The Argument from Design purports to demonstrate the existence of God by citing as evidence the appearance of design or purpose in the natural world. This argument differs from the others we have considered in that it does not depend on any exotic or potentially controversial metaphysical assumptions. The ontological argument assumes as a premise that existence is a perfection; the cosmological argument assumes that every positive fact must have an explanation. And whatever else one thinks of these arguments, their appeal is limited precisely because the rational atheist or agnostic seems to be entirely within his rights in resisting these assumptions. The Argument from Design, by contrast, appeals only to evident facts of experience and to a principle of reasoning that seems to be firmly embedded in ordinary common sense and in scientific thinking.

In class I will run through a detailed version of the argument due to William Paley (pictured here). In these notes I will consider only the following simplified version.

(1) It is an indisputable and yet remarkable fact that many natural objects appear to have been designed for a purpose: the eye for seeing, the hand for grasping, etc.

(2) The only reasonable explanation for this appearance of purpose is that natural things are ultimately the product of an immensely powerful supernatural intelligence, namely God.

(IBE) If an hypothesis H is the only reasonable explanation of a remarkable fact F, then it is reasonable to believe that H is true.

(3) Therefore, it is reasonable to believe is that God exists.

The argument is certainly valid. The only question is whether we find ourselves with reason enough to believe the premises.

The Argument from Design is an argument of a distinctive sort. It proceeds by means of a principle which is sometimes called the Inference to the Best Explanation or IBE. This principle can be formulated in several ways, and we will discuss it at length when we take up the rationality of scientific inference in the next section of the course. For present purposes it is enough to note first that the principle seems to be implicated quite centrally in ordinary reasoning. Whenever a detective infers from the presence of certain clues or traces that the Jones murdered Smith, we can think of him as proceeding as follows. He notes several remarkable facts: Smith is dead; a bloody poker was found near his body; Jones was holding the poker at the time; Jones hated Smith, etc. After assembling his data, the detective canvasses explanations or accounts of the data. It is consistent with the data that Jones murdered Smith; it is also consistent with the data that invisible trolls murdered Smith and then placed the poker in Jones's hands. The detective then asks which of these accounts does the best job of explaining why the evidence is as it is. The version of the principle we exploited above licenses a conclusion when a single explanation emerges as vastly more reasonable than any other. In that case, we are obliged, according to our principle, to infer that the best explanation is the true one.

It is possible to imagine stronger versions of the principle, e.g.:

(IBE+) If H is the best explanation of the observed facts, then it is rationally obligatory to believe that H is true, even if there are other, reasonably good explanations available.

We can also imagine a weaker version:

(IBE-) If H is the only reasonable explanation of the observed facts, it is rationally permissible (though not obligatory) to infer that H is true.

One central question in the philosophy of science is which (if any) of these versions of the principle best captures the procedures of science itself. We will postpone that discussion. But you should try to be alert to how these alternative versions of the principle might affect the argument.

The claim that natural objects exhibit the appearance of design can be understood in a number of ways.

• Some versions of the argument stress the structural complexity or intricacy of many natural things: crystals, cells, spider webs and the like are all remarkably complex, but they are also remarkably regular or orderly. This notion of orderly complexity is somewhat obscure, and I really have no idea how it might be measured. A skeptic about the notion might say, for example, that any collection of rubbish in a junkyard is a certain sense remarkably complex: it is exceedingly unlikely that any chance process would have produced just this arrangement of junk and not some other. So there is no sense in which, say, the human brain is more complex than any other spatially distributed collection of a similar multitude of parts. Anyone who is concerned to defend a version of the argument from design that begins from a premise about the remarkable structural complexity of nature must first say how exactly this notion is to be understood. If you are attracted to this sort of argument, you might pursue the matter in one of your papers. There is a useful introductory discussion of the problem in Chapter 1 of Richard Dawkins' The Blind Watchmaker.

• The most familiar versions of the argument stress lay more emphasis on the fact that many natural objects appear to serve a function. This is not true in general: most rocks do not appear to be for anything. The organs of animals and plants, on the other hand, all seem to function as tools the customary use of which serves the animal's good or interest. Eyes are for seeing; animals use them to glean information about their environments, and if they did not have them they would be much worse off. This is not to say that we can assign a function to every animal part, or that we can know in advance that every part must have a function. But it seems incontestable that a great many natural objects are functional in this sense.

• An obscure but intriguing version of the argument stresses neither intricacy nor function, but rather the fact that laws of nature are compatible with the existence of life or perhaps with the existence of intelligent life. The fact that there are living things (and some intelligent ones at that) shows beyond question that the fundamental laws, whatever they are, are compatible with the existence of life. This would not be remarkable if it turned out that just about any collection of laws was compatible with the emergence of life. But apparently this is not the case. I must stress that I can claim no expertise here. I am not in a position to understand much less evaluate the arguments that have been offered on this score. But it is sometimes asserted by theologically minded scientists that if the constants that figure in the fundamental laws of nature had been ever so slightly different then the emergence of intelligent life would have been impossible. The claim, in effect, is that if we consider the space of all possible universes, only a vanishingly small region will be habitable. And if this is true, then it is surely a remarkable fact that the our universe -- the real universe -- has the highly improbable feature of being compatible with the emergence of intelligent life. It is as if, one might say, the universe were made for living things to inhabit.

Here are some references to more detailed versions of this argument. I cannot vouch for any of the facts. You will have to decide for yourselves how seriously any of this deserves to be taken. The topic is a magnet for cranks of all sorts. This is not an area where one can easily defer to the expertise of someone who seems to have mastered some technical science.

The first premise involves two distinguishable claims: The first is that the alleged facts about the natural world are genuine facts -- the universe really is "highly ordered" (whatever that means); the fundamental laws really are such as to support the emergence of intelligent life, etc. The second is that these genuine fact demands explanation. It is important to make this distinction because not every fact cries out for explanation. Suppose it turns out upon inspection that I have exactly 3,009, 887 hairs on my head. There is a sense in which my having just that number of hairs is quite improbable. And yet no one would think that there had to be an answer to the question, "But why just that number?". What this example shows that in many cases there is nothing wrong with regarding an improbable fact as inexplicable. And this means that it will not be enough to establish that the existence of function or intelligent life or whatever is in some sense antecedently improbable. A proponent of the argument must show that the facts he sites are the sort of improbable fact that cry out for some account. And I confess that I do not know how, in general, to draw the distinction between facts that demand explanation and those which do not.

Here is an example that came up in my precept several years ago that may help to illustrate the distinction. Suppose I tell you that I have in my possession a device that generates dot matrix pictures. The matrix is a grid of squares, 1000 x 1000, and each square can be either be either black or white. The device determines the pattern of pixels, and your job is to see what you can infer about the device from the patterns it generates. Now suppose that the first time I run the device in your presence, it produces an apparently random jumble of black and white squares. There is a sense, of course, in which this particular pattern is highly unlikely. The probability of getting just this pattern -- whatever it may be -- from a device that randomly assigns colors to squares in the grid is 1/2^ (10^6), which is very small. But still, despite the improbability of getting this particular outcome, the fact that you got it has no tendency to show that my device is anything but a random generator of dots.

But suppose instead that on the first run of the device it produces a recognizable image of George Washington. This particular arrangement of dots is in a certain sense no less likely than the unremarkable arrangement described above. However this time, I take it, the result of the trial gives us excellent reason to believe that the device is not a random dot generator, but is rather guided by some sort of intelligence. This evidence is potentially fallible. If subsequent runs produced only a series of unintelligible jumbles, we might conclude that the first run had been a fluke. Still, it seems clear no reasonable person good just shrug off the appearance of the George Washington image by saying: so what? Improbable things happen all the time and most of them don't call for any sort of explanation.

This is just an illustration of the distinction between an improbable occurrence and a "remarkable", explanation-demanding occurrence. The Argument from Design presupposes that the natural facts it cites demand explanation of some sort. So one way to resist the argument is to grant that the facts really obtain but deny that they are all that remarkable. For example, it is often said that the existence of life in the universe is not all that remarkable. Given the age of the universe and the vast number of potentially habitable planets, it would be surprising -- so the story goes -- if life hadn't emerged, not just here but in a number of places. Now I don't know whether this is true; but if it is then it deprives the purveyor of the design argument of one potential starting point. If the same case could be made for each of the "remarkable" facts cited above, the argument would be neutralized before it could get started.

If we grant that the alleged facts are real and remarkable, we are under some pressure to entertain explanations for them. And given IBE as we have formulated it here, should one explanation emerge as clearly superior to its rivals, we are obliged to accept that explanation as true, at least until a better one comes along.

The theist offers a straightforward account: natural objects have the functions they do because they were created by an immensely powerful craftsmen who was concerned (in some measure) for the well-being of animals. He gave them eyes to see with and hands to grasp with because he knew that they would be better of with these things than without them. He created a universe whose laws were compatible with the existence of intelligent life because he wanted there to be intelligent life, and so on.

Can the atheist offer any alternative explanations? In some cases, she clearly can:

• Darwin's theory of evolution by natural selection provides an elegant account of how functional adaptation in the parts of living things can emerge over time through a process that involves no intelligence or foresight. (See the selection from Richard Dawkins in the course pack for just one illustration of the explanatory power of evolutionary biology.)

• Some of the ancient atomists seem to have held that the universe is infinite in both time and space, and that over time every possible arrangement of particles is realized. If this were the case then it would not be surprising that at some point in the history of the world there should be arrangements like the arrangements we find -- arrangements which present an illusory appearance of design.

• In a related vein, some speculative cosmologists have maintained that the physical universe we inhabit is one of infinitely many such universes. On this view, every possible universe -- i.e., every possible set of laws of nature; every possible configuration of particles in space -- is actually realized somewhere in a vast "pluriverse". If this were the case, then again, it would not be surprising that some universes support intelligent life and display the hallmarks of apparent design. Since a world of this sort is clearly possible, the "many worlds" hypothesis explains why such a world is actual.

These alternative explanations refute the extreme view that the theological hypothesis is the only conceivable explanation for apparent design in nature. But our version of the argument asserts only that theism is the only reasonable explanation. We have not said very much about what it is for an explanation to be reasonable. By any standard, the Darwinian account of the adaptation of organisms to their environments is a reasonable one. So it would be very difficult for a purveyor of the argument to restrict himself to these biological phenomena. It is just barely conceivable that he might show, after a detailed review of the available facts and their purported explanations within evolutionary biology that the theological explanation is more reasonable than the Darwinian one. But that is not an argument we can pursue in relative ignorance of the relevant science, which is where we now stand. From our standpoint it is enough to note that the Darwinian account of adaptation is clearly a contender, and hence that the Argument from Design cannot proceed without considering the explanatory power and promise of evolutionary biology in all its gory detail.

Darwin will do us no good, however, if we are impressed by the "facts" about the fundamental laws of nature sited above. A Darwinian explanation presupposes a domain of entities which exhibit heritable variation and random mutation. Living organisms fit this bill nicely; but the universe as a whole does not. We can give no sense to the claim that the fundamental laws of nature are what they are as the result of a process of natural selection. What would correspond to reproduction or the competition for resources? Universes, unlike animals, are not engaged in anything like a struggle for survival. So evolutionary theory cannot get a grip.

The ancient hypothesis of a universe that cycles through every conceivable arrangement of particles and the more modern "many worlds" cosmology do not suffer this defect. But it must be confessed that they are very strange theories -- as strange if not stranger than the theological hypothesis, which has at least the advantage of being familiar. I will leave it to you to assess the cogency of the argument from design when it proceeds from facts about the fundamental laws. Obviously we cannot even begin to assess the argument until we have a clearer sense of how remarkable the observed values of the constants really are. And even then, the fact that you and I cannot think of a non-theological explanation of some fact of this sort should probably count for very little. We are not trained to think about these things; so what really matters is what a competent physicist or cosmologist might say in response to the challenge to explain the relevant facts. It might be interesting if some of you looked into this matter.