The principle quoted above begins an early paragraph of the Groundwork (Kant 1997a, 8/4:395). It introduces an argument whose importance has too often been overlooked. But it also appears to inject an embarrassing natural teleology into Kantian ethics; and for this reason most readers today are prone to ignore it. What is the argument? What is so important about it? and Why does it seem embarrassing?

The teleological argument. Suppose the teleological principle above, about the adaptation of parts in organized beings, is correct. In that case, something about the role of practical reason seems to follow. Practical reason—the application of reason to conduct—would serve some end that could not be better served by desire or inclination. So considering that the latter would be far more effective in bringing about well-being and happiness, reason, in its practical use, cannot be expected to guide conduct aiming at those ends. It must serve some other end instead; or it has no practical function at all. The end it serves, Kant concludes, would be to establish a good will.

Importance of the argument. If sound, the teleological argument just presented undermines all moral theories that are today called “consequentialist.” Theories of this type assume that the ultimate end or purpose of moral action, and so of practical reasoning, is to bring about some good consequence in the world. But the teleological argument shows that if that is what morality is about, then nature would not have made us rational. Kant did not put the point in terms of a divine creator; but he could have, in the following way. The creator would not have made us rational if the morality of actions depends upon their good or bad consequences. He would know that, for us, sensible pleasures and pains would be more effective guides to action than rational insights about actions’ consequences. Lying, for instance, tends to produce more unhappiness overall than telling the truth. Knowing this, God, or “nature,” would not have aimed at preventing us from lying by giving us the rational ability to discover lying’s bad consequences. The most efficient arrangement would have been to instill in us a strong abhorrence of lying, so we would naturally avoid it, as, for example, we naturally avoid incest. An efficient God or system of nature would make the very thought of forbidden actions revolting; and no one would need to understand why they are forbidden. So it would be pointless for us to reason about how we ought morally to act, and pointless even for us to have that ability. We would need no concepts like “right” and “wrong,” or “ought”; we would need only to follow our “gut.” What could be simpler, and more natural?

The trouble with teleology. The main reason people tend to reject teleological arguments today is that biology and the life sciences have been so successful without them. Evolutionary thinking tells us that what may appear to be intelligent design imposed on nature is more parsimoniously explained in terms of thoughtless natural selection. The evidence of fossil records seems to show that species adapt to their changing environments over vast stretches of time. Kant’s assumption that nature or the deity has endowed human beings with practical reason, for some supernatural purpose, might have been respectable for the science of his day, but it is outdated now. Perhaps because of the trouble with teleology, Potter has called Kant’s argument here “unusual and not very convincing” (Potter 1998, 37). Guyer has criticized it on the grounds that it assumes that “every natural organ or capacity has one and only one natural end” (Guyer 2005, 194). Allison points out that if practical reason cannot serve happiness or well-being, it does not immediately follow that its purpose is, as Kant concludes, to develop good will (Allison 2011, 84). Ross objected that “Even if we regard teleology in nature as real, we are not bound to regard every arrangement existing in nature as as perfectly adapted to secure the end it is directed towards” (Ross 1954, 13).

Replies to Criticism. Kant did not claim that every natural “instrument” would be perfectly adapted to its end. He claimed only that it would be better adapted in comparison with any other instruments in the same organism. So his point, which Ross seems to miss, is that the natural end of happiness would be better served by instinct or inclination than by practical reason, because the latter would not be as effective as the former. In fact, and this is the general point I like most about Kant’s argument, instinct or inclination would better serve any end that could be brought about in the world through human action. So, assuming we are practically rational, as the consequentialists also assume, rationality must serve for some other purpose than bringing about good consequences. The only alternative, Kant thought, would be that rationality serves for goodness of will. Allison is wrong in his objection: that just because it would not serve for happiness does not mean that practical rationality would serve for good will. Goodness of will has just been shown to be independent of consequences. Practical reason has just been shown not to serve to bring about good consequences. Kant will later distinguish clearly between two types of “good”: well-being and moral good (Kant 1997b, 51/5:59). If practical reason cannot serve the one end, then, if it serves a purpose, it must serve the other. Allison evidently imagines a third alternative, but here there is none. Guyer, for his part, is attempting to deny a logical point. Kant is perfectly correct to assume “one means–one end.” This is required by the teleological method. If a natural “instrument” could have more than one end or proper effect—say it has two, for simplicity—then it would not work in conflict cases. It would not work when the way to achieve one proper effect prevents the other from being achieved. It could work only if one end were subordinated to the other. But such subordination would mean that bringing about the one effect is the proper end or function. Hence: one means-one end.

Skepticism about teleology. The teleological argument, as I see it, convincingly shows this: assuming that reason is capable of guiding human conduct, the purpose for which it would do so is not to bring about something in the world that could be brought about better by instinct or inclination. Thus, practical reason does not have a “consequentialist” orientation. An evolutionist might balk at the suggestion that reason, or anything else, has a “natural purpose.” But that would be overstepping her area of expertise. This is not a scientific but an ethical controversy. The biologist can say that nothing in her field shows evidence for any natural purposes. Kant would agree completely, and add that, nevertheless, this does not show that practical reason has no natural purpose.