ABSTRACT
Let us call a ‘theistic argument’ any argument intended to support the claim that God exists. In this essay I propose a strategy for rebutting theistic arguments based upon David Hume’s paradigmatic demolition of the argument from analogy. I show that Hume’s plan of attack can be generalized and made applicable to many different kinds of theistic arguments. To illustrate, I employ the Humean strategy to two theistic arguments: Richard Swinburne’s inductive version of the cosmological argument and William Lane Craig’s defense of the ‘ ne-tuning’ design argument. In Hume’s day, the leading theistic argument inferred an analogy between nature and human artifacts. According to this argument, we nd throughout the natural world an arrangement of means-to-ends relationships exactly resembling, though vastly exceeding in scale, accuracy, and intricacy, the objects produced by human contrivance. For instance, just as a watch is designed to tell time, so an eye is apparently designed to see. It ostensibly follows that nature is the product of an intelligent designer whose mind resembles the human mind, though of vastly greater power and creativity. Hume deployed three tactics in his refutation. First, he extended the analogy, that is, for the sake of argument, he conceded the analogy between nature and human design. However, he drew conclusions that were highly uncongenial to the proponents of the argument. For instance, just as we blame the architect if the building is poorly designed, so nature’s aws should lead us to infer incompetence in its designer. Hume contends that the pain, waste, and ugliness in the natural world points to an incompetent designer. Further, even a competent designer might not be eternal, in nite, all-powerful, all-knowing, or good, which are all essential properties of the God of theism. Hume’s second tactic was to question the assumed analogy at the heart of the design argument. Proponents of the design argument compared the universe to a great
machine. But the universe is not really much like a watch, or any other product of human craft. For instance, there is a difference of materials. We know that bits of glass and metal will not spontaneously come together to form a watch, but can we be sure that natural materials will not self-organize into complex and orderly arrangements? In fact, we know that they do. We frequently observe order, even means-to-end order, spontaneously arising in natural systems. The best-known process of natural self-organization is, of course, natural selection (a process unknown to Hume, though he comes tantalizingly close to stating it in some passages). Natural selection is an observable process whereby successive populations of organisms improve their meansto-ends adaptations to their environments. However, natural selection is only one of many observable processes of self-organization in nature (see Shanks 2004). We can generalize: the means-to-end organization of something like a watch is always imposed from outside by an extrinsic designing and manufacturing process. In nature, on the other hand, so far as we can observe, we see that such order always arises from spontaneous self-organization brought about by impersonal causes intrinsic to nature itself. A watch and an eye are therefore crucially different. Each is a highly ordered system exhibiting means-to-end organization. But insofar as we have observed the origins of order in natural things, we see that this order comes about in a radically different manner than with arti cial things. Prima facie, therefore, and despite super cial similarities of organization, orderly natural systems appear to originate in a very different way than manufactured items. At this point, defenders of design will likely admit that nature is full of self-organizing and self-maintaining systems, but they will insist that explaining this is precisely what requires a designer. Suppose we come to a clearing in a jungle and nd a fully automated and self-maintaining factory. Obviously, an intelligent designer set up this factory. By analogy, the universe with all of its self-organizing and self-maintaining systems – much like those our imagined factory, only immeasurably grander and more complex – must have been set up by a stupendously intelligent designer. Hume would reply that we cannot know a priori that order, even means-to-ends order, requires an intelligent designer. We can support the designer inference only when we encounter the particular kinds of orderly systems that experience has shown us are likely to result from design. Experience shows that factories are set up by intelligent designers and do not originate in any other way. However, a factory is just not suf ciently like the universe to justify a similar inference about origins. For one thing, an essential feature of a factory is that it has an identi able purpose: the manufacture of a given product. The universe has no apparent overall purpose. If, as some suggest, the aim of the universe is to produce intelligent life, it seems a remarkably inef cient, roundabout, and risky means of achieving that end. How, then, did order originate in the universe? Hume notes that the way that ideas originate and organize themselves in a mind is a process at least as mysterious as the way order arises in material systems. Therefore, it explains nothing to trace the order in matter back to a mind. We simply replace one particular mystery with another equally singular one. We save an unnecessary step if we just say that the physical universe originally contained intrinsic orderliness, which, through the operation of
lawful but impersonal physical processes and interactions, developed into the many orderly natural systems we currently see. If we must use an analogy, the universe developed like a mighty oak growing from an acorn. Hume’s nal tactic, therefore, is to propose a counter-analogy: it is more reasonable to compare the origin of order in the universe to the generation and development of an organism, which occurs automatically and is guided by the organism’s innate organization, than to compare it to the imposition of extrinsic order on an artifact. This was Hume’s strategy, to extend the analogy, question the analogy, and propose a counter-analogy. (The account of Hume’s critique of the argument from analogy as employing these three tactics is given in Moore 1986.) As a response to the design argument of his day, these arguments were very cogent. However, Hume’s riposte did more than rebut a particular argument; it made a permanent contribution to the art of philosophical polemics: Hume gave us a generalizable strategy for refuting theistic arguments. We may generalize Hume’s strategy as follows:
1 Concede the argument. And then show that it falls far short of establishing the existence of the theistic God.
