David Hume and the Argument from Design

Steven Dutch, Natural and Applied Sciences, University of Wisconsin - Green Bay
First-time Visitors: Please visit Site Map and Disclaimer. Use "Back" to return here.

David Hume and the Argument from Design

"The Argument From Design was thoroughly discredited by David Hume 250 years ago."

That claim is repeated in so many places, with exactly the same tone (essentially as an argument from authority) and with such virtually identical wording that it creates the powerful impression that the person making the claim has never actually read Hume, but is merely parroting an assertion he or she read somewhere else.

Hume criticizes the Argument from Design in his Enquiry Concerning Human Understanding (best known as his essay on miracles) but his most complete analysis is in Dialogues Concerning Natural Religion, published shortly after his death in 1776.

Dialogues Concerning Natural Religion

Dialogues Concerning Natural Religion is an imaginary conversation between three persons (The "di" in dialogue actually is part of the Greek root "dia" ("through") and has nothing to do with the number two. It is perfectly possible to have a dialogue among any number of people). In contrast to Galileo's famous Dialogue Concerning the Two Chief World Systems, the characters and positions in Hume's Dialogue are much less clearly defined. Hume's three characters are:

The infirmities of our nature do not permit us to reach any ideas, which in the least correspond to the ineffable sublimity of the divine attributes.

In addition there are two background characters, Pamphilus the observer and Hermippus, the intended reader. "Natural Religion" refers to what can be known of God from wholly natural phenomena, not preconditioned by prior beliefs or influenced by revelation.

In his analysis of the Dialogue, Norman Kemp Smith believes Philo most clearly represents Hume's views, but Cleanthes definitely reflects some of Hume's formative thinking as well. Hume certainly uses Philo as a vehicle to express some of his more abrasive views. On the other hand, Hume was never fully able to reconcile the results of complete skepticism with the reality that some things are known for certain even if the basis for the knowledge cannot be clearly articulated, and Cleanthes most clearly reflects that side of Hume.

Outline of the Book

In the prologue, Pamphilus explains to Hermippus why he has adopted a dialogue format for his discourse. By the mid-1700's this approach had largely fallen into disuse, but Pamphilus explains it is well suited to dealing with complex topics that would be intolerably dry if approached in a more didactic style.

Part I     The three players lay out their basic positions with regard to religion. Demea notes that he would not teach theology at all until late in a child's upbringing until after an instinctive devotion had been ingrained (a modern skeptic would call this brainwashing). Philo sarcastically endorses this position because he feels that, once we leave the realm of common experience, so many philosophical positions become defensible that it is impossible to choose between them (In other words, it does no harm to delay because it's all worthless anyhow). Cleanthes notes that extreme skeptics nevertheless do not live up to their own statements because they continually behave is if some position or other were really more valid. Philo notes that theologians at various times have embraced or opposed reason as it suits their purposes, to which Cleanthes replies, so what? Everyone makes use of the ideas that best supports his position.

Part II    Philo and Demea both assert that any real understanding of God is unattainable, Demea because God is so inscrutable, Philo because none of our conclusions can be trusted. Cleanthes basically lays out the Argument from Design. Demea is mortified at the comparison of God to a human artisan. Philo argues that we cannot validly argue by comparing what we observe in a small part of the universe to the universe as a whole. [In Hume's day, even the existence of the planet Uranus was unknown. Galaxies were visible in telescopes as fuzzy patches but were not known to be distant star systems. Spectroscopy, with its incredible power to probe the nature of stars, was a century in the future. The only laws of nature that were really known in rigorous quantitative terms were Newton's laws of motion and gravitation. It never occurred to Hume or anyone else that we might one day have ways of testing whether the laws of nature on Earth applied through the cosmos. In modern terms, Philo is dead wrong.]

Part III    Cleanthes continues to argue for Design. He points out that if we were to hear a voice coming from the clouds, we would not hesitate to ascribe to it intelligence, and if books grew on trees like fruit, we would certainly deduce from their contents that there was an intelligence behind their design. Although these are rather weird and contrived arguments, Cleanthes points out that what is actually observed in nature could hardly be more surprising, or more clearly indicative of intelligence. Demea continues to flog the dead horse of being unable to ascribe humanly comprehensible attributes to God.

Part IV    Cleanthes demolishes Demea by pointing out that a mind which has no thought is not a mind, and a view of God that makes him utterly unknowable is not very different from atheism. Philo, meanwhile, has caught his breath and wades back in. He attacks the idea of reasoning from earthly particulars to the universe as a whole, and pretty well lays out the modern skeptical opposition to the Argument from Design, asking why, if ideas in God's mind organize themselves without cause, why matter couldn't do the same thing. Cleanthes retorts that, regardless of Philo's "abstruse doubts, cavils and objections," the chain of inference from order to designer is too clear and straightforward to doubt.

Part V    Philo fires a broadside against Cleanthes. He points out that the then-new discoveries in astronomy and under the microscope undermine Cleanthes by making the universe less and less like human design. He points out that humans are in no better position to judge whether the universe is well designed than an illiterate peasant is to evaluate the Aeneid. He argues that we have no way of knowing whether the universe might not be the work of a team of deities, or perhaps a discarded trial run [positions that run counter to orthodox Christianity, of course, but not to the idea of intelligent design]. Cleanthes replies that not one of these arguments successfully refutes the fact of design itself.

Part VI    Philo presses on, urged on by Demea, who sees all these contradictory views of God as further proof of his own position. He argues that, if we are going to argue by analogy, then the universe is much more like an organism than a machine, with God as its soul, so that Cleanthes' design is tantamount to anthropomorphism (or perhaps pantheism). Cleanthes retorts that Philo's view suggests that the universe is infinitely old, a claim that he refutes. [Since this all happens before there is any real understanding of the age of the universe or the earth, the arguments are pretty much irrelevant to modern ears.]

Part VII    Philo continues, arguing that if the universe resembles an organism, it is much more likely to have originated by generation from matter itself. "Wait," says Demea, "I'm lost. Explain." Philo speculates that perhaps comets are seeds of solar systems, to which Demea asks what data there is to support such a view. Philo replies, "my point exactly. There are no data to allow any theories of cosmogony." [Again, this argument is hopelessly dated] Philo cites the examples of plants creating order from seeds, and rather cavalierly dismisses the possibility that the entire process might be proof of design. [One wonders how Philo would deal with DNA and the genetic code.] Cleanthes is not impressed and says "such whimsies as you have delivered, may puzzle, but can never convince us."

Part VIII    Philo argues that the appearance of order in nature could simply derive from the nature of matter itself. Cleanthes argues that the features of nature that are advantageous to humans proves the existence of a benevolent intelligence. [This is obviously a pre-Darwinian outlook, and atheists pounced with glee on natural selection as an answer to this argument.] Philo points out "thought has no influence upon matter except where that matter is so conjoined with it, as to have an equal reciprocal influence upon it. No animal can move immediately any thing but the members of its own body." to argue that pure intelligence could not manipulate matter.

Part IX    Demea argues that since arguments a posteriori are inconclusive, the only way to go is an a priori declaration of faith. Cleanthes points out that the usefulness of an approach doesn't make it valid, then, to anticipate Philo's arguments, attempts to show that one cannot demonstrate the existence of God a priori.

Nothing is demonstrable, unless the contrary implies a contradiction. Nothing, that is distinctly conceivable, implies a contradiction. Whatever we conceive as existent, we can also conceive as non-existent. There is no Being, therefore, whose non-existence implies a contradiction. Consequently there is no Being, whose existence is demonstrable.

Philo adds that some patterns in mathematics that appear like wonderful design can be shown to arise from the nature of numbers themselves; maybe the order in nature is similarly merely a result of the nature of matter.

Part X    Demea argues the instinctive nature of religious belief, the wretched state of humanity and the universal hope of a better world. Philo counters by asking why the world is so wretched if it is controlled by a benevolent intelligence. Demea argues that the pains of this life might be balanced by rewards somewhere else, a position that Cleanthes rejects as unprovable. He states "the only method of supporting divine benevolence is to deny absolutely the wickedness and misery of man." Philo shreds this argument, noting that pains are far more intense than pleasures and that the natural state of man offers no grounds whatever for deriving moral principles.

Part XI     Cleanthes explores the idea that perhaps evil is the price of achieving some greater good. Philo asks if the world is the sort of thing a rational, impartial observer would picture as the work of a benevolent Being. [C.S. Lewis makes very similar arguments in Mere Christianity with quite different ends in mind.] Philo ascribes the misery in the world to four causes:

Philo goes on to argue that we can frame four hypotheses about the first causes of the Universe: it was all good, it was all evil, it was both good and evil, or it is neither good nor evil. We can dismiss the first two outright since we see both good and evil in the Universe. The third is essentially Manichaeanism, and seems hard to reconcile with the uniformity of general laws of nature [a fairly dubious line of reasoning] so he opts for the fourth. Demea realizes to his horror where Philo is going with all this and leaves.

Part XII    Philo and Cleanthes conclude on a more cordial note, with Philo admitting that his hatred of superstition and corrupted religion sometimes carries him to extremes, while Cleanthes argues that even corrupted religion is better than none. At the end of the book, Pamphilus observes:

I cannot but think that Philo's principles are more probable than Demea's, but those of Cleanthes approach still nearer the truth.

That Puzzling Final Remark

It's clear from the flow of the book and other writings by Hume that this remark does not represent Hume's views, as many commentators of the 18th and 19th centuries liked to comfort themselves. The purpose of this particular comment has perplexed generations of scholars. Philo pretty much has Cleanthes on the ropes in the last few chapters of the book. If the book is a chronicle of Hume's own evolving views, it's clear that toward the end he has come to regard the Argument From Design as seriously flawed. If Hume always regarded the Argument from Design as flawed, then he first allows Demea free run at the beginning in order to dismiss the idea that we cannot reason about God, next he allows Cleanthes to take his best shot, then he allows Philo to dismantle the Argument From Design while Cleanthes offers progressively more feeble counter-arguments. 

Remember that even on its own terms the Dialogue is a work of fiction, in which Pamphilus spends some time defending the use of a dialogue format before telling the story. So why have Pamphilus toss in this comment? Is it a pro forma sop to religious believers? Or is Hume perhaps saying that the Argument from Design may have some merit but is not nearly as conclusive as its proponents tended to believe? Is he saying that the Argument from Design is a possible interpretation but not the only one consistent with what we observe?

In any event, simple intellectual honesty would demand that people who believe Hume dismantled the Argument From Design should at least acknowledge this statement and deal with its implications. And they rarely do.

The Case Against the Argument From Design


Philo emerges as the winner. Frequently Cleanthes and Demea are allowed to dismantle each others' arguments so that Philo is spared the onus. However, the fact that an author writes something doesn't necessarily make it true, and the fact that Philo wins the debate (as stage directed by Hume) doesn't necessarily prove his position is valid. The way many people cite Hume suggests strongly that they think Philo's winning the debate proves something, whereas it merely reflects the way Hume wanted the debate to come out. Whether Philo's arguments are actually valid depends on their own logical consistency, which Hume may or may not have analyzed rigorously.

Philo wins because Hume writes it that way. We might just as well say that Moby Dick proves whales are evil. Furthermore, it's an intellectually dishonest approach. Hume makes Demea a simpleton instead of having a foil on a par with, say, Thomas Aquinas or Augustine. (I guess we could level the same charge against Galileo for making Simplicio a caricature, as well.)

Science Then and Now

The nature of science has changed dramatically since Hume's day, and the changes illustrate some holes in Philo's reasoning that probably would not have been apparent to Hume or anyone else at the time.

A Sample With N = 1

When two species of objects have always been observed to be conjoined together, I can infer, by custom, the existence of one wherever I see the existence of the other. And this I call an argument from experience. But how this argument can have place, where the objects ... are single, individual, without parallel, or specific resemblance, may be difficult to explain..... To ascertain this reasoning, it were requisite that we had experience of the origin of worlds...

Here's another case where Hume utterly (and nobody can blame him) failed to foresee the growth of science. He would doubtless be astonished at how much we can infer about the origin of the universe.

But let's look at the validity of the argument even in Hume's day. The French mathematician Laplace would publish the first speculation on the origin of the solar system shortly after Hume's death. It simply is not true that we cannot reason about a unique event.

Hume's Central Circularity

The issue to be decided is whether the order in nature is the result of intelligent design. If it is, then the properties of matter (the color, luster and density of gold, for example) are also the result of intelligent design. Postulating a dichotomy between intelligent design and the properties of matter therefore amounts to postulating a priori that there is no design in nature. Hume (and all who follow him) essentially follow a grand circularity:

Nowhere is the circularity more blatant than Jacques Monod's Chance and Necessity. Monod starts by asking what criteria one would use for deciding something was intelligently designed and defines two criteria: repetition and geometric regularity. But, he hastens to add, these criteria apply only at the microscopic level. The distinction is entirely ad hoc because, if we applied it at the microscopic level, the Argument from Design would follow automatically. In any case, the distinction is nonsense, since macroscopic non-biological structures like crystals, cloud patterns and orbital resonances in the Solar System display repetition and geometric regularity, and microscopic structures like computer chips are of clearly intelligent origin (the bug in the first Pentium chip notwithstanding).

Lacuna Matata (Don't Sweat the Holes)

There's probably no greater lacuna in Hume's reasoning than in Part IX. Demea asks why there should be something rather than nothing, and why the universe we know instead of something else. By definition there can be no external cause, hence the only explanation is a logically necessary Being who "carries the Reason of his existence in himself, and who cannot be supposed not to exist without an express contradiction." (I run into people who quite literally cannot conceive of their religious beliefs being wrong, for whom the very concept of the Bible or the Koran being wrong is a logical contradiction.)

Hume puts the reply in the mouth of Cleanthes, who says it only to forestall Philo. Hume presumably assigns this role to Cleanthes to spare Philo the burden of attacking every religious doctrine and thereby alienating readers, but it's quite definite here that Cleanthes is voicing Hume's own convictions. He says:

Nothing is demonstrable, unless the contrary implies a contradiction. Nothing, that is distinctly conceivable, implies a contradiction. Whatever we conceive as existent, we can also conceive as non-existent. There is no Being, therefore, whose non-existence implies a contradiction. Consequently there is no Being, whose existence is demonstrable.

This is essentially disproof by outright denial. We can conceive of God as non-existent, hence God cannot be logically necessary. The only thing missing is any proof that the conception of God as non-existent is valid even in principle. Cleanthes shortly cites the idea that 2 + 2 = 5 as an example of a logical contradiction. By the logic above, if I say I can conceive of 2 + 2 = 5 not being a contradiction, that makes it non-contradictory, since I can conceive of the contradiction being non-existent. 

Cleanthes (Hume) goes on to say that the universe might well contain hidden attributes that would make its non-existence seem as contradictory as the non-existence of God.  Considering how hard Hume hammers elsewhere at the unproven nature of religious conjectures and their ad hoc nature, postulating wholly unknown properties for the universe is a nice case of the pot calling the kettle black.

But there's more. Cleanthes (Hume) says the only argument that persuades him the universe (rather than God) is not the necessarily existent entity is the fact that we can validly conceive of it being different. "But it seems a great partiality not to perceive, that the same argument extends equally to the Deity." Let this soak in. The entire thrust of Hume's arguments up till now, mostly as voiced by Philo, have been that we cannot reason from our knowledge of design on earth to design in the universe. Now Hume uses Cleanthes to assert that because we can conceive of matter being different than it is, we are justified in conceiving of God being different than he is. If there is something about God that makes him logically necessary and immutable, it must be attributes we do not know, and we have no way of knowing whether the same qualities might not reside in nature. We're back to proof by postulating unknown hypothetical properties of the universe.

Hume essentially gets away with using mutually contradictory arguments by putting them in the mouths of two different characters, but nothing could be clearer than that Cleanthes, in this section, is relating Hume's ideas exactly.

A Truly Weird Argument

In Part VII, Philo describes that in Indian mythology, the universe was spun by a great spider, and goes on to say:

And were there a planet wholly inhabited by spiders (which is very possible) this inference would there appear as natural and irrefragible as that which in our planet ascribes the origin of all things to design and intelligence, as explained by Cleanthes.

This argument is so wonderfully naive and blatantly anthropomorphic one hardly knows where to begin. A good starting point is to say duh..., any primitive intelligent species is likely to picture God as a larger version of itself. Spiders on this planet don't, as far as we know, have any conception of God so we don't need to worry about what they think, if indeed they have any more cognition than a personal computer.

So Philo's hypothetical spider planet would have to be inhabited by intelligent spiders. Just as human thinkers in Hume's day had long since concluded that it was mere imagination to picture God in human form, by the time Philo's spiders reached a similar level of development, the more intelligent ones should have gotten over their primitive arachnomorphism and reached the same conclusion. No doubt their artists would still picture God as a spider, but they would realize that was an image, not necessarily reality. (The only people on our planet who really seem to believe God is an old man with a long white beard are cartoonists and atheists.)

Spider webs, of course, are a matter of biology rather than conscious effort, but a race of spiders that remained at that level would be unlikely to have enough consciousness to speculate about God. Who knows what sorts of webs intelligent spiders might spin? (Just imagine the webs a spider mathematician might weave!) And even if we postulate a race of spiders that did nothing but spin webs and engage in philosophy while waiting for hapless flies, this argument still embodies the fundamental circularity of assuming that complexity in nature (spinnerets and spider webs in this case) are not the result of design, then using their "naturalness" to "prove" that the complexity is not the result of design.

The Problem With Disproofs Of Design

When I was attending geology field camp, a favorite pastime was hunting for arrowheads. There was no doubt about their authenticity. They were exquisitely chipped on both sides (in contrast to the clumsy modern ones sold in tourist traps) and made of obsidian, a rock not found in our field area. Most of my classmates did well to find one or two. I found dozens. In fact it got downright obscene. I found arrowheads in camp along trails that people had walked every day for weeks. The most surreal case was finding an arrowhead in my shoe one morning. I presume it was the gift of a pack rat because I guarantee none of my classmates would have given up a nice arrowhead for a practical joke!

I also found quite a few chips of red and yellow chert. They occurred in batches, and this rock, too, was foreign to the area, but I never found a full-fledged artifact of these materials. All things considered, I think it's safe to assume they were left by Indians.

At the opposite extreme of probability, the late anthropologist Louis Leakey spent a number of years excavating a site in the Calico Hills, California, convinced that he had discovered artifacts that pushed the arrival of humans in the Americas back to 100,000 years or more. The purported artifacts included flakes and crudely chipped pebbles. The problem is that the claimed artifacts come from an alluvial fan, not the gentlest depositional environment. With fist-sized pebbles being carried by flash floods, it would be surprising if there weren't a few that got chipped en route in a manner reminiscent of an artifact. Most geologists and anthropologists are convinced that's exactly what happened. (You'd never know this to look at most of the Web sites on the subject, which start at taking the artifacts at face value and drift off to the fringe from there. Very few pictures of the claimed artifacts are on line, something that hardly speaks well for their status as revolutionary discoveries or the confidence of their backers.)

So here we have chips, flakes, and chipped rocks. Are they artifacts of intelligent origin? In one case circumstantial evidence suggests they are, in the other case, not. But they're ambiguous. Maybe some single band of early humans got to Calico through some improbable series of adventures, lived for a while, then died out. We can never rule out the hypothesis short of a time machine.

Moral: If something looks complex enough to be of intelligent design, one possible interpretation is always that it is of intelligent design. It may not be, but in the absence of disconfirming evidence, intelligent design is always a viable hypothesis. We can say that it's not the only possible explanation, maybe even that it's not the most likely explanation, but it's extremely hard to dismiss the idea entirely. Intelligent design is always a possible interpretation of any sufficiently complex object.

Why Try to Disprove Design?

In his Enquiry, Hume accurately described the Argument From Design as "useless" because in and of itself it can never "establish any new principles of conduct and behavior." The Argument From Design only shows at best that there is intelligent design in the universe; it tells us nothing about whether the entity cares about human beings, communicates with them, or has moral scruples. Of itself, intelligent design does not validate any theology beyond deism.

On the other hand, intelligent design does not violate any known facts or logical principles. So why does it meet with such fierce opposition? True, many people leap immediately from the notion of intelligent design to the theology of their particular sect, but the proper response by anyone who claims intellectual rigor is to show the hidden assumptions in that leap of reasoning.

Still, it's legitimate to raise the possibility that order in the universe arises solely from the properties of the universe itself. Or is it? We know that some cases of complex order are the result of intelligent design. We do not know that any other origin for complex order is possible. What basis do we have even for postulating such a possibility?  The bottom line is that none of the criticisms of the Argument From Design are compelled by any empirical or logical evidence; they are inspired solely by the desire to discredit the Argument From Design for the sake of discrediting it.

The Red Herring

Parts X and XI are given over to a debate about suffering and its implications for a benevolent intelligence. Although any discussion about the nature of God has to confront this issue, it is completely premature in a debate about the existence of God. Most of Hume's readers, of course, would have thought in terms of either the God they pictured in their own theology, or no God at all (other conceptions of God being merely spurious fantasies). They would have considered Cleanthes' arguments for a benevolent intelligence inseparable from the issue of design in general. But that's no excuse for a modern reader to conflate the two issues. 

Where Should We Look For Design?

It's worth looking at a list of what was not known in Hume's day:

So, although science was making thrilling progress in understanding the workings of the world, the "laws of nature" in Hume's day were mostly restricted to empirical descriptions of phenomena, and therefore the argument for and against design was waged mostly in the realm of complex natural phenomena. For every case Cleanthes can cite of two natural phenomena meshing smoothly together, Philo can cite a counter-example where the feedback fails. A century later Darwin's theory of evolution by natural selection would pretty much demolish this line of reasoning.

However, a modern day Cleanthes (as opposed to his semi-literate wannabes) wouldn't point to the way a hummingbird's bill is precisely shaped to fit a certain flower (or vice versa). He would point instead to the symmetries found in physics and the way a relatively small set of fundamental laws give rise to such a vast range of complex phenomena (the Cleanthes wannabes are mostly too illiterate even to have any idea of physics). He would also point to the fact that a very slight change in some fundamental constants would either have prevented complex atoms from forming at all, or would have allowed the stars to exhaust their nuclear fuel before life could evolve.

By way of analogy, consider a chess board well into the game. Is this a product of intelligent design? The pieces might very well look randomly arrayed, although a skilled player could easily spot that some arrangements cannot be achieved in normal play. For example, one side cannot lack a king, or have a pawn in its own back row. There is a checkmate position involving a king and two knights against a lone king, but it cannot be achieved in normal play (the lone king can always evade checkmate). But there would be many positions that are ambiguous. For every masterful position Cleanthes cites as clear evidence of intelligence, Philo cites one with two passed pawns. Cleanthes suggests that some positions are artificially created as problems; Philo accuses him of creating ad hoc excuses to avoid acknowledging failures of the design hypothesis. Cleanthes argues the pieces are clearly of intelligent manufacture; Philo notes that seashells are even more intricate and are wholly natural. Cleanthes points to the geometric regularity of the board; Philo cites crystals and honeycombs as equally regular but natural structures.

The reason chess has been played as long and seriously as it has is not because people like the pieces and the board (a nice chess board and set has esthetic appeal because of our reverence for the game), but because a very sparse set of rules leads to a fantastic variety of possible outcomes. And it's in the rules, not in the pieces, the board, or any particular arrangement on the board, that we must look for intelligence. (Chess isn't a perfect analogy. There are ad hoc rules like capturing en passant. The Japanese game Go might be a better analogy for pure intelligent design, but chess is more familiar.)

Philo, of course, has not been asleep at the wheel. He would point out that the beauty of chess is rooted in the underlying laws of logic and mathematics in the universe. He would also point to hypotheses that perhaps there are many universes, and we merely occupy one where the laws of physics allow matter to evolve into complex forms.

Cleanthes would wonder, in turn, why it's valid to criticize theological concepts as ad hoc, while it's simultaneously permissible to postulate the existence of universes whose existence is entirely unproven and which may be forever untestable. He'd also wonder why the order in the universe is such a pressing scientific problem as to justify postulating a vast number of alternate universes but not to justify postulating an intelligent designer.

That Chess Position

This surprised me when I first found it, and I've never seen it in any chess book. X is the checkmated king. It's possible to achieve this position without violating the rules but impossible to force it. The two knights always leave holes the lone king can slip through. The losing player has to be very cooperative or very dumb to end up in this position.
|  |  |  |  |
|Kt|K |  |  |
|  |  |Kt|  |
| X|  |  |  |

Return to Pseudoscience Index
Return to Professor Dutch's Home Page

Created 11 March, 2002, Last Update 05 December, 2011

Not an official UW Green Bay site