothing greater can be conceived.
So the crucial step in Anselm’s argument is this: if (b) is true, and no perfect being exists, then (a) must be false—the Fool is not thinking of a perfect being, because a perfect being has, among its other perfect-making properties or features, existence. Put the other way round: if (a) is true—if the Fool is genuinely thinking of a perfect being—then (b) must be false, and so God, the perfect being, exists.
Both Dawkins and Hitchens suggest that Kant uncovered Anselm’s mistake—and Kant certainly had an influential objection. In his Critique of Pure Reason he claims that “‘Being’ is evidently not a real predicate,” by which he means that existence is not a property or a feature of a thing. To say that dragons are green, or scaly, or ferocious, is to attribute certain properties or features to dragons. To say that dragons exist is not to attribute yet another property to them, it is simply to say that there are dragons. And if existence is not a property or feature of things, Anselm’s argument fails: a perfect being has all the perfections, including the properties of being all-good and all-knowing, but not including the property of existing, simply because there is no such property.
Kant is on to something here. If existence is a property of things, it is a rather peculiar one: you can find a blue marble, and also a non-blue marble (a red one, say), but you cannot find a nonexistent marble—a marble that lacks the property of existing. Of course, that does not mean Kant is right: a peculiar property is still a property. And in fact, according to many philosophers, Kant is wrong: existence is indeed a property, albeit a very undiscriminating one, because everything has it.
A better objection to Anselm’s argument is that he has conflated two readings of “The Fool is thinking of a perfect being.” Compare “J. R. R. Tolkien is thinking of a scaly existing dragon,” which can be read in two ways. On one reading, this sentence can be more perspicuously rendered as, “There is a scaly existing dragon, and Tolkien is thinking about it.” On that reading, the sentence is true only if at least one scaly dragon exists. But on the second, more natural reading, “Tolkien is thinking of a scaly existing dragon” can be true even if dragons do not exist. Let us ask the man himself: “Hey, Tolkien, what are you thinking about?” He replies: “I am thinking about a dragon.” “Oh, I see, you are thinking about an imaginary dragon.” “No, I am thinking about a real flesh-and-blood dragon.” Tolkien was not a postmodernist whose novels are populated with paradoxical, metaphysically insubstantial, nonexistent dragons—he wrote and thought about existing dragons. But for all that, dragons do not exist.
Now there is a similar ambiguity for “The Fool is thinking of a perfect being.” On one reading, it means, “There is a perfect being, and the Fool is thinking about it.” On the other reading, it simply characterizes the Fool’s thought: the Fool is thinking of a perfect being in the innocuous sense in which Tolkien is thinking of a scaly existing dragon.
Anselm is thus caught in dilemma. What is the intended reading of (a), “The Fool is thinking of a perfect being”? If it is “There is a perfect being, and the Fool is thinking about it,” then God’s existence immediately follows. However, Anselm has given us no reason at all to suppose that, on this reading, (a) is true, because he has not already shown us that there is a perfect being. On the alternative reading, where (a) is read as simply characterizing the Fool’s thought, we may grant that (a) is true, but it is perfectly consistent with a Godless universe.
There are other versions of the ontological argument, and the exact interpretation of the argument in chapter two of the Proslogion is a matter of dispute. Descartes offered an Anselm-inspired argument in his Meditations (it was this version that Kant criticized), and other variants can be found in Anselm’s own writings. These arguments have been subject to elaboration and repair at the hands of contemporary philosophers, Plantinga included. Graham Oppy’s Ontological Arguments and Belief in God is an exhaustive survey. However, although this work has produced much enlightenment about topics of interest to metaphysicians, it is pretty clear that a philosopher in search of God has to rise from the armchair.
Although the design argument can be traced to the ancient Greeks, it received one of its most careful and elaborate formulations from William Paley, an eighteenth-century English clergyman and philosopher, in his Natural Theology; or, Evidences of the Existence and Attributes of the Deity. That book was published in 1802, a few years before Paley’s death and more than half a century before the publication of Darwin’s On the Origin of Species.
Paley begins by contrasting the discovery of two objects while “crossing a heath”: a stone and a watch. The presence of the stone requires no explanation in terms of a designer—indeed, Paley supposes that the hypothesis that “it had lain there forever” might well be correct. The presence of the watch is another matter entirely, for on examination “we perceive—what we could not discover in the stone—that its several parts are framed and put together for a purpose, e.g. that they are so formed and adjusted as to produce motion, and that motion so regulated as to point out the hour of the day.” And the inference from these observed facts, Paley thinks, “is inevitable; that the watch must have had a maker.” Importantly, that is not because we know that watches are, in fact, usually the product of design: the conclusion, Paley says, would not be weakened if “we had never known an artist capable of making one.”
All that seems reasonable enough. The design argument that Paley then proceeds to give replaces the watch with terrestrial flora and fauna and their intricate parts. Paley—evidently a keen amateur naturalist—gives many examples, from the diverse mechanisms of seed dispersion to the tongue of the woodpecker, but his example of the eye is the one typically quoted. How could such a “complicated mechanism” have arisen, Paley asks, if not by the action of a designer? “In the human body, for instance, chance, i.e., the operation of causes without design, may produce a wen, a wart, a mole, a pimple, but never an eye.” In the case of the watch, the reasonable conclusion is that a designer produced it. And similarly, Paley thinks, in the case of the eye and other biological structures. Admittedly, we have no idea how the designer managed to construct the eye, and we have “never known an artist capable of making one.” As Paley says, however, these points of disanalogy do not seem to ruin the argument.
Unlike the ontological argument, the design argument is not supposed to prove God’s existence. Rather, it is an “inference to the best explanation,” like the inference that there are mice in the kitchen because this hypothesis best explains the missing cheese. The hypothesis of a designer is one of many possible “scientific explanations” of Paley’s watch on the heath, and similarly of the eye. The frequent complaint that intelligent design is “not science” (as opposed to “bad science”) only succeeds in muddying the waters.
An inference to the best explanation can be overturned by more evidence. Perhaps, on further investigation, it turns out that another hypothesis—say, that the au pair has been snacking in the early hours—is the best explanation of the missing cheese. And that is the standard reply to Paley: we now know that the best explanation of the apparent design of the eye is not “the hand of an artificer,” but Darwinian evolution. To borrow from the title of an earlier book by Dawkins, a blind watchmaker—the impersonal forces of natural selection—made the eye.
This reply crucially hinges on the assumption that modern biology can explain all instances of apparent design, and it is here that sophisticated proponents of intelligent design, most notoriously the biochemist Michael Behe, have seen an opportunity to dust off and burnish Paley’s argument.
In his first book, Darwin’s Black Box, Behe argues that while evolution by natural selection “might explain many things,” it cannot explain what he calls “irreducible complexity.” The notion is straight out of Paley, who writes of the watch that “if the different parts had been differently shaped from what they are, or placed after any other manner or in any other order than that in which they are placed, either no motion at all would have been carried on in the machine, or none which would have answered the use that is now served by it.” A watch is “irreducibly complex” in the sense that many of its main parts are essential to its proper functioning—remove the balance wheel, or the escapement, and all you have left is a paperweight. Irreducible complexity is everywhere in nature: Behe’s poster children are the blood-clotting system and the bacterial flagellum, but he also quotes Paley’s observation that “The heart, constituted as it is, can no more work without valves than a pump can.” According to Behe, a process of small step-by-step alterations of the sort found in natural selection is wildly unlikely to produce irreducibly complex systems.
Obviously it is a matter of great importance whether Behe’s criticism of the cornerstone of modern biology is correct. (For a clear explanation of why it isn’t, see H. Allen Orr’s review of Behe’s book in Boston Review, December 1996/January 1997.) But here the debate took a crucial turn too hastily: focusing attention on whether evolution by natural selection can explain the origin of the bacteriological flagellum is to obscure the fact that the design argument fails even if Behe is right.
David Hume’s Dialogues Concerning Natural Religion (published posthumously in 1779) presented the key objections, more than twenty years before Natural Theology. Two of Hume’s objections are especially acute. First, if the argument works at all, its conclusion is much weaker than might have been hoped. The argument does not indicate anything about what the designer is like: whether it is benevolent or a suitable object of worship. Even the intelligence of the designer is up for grabs—terrestrial biology might be the product of long trial and error, with the designer’s many previous attempts “botched and bungled.” Or perhaps the designer is “a stupid mechanic,” who imitated other much cleverer designers who practiced their art in far-off galaxies. Further, the designer could have died long ago—the eye and such might have been “the production of old age and dotage in some superannuated deity.” And finally, since the design of something complicated is usually a collective endeavor—“A great number of men join in . . . framing a commonwealth,” for example—we can hardly presume that there was exactly one designer. At best, the design argument shows that some designer or designers, whose motives, talents, and present whereabouts are all unknown, existed at some time. The proponent of the argument is at liberty afterwards “to fix every point of his theology, by the utmost license of fancy and hypothesis.” Perhaps life on Earth was designed over millions of years by successive committees of incompetent and thoroughly despicable space aliens, who are now fortunately all dead.
Paley had read Hume, and he tries to reply to this objection. Paley concedes that if the design argument simply concerns individual biological structures like the eye, then the proper conclusion is indeed weak: “there must have existed, at some time and at some place or other, an artificer or artificers.” However, he thinks a more careful study of the biological world as a whole reveals that there is exactly one designer (or at any rate one chief architect), who possesses the usual divine attributes. But Paley’s arguments on this score are feeble. He notes the general similarities in the body-plans of animals, and concludes that this “bespeaks the same creation and the same Creator,” forgetting Hume’s point that multiple designers can act in concert, or that one designer can pick up where another left off. And in support of the goodness of the deity, Paley declares, “It is a happy world after all.” Rural England is, anyway: “A bee amongst the flowers in spring is one of the cheerfullest objects that can be looked upon.” A more plausible theological conjecture is the remark attributed to the biologist J. B. S. Haldane, that the creator had “an inordinate fondness for beetles.”
Hume’s first objection is that the design argument can only establish the existence of at least one designer. His second objection is that the argument does not establish even this much. Paley claims that the evidence points to the conclusion that, by means entirely unknown, the biological world is the product of design. But why favor this over the hypothesis that, also by means entirely unknown, flora and fauna were produced by, as Paley puts it, “the operation of causes without design”?
As Paley himself emphasizes, his initial watch analogy is far from perfect: watches, unlike organisms, do not reproduce. The eye has not been found lying on its own on the heath, but in the bodies of countless creatures and their ancestors. And offspring differ in various ways from their parents. So one possibility is that the operation of causes without design, operating over “a hundred millions of years,” somehow allows, after numerous generations, a “round ball” to “acquire wings,” eyes, and so forth. Paley’s strategy for dismissing no-design alternatives wholesale is to object to the specific evolutionary theories of his day (for instance that of Erasmus Darwin, Charles’s grandfather). But this is rather like saying that because this apple and that pear are rotten, vegetables are better than fruit. What Paley needs is an argument for choosing the general hypothesis of an unknown designer or designers operating by unknown means over the general hypothesis of an unknown blind process operating by unknown means, and he signally fails to supply one.
An example that briefly appears in Darwin’s Black Box nicely illustrates the point. In 2001: A Space Odyssey, a magnetic anomaly in one of the moon’s craters leads to the discovery of a perfectly regular slab buried under lunar soil. The characters have no idea how the slab was constructed, or what it is for, and have never known an artist capable of making one; nevertheless they reasonably conclude that it was designed. But that is precisely because the characters are not in Paley’s position. They know enough about lunar geology, astronomy, and the possibility of extraterrestrial life to discredit the rival hypothesis that the monolith is a natural object (a big crystal, say) that formed on the moon or collided with it. Paley, on the other hand, had no reason, other than the failure of his imagination, to dismiss the hypothesis of “causes without design.”
Darwin’s Black Box exactly recapitulates Paley’s mistake. “Might there,” Behe asks after he has disposed of Darwin’s theory, “be an as-yet-undiscovered natural process that would explain biochemical complexity?” Assuming for the sake of the argument that Darwinism is false, Behe is surely right that “if there is such a process, no one has a clue how it would work.” But of course that is quite different from saying that there is no such process. Moreover, intelligent design is in the very same boat: if there is such a process, no one has a clue how it would work either. Why is one mysterious unknown process to be favored over another? After all, as Behe clearly brings out, biochemistry is fantastically baroque, with many unanswered questions and unsolved problems.
The version of the design argument on which Paley rests his case begins with certain features of organisms. Other versions start from the observation, in Hume’s phrase, that the entire universe is “one great machine, subdivided into an infinite number of lesser machines.” And one of these other versions has received a great deal of attention in the recent philosophical literature: the so-called fine-tuning argument. The fine-tuning argument is also in Natural Theology, although Paley is not usually credited in contemporary discussions. There are, Paley says, an “infinite number of possible laws” that could have governed material objects (in particular, the heavenly bodies), and out of “a boundless variety of suppositions which were equally possible,” and despite “a thousand chances against conveniency,” the laws that do in fact obtain are “beneficial.” The universe, in other words, is fine-tuned for life. The remarkable fact that the universe is so hospitable needs an explanation, and isn’t the hypothesis of a designer the best one?
One might object that explanation has to stop somewhere. The eye is not a credible candidate for a stopping point, but perhaps the basic physical laws are the sorts of things that have no explanation. If so, the fine-tuning argument does not get started. But let us (perhaps generously) admit that an explanation is required: why, we may ask, is the universe apparently made for life?
The fine-tuning argument did not appear in Darwin’s Black Box, but it has a starring role in Behe’s latest book, The Edge of Evolution. One of the most extensive discussions of the argument in the philosophical literature is John Leslie’s Universes, and—as the title hints—a rival explanation of fine-tuning is that our universe is only one of many universes, just as our sun is a single twinkle in the sidereal plenitude. If universes exist in “boundless variety,” each with a distinct set of basic physical laws, then the fact that the laws of our universe are “beneficial” would seem to be nothing to get excited about.
This “multiverse hypothesis” stands to the fine-tuning argument for God’s existence as Darwinism stands to Paley’s biological design argument: it is an alternative “no-design” explanation of the data. If the fundamental organizing principle of modern biology is pitted against a rival hypothesis that receives no serious consideration in professional journals, the outcome is not in doubt. But if the alternative to design is cosmological speculation (by philosophers, no less!), the contest looks to be back on a much more equal footing.
Dawkins, then, makes a significant concession when he turns in The God Delusion to the fine-tuning argument. He replies in exactly the same way he does to Paley, by arguing that the multiverse hypothesis should be preferred over the “God hypothesis,” because the former is considerably more “simple.” Well, maybe—but unlike the Darwinian reply to Paley’s argument, this point is eminently debatable. And in any case, the idea that the multiverse hypothesis could provide any kind of explanation of why our universe is fine-tuned is controversial.
Hume suggests a more convincing rebuttal. His two objections apply equally well to the fine-tuning argument for the existence of God. First, the fine-tuning argument is silent on the number and attributes of any designers. Further, it is quite unclear what the designer or designers could be like, which contrasts the fine-tuning argument unfavorably with the biological design argument. At least we may intelligibly hypothesize about the designers of the eye—perhaps a race of extraterrestrials visited the earth about half a billion years ago to manufacture the early prototypes. But if any sense can be made of agents creating the totality of space-time, it cannot be by comparison with familiar artisans like watchmakers, quilters, and pastry chefs, who do their work at particular times and places according to ordinary causal laws.
Hume’s second objection is that there is no reason to favor the (unspecific, and perhaps not even intelligible) design hypothesis over the (also unspecific) hypothesis that fine-tuning can be explained in some other way. How could we be in a position to rule out all the no-design alternatives? Hume sketchees a number of possibilities (including an ancient version of the multiverse hypothesis), of which perhaps the most interesting compares the structure of the universe to structures found in mathematics. The explanation of arithmetical structure, as any “skilful algebraist” will tell you, is not to be found either in “chance or design,” or the hypothesis of a multiplicity of other structures, but instead in the “nature of . . . numbers.” Likewise, perhaps “the whole economy of the universe is conducted by a like necessity, though no human algebra can furnish a key, which solves the difficulty.”
If a persuasive argument for the existence of God is wanted, then philosophy has come up empty. The traditional arguments have much to teach us, but concentrating on them can disguise a simple but important point. As Anselm and Paley both recognized, the devout are not exactly holding their collective breath. For the most part, they do not believe that God exists on the basis of any argument. How they know that God exists, if they do, is itself unknown—the devout do not know that God exists in the way it is known that dinosaurs existed, or that there exist infinitely many prime numbers. The funny thing about arguments for the existence of God is that, if they succeed, they were never needed in the first place.
