Part I of Kant’s Groundwork of the Metaphysics of Morals presents a derivation of the formula of universal law. This is the principle that “I ought never to act except in such a way that I could also will that my maxim should become a universal law.”Footnote 1 But several well-known commentators have criticized the argument by which Kant derived this formula, claiming to find one or another “gap” in his reasoning. My objective here will be to defend that argument against these charges. I’ll attempt to show why the formula of universal law actually does follow as a valid conclusion from the premises and presuppositions of Kant’s argument in Groundwork I. In doing so, I’ll also call attention to some neglected ideas that can be found in the argument and its wider context. Some of these, I think, can help us better understand the Kantian project of grounding morality in practical reason.
I should point out at the beginning that I am not the first to attempt a defense of the argument of Kant’s derivation. Several others, who agree that there are gaps in his reasoning, have already suggested different ways to fill these gaps.Footnote 2 They have pointed to ideas or principles that Kant would have endorsed on independent grounds, offering them as plausible, stop-gap premises for his argument — ideas or principles such as “humanity as an end,” “transcendental freedom,” or “ought implies can.” But my strategy here will be different, mainly because I am not persuaded that these offered suggestions are helpful.Footnote 3 Here I shall attempt to show why there really are no gaps in the argument of Kant’s derivation. I think his argument can be defended against critics’ charges just by clarifying its existing premises. Four points I’ll be emphasizing in what immediately follows are these: (1) that the derivation presupposes a special merit in acting purely for morality’s sake; (2) that Kant attempted to derive a formula expressing the contents of conceptions like “right” or “moral” that would guide someone in so acting; (3) that a rational will is required for actions guided by such conceptions; and (4) that reason’s native interest in systematic unity is the key to discovering their contents.
In section I, below, I elaborate these four points more fully. In the section after that I lay out Kant’s argument step by step, as it appears in Groundwork I. There I also identify two places where critics claim to find gaps in his reasoning. Then I show, in sections III and IV, how the points I have indicated just above, especially points 3 and 4, answer the critics’ charges. Next, in a fifth section, I turn to address a set of possible objections to my defense of the argument of Kant’s derivation, before concluding. As I discuss that derivation below, laying out the premises and presuppositions that make the argument work, I’ll have little concern for whether it may be sound. That is, my aim will not be to demonstrate why it is true that we ought to act only on maxims that we can will to become universal laws. Instead, I shall concentrate on showing why the premises and presuppositions of Kant’s argument, assuming we accept them, should lead us to this conclusion.
1. Four Preliminary Points
1. The derivation’s leading presupposition is that acting with “moral worth” requires acting purely for morality’s sake, guided by a conception of one’s action as “right” or “moral.” But many are skeptical about this requirement, because they believe that actions can be morally admirable even without being guided by such ideas. And some go so far as to say that, depending on the circumstances, being moved to action by a conception of one’s action as right or moral can actually be unadmirable. Nevertheless, the critics of Kant’s derivation have not faulted his argument for its dependence upon this controversial presupposition. So for present purposes we need not enter into the otherwise interesting debate over the requirements for actions’ moral worth.
2. Assuming, then, that there would be some special merit attaching to a person’s acting purely for the sake of what is right or moral, a clever strategy of argument emerges. Conceptions like “right” or “moral” have to matter to a person if they are supposed to be capable of guiding her in acting. Hence, if we can discover how or why these ideas would matter, it should be possible to understand their contents — it should be possible to determine what makes actions right or moral. This is the underlying strategy of Kant’s derivation of the formula of universal law, as it appears in Groundwork I.
3. In the early paragraphs of the Groundwork Kant provided a compelling argument to show that reason alone would determine us to action guided by conceptions like “right” or “moral.” But today this argument is often dismissed out of hand, or simply ignored.Footnote 4 It is the argument that makes a point of reason’s ineptitude as a guide to bringing about well-being or happiness. And it is so little heeded nowadays because it depends upon a teleological premise: that each of the core faculties of a natural creature would be best adapted by nature for achieving a specific purpose — best adapted for its purpose, that is, compared with any of the creature’s other faculties.Footnote 5
The argument is familiar enough that it need not be quoted here. It purports to show that reason’s function, in its practical use, is not to guide us in bringing about good consequences or happiness. This is because natural instincts and inclinations would be better suited than reason for this purpose. So, the contents of concepts like “right” or “moral” would not be specifiable in terms of any objects that instinct or inclination might bring about more effectively than reason.
4. The last of these preliminary points focuses on what I have called the “native interest” of reason.Footnote 6 I have suggested that reason’s native interest can inform us about the contents of ideas like “right” or “moral,” as action-guiding conceptions. It is not difficult to understand that reason would have an interest in consistency. And since self-defeating actions are practically inconsistent, it is easy to see how it would be in reason’s interest to avoid them. Thus, it may be tempting to offer an analysis of the contents of conceptions like “right” or “moral” in terms of, say, the consistency or non-self-defeatingness of actions with regard to their ends.Footnote 7 But as Kant saw it, reason’s interest is actually much deeper. The drive to consistency is only the tip of the iceberg. Consistency matters to reason because it is a requirement of systematicity, or completeness.
In the Critique of Pure Reason, Kant had explained that reason in its speculative employment seeks “the unconditioned for every condition.”Footnote 8 Given any effect, that is, reason seeks out its cause, and the causality of the cause, and even the entire causal series. Its quest for the unconditioned ultimately directs reason toward the totality of the world’s causal relations, as a natural system. Consistency is important for speculative reason, therefore, because inconsistency would block its path to complete comprehension of the system of nature.Footnote 9 Practical reason, on the other hand, would be interested not in comprehending the world, but in altering it. Or, better, we can say that practical reason’s aim is to bring about an entirely different world: a system not of natural causes and their effects, but of free actions and their ends.Footnote 10 And inconsistency between actions and their ends would be an obstacle to practical reason’s realization of that system — just as inconsistency in the laws of causes and effects would be an obstacle to speculative reason’s interest in comprehending the natural world. Accordingly, the contents of conceptions like “right” or “moral,” considered as guides to action for a rational will, should be analyzable in terms of reason’s interest in the systematic order of actions and ends; but more on this later.
The four points preceding are important background considerations to bear in mind as we now turn to examine the argument of Kant’s derivation.
2. The Derivation and Its Alleged Gaps
Included among the preliminaries to Kant’s derivation of the formula of universal law, in Groundwork I, is a problematic argument consisting of three propositions. Especially peculiar about this argument is that Kant seems never to state its leading premise. At one point in the text he introduces a statement as the second proposition, but previous to this point he does not identify any first proposition.Footnote 11 A statement he subsequently offers as the third proposition is said to follow from the two preceding, but it has been difficult to see how.Footnote 12
At least the second proposition is clear; and it seems well supported by Kant’s otherwise controversial analysis of moral worth and acting from duty. Here is his statement of that proposition:
The second proposition is this: an action from duty has its moral worth not in the purpose to be attained by it but in the maxim in accordance with which it is decided upon, and therefore does not depend upon the realization of the object of the action but merely upon the principle of volition in accordance with which the action is done without regard for any object of the faculty of desire. (13/4: 399–400)
This proposition has two parts, one negative and one positive. The negative part tells us that an action’s moral worth does not depend upon the value of any purpose or consequence to be achieved by it. The positive part tells us that moral worth depends upon the maxim, the subjective principle of the agent, and thus upon the conception guiding her in acting.
Kant’s derivation of the formula of universal law begins with this second proposition, or with what appears to be a paraphrase of its negative part. The first premise of the derivation is,
I. “[T]he moral worth of an action does not lie in the effect expected from it and so too does not lie in any principle of action that needs to borrow its motive from this expected effect” (14/4: 401).
In the next step of the derivation, which is the subsequent sentence in the text, Kant alludes to his earlier, teleological principle about the faculties. This step is offered in support of the previous premise, as a kind of reductio. So here I provide the following paraphrase of the second step, in order to make this explicit:
II. Suppose the moral worth of actions does depend upon the value of their expected effects (denial of premise I); since “these effects … could have been also brought about by other causes” — like natural instinct or inclination — “there would have been no need, for this, of the will of a rational being.”Footnote 13
To assume that by reasoning we can discern what ought to be done, and guide ourselves in doing it, is to assume that we have a rational will. This assumption was shared by Kant and his consequentialist opponents, whose views were his target in this passage. So his sub-argument here was that if moral action is expected to bring about consequences that could be brought about more effectively by “other causes,” like instinct or inclination, then for this — i.e., for moral action — we would not need a rational will. And if we would not need a rational will, then nature would not have allowed reason to “break forth into practical use,” as Kant had written earlier in his argument based in teleology. Hence, because we have a rational will, it must follow that the moral worth of an action cannot be derived from any valued effects it could be expected to bring about. And this is just what was indicated in premise I.
So with that premise supported in this way by step II in the argument, Kant next concludes that (paraphrasing again):
III. Therefore, the moral worth of an action must depend not upon some hoped-for effect of the action, nor upon any principle the motive for which would have to be borrowed from the allure of that effect, but instead upon the agent’s own “representation of the law in itself, which can of course occur only in a rational being.”
Here, then, Kant makes a transition from the negative part of the preliminary, “second proposition” to its positive part.
And now comes a rhetorical question: “But what kind of law can that be, the representation of which must determine the will [in moral action], even without regard for the effect expected from it? …” (14/4: 402). Kant’s single-sentence answer to this question presents the final two steps of his derivation, which I quote:
IV. “Since I have deprived the will of every impulse that could arise for it from obeying some law, nothing is left but the conformity of actions as such with universal law, which alone is to serve the will as its principle, that is …
V. “I ought never to act except in such a way that I could also will that my maxim should become a universal law.”
Thus concludes the argument of Kant’s derivation of the formula of universal law.
But in the past several decades, commentators have claimed to find some “gaps” in its logic. According to Paul Guyer’s analysis, for example, there is a gap in the argument that would fall just before step III, as it is represented above. This is where Kant concludes that the moral worth of an action must depend upon a “representation of the law” guiding the agent, and not upon the value of any object the agent is determined to bring about. Guyer grants that in his argument Kant succeeds in ruling out inclinations as possible motives for morally worthy actions, and that this would rule out any object of inclination as a possible basis for morality. So he agrees with Kant that morality’s fundamental principle cannot be formulated as a command to act in ways that would bring about any object of inclination. But he charges that Kant has nevertheless not shown that morality’s principle cannot direct us to bring about any object at all. And this is what he would need to show in order for his derivation to be valid. So Guyer thinks it does not follow, as Kant puts it in step III, that only a “representation of the law” can guide us in morally worthy action. And thus it does not follow, ultimately, that only the formula of universal law can serve as morality’s principle. Guyer puts his charge this way:
[E]ven if it is obvious that no object of merely contingent inclination could serve as the basis of morality, there still might be some sort of necessary object, perhaps of pure reason rather than inclination, which is the basis of the moral law, and if so then the fundamental principle of morality could be the substantive requirement to act only on maxims that would bring about that necessary object rather than just the formal requirement to act only on maxims that should also be universal laws.Footnote 14
So Guyer’s objection is that Kant leaves a gap in his argument because he does not rule out something he needs to rule out in order to draw his conclusion. In his argument he assumes that in actions with moral worth, rational agents would be guided by either a thought of an object of inclination, or a representation of the moral law. And his argument may well rule out the first alternative. But, says Guyer, this does not mean that agents must be guided in morally worthy action by nothing other than a representation of the law, or only by some conception of an action as “right” or “moral.” The guiding conception appropriate for such action could instead be some necessary object of reason. We shall see later what Guyer means by this.
A similar complaint has been lodged by Samuel Kerstein, who charges that the argument of Kant’s derivation is too sweeping. Like Guyer, Kerstein agrees that the first premises disqualify any inclination-based formula from serving as morality’s supreme principle. But he wonders why the formula of universal law would be the only principle left standing after all inclination-based formulae are ruled out. Kerstein poses this critical question: “Might [a theological moralist] not instead maintain that an agent’s notion that God wills that she do something itself can give her sufficient motive to do it?” (143 [see n. 3]). Kerstein sees no basis for assuming that action guided by the thought of obedience to God’s will must depend upon inclination. For it could be that this thought is sufficient to motivate moral action, apart from any inclination. So, like Guyer, he does not see Kant’s argument as showing, necessarily, that only a representation of the law can guide agents in morally worthy action. He therefore does not see the formula of universal law as the only formula that could follow from the premises of Kant’s argument.
The preceding criticisms focus on Kant’s transition from steps I and II to step III. But there is a second line of critical complaint about the argument of the derivation, one that alleges a gap in Kant’s reasoning from step IV to step V. Bruce Aune seems to accept the argument as valid up until step IV, where Kant concludes that “nothing is left but the conformity of actions as such with universal law, which alone is to serve the will as its principle.” But he does not see how Kant is entitled, from there, to derive the formula of universal law in step V. An action’s conformity with a law that may be valid for all rational agents is one thing, says Aune. Its conformity with the formula of universal law is something else. So Kant cannot infer the latter idea directly from the former, as if these two statements are logically equivalent.Footnote 15 Another way to put the charge is this: If all we know is that morality requires conformity to laws valid for all moral agents, as seems to be indicated in step IV, this does not show us which actions are required or ruled out by morality. But the formula of universal law is supposed to show us this. Or so Kant assumes. So there is evidently more in his concluding formula than follows logically from his premises.
3. Denying the First Alleged Gap
I want to show now that the premises of Kant’s derivation are actually more powerful than critics have recognized, and that there are no gaps in the argument. In this section I’ll emphasize the neglected teleological principle that Kant alluded to in step II of the derivation. In the following section I’ll attempt to develop more fully his thoughts on “conformity to universal law,” as expressed in the problematic step IV.
We may begin with Kerstein’s criticism. It was, roughly, that Kant’s argument did not preclude the prospect that an action guided neither by an object of inclination, nor by a representation of the law, might still have moral worth. An agent might be guided solely by her conviction that she is doing what God commands, and that might make her action morally admirable. Or at least someone might plausibly hold this view; and Kant’s argument does not seem to rule it out. Yet in offering this critique Kerstein has evidently overlooked the teleological allusion in step II. To Kant’s mind, our ability to reason with moral conceptions, and act guided by such reasoning, is evidence that we act with moral worth only when we are determined to action by reason alone. If, instead, we are expected to act morally guided by a conception of our action as commanded by God, then God would have designed human nature poorly in giving us a rational will. His having given us an instinct for acting as he willed would have been a design more worthy of his intelligence. It would have served his purpose less well to expect us to act guided by a conception of obedience to his command — and even less well to expect us to act by means of a rational will.
Step II of Kant’s argument, therefore, once its allusion to the teleological principle is made evident, actually does preclude the prospect that Kerstein claims is not precluded. The same teleological consideration that rules out morally worthy actions’ being motivated by instinct or inclination — because then nature wouldn’t have allowed reason to be practical — also rules out their being guided by a conception of obedience — because then God wouldn’t have allowed reason to be practical.Footnote 16
Guyer’s charge of a gap in the argument resembles Kerstein’s, insofar as he emphasizes an alternative conclusion to Kant’s argument — a conclusion the earlier premises of the derivation appear not to have ruled out. As Guyer sees it, Kant’s premises rule out only that the guiding conception for a morally worthy action is a thought of an object of contingent inclination. His charge is that the argument has not precluded the prospect that this guiding conception is the thought of an object of reason — something that we, as rational agents, would have a necessary interest in bringing about — something like “humanity.”Footnote 17 And Guyer’s point here is compelling; because as Kant would later postulate in Groundwork II, humanity is a necessary object of reason. It is what we are expected to treat always as an end in itself.
Yet this line of criticism will fail if “humanity” is considered to be an object that morally worthy actions are expected to bring about — something like human life, or human flourishing. And Guyer himself thinks of the idea of “humanity” in this way, as the text of his objection shows. His criticism will fail when so presented because of the teleological principle alluded to in step II. For suppose it is true that morally worthy actions are expected to bring about an object like “humanity.” In that case, an instinct or inclination would be better suited than reason for this purpose — something like empathy, or sympathetic feeling, or perhaps what Hume sometimes called the “sentiment of humanity.”Footnote 18 And then nature would not have allowed reason to be practical.
As Kant saw it, however, humanity is not an end that morally worthy actions can be expected to bring about. It is instead an “independently existing end,” and thus a “limiting condition” on our actions, as he put it later in the Groundwork. It is “that which must never be acted against and which must therefore in every volition be estimated never merely as a means but always at the same time as an end” (45/4: 437). Only reason could guide us in limiting conduct in accordance with Kant’s idea of humanity as an end. This is because having an interest in limiting our conduct in this way requires understanding the logic of the means-end relation. Anyone acting on a principle respecting humanity as an end would recognize that her acting otherwise would involve the inconsistency of treating an end in itself as a means to something else. Positively stated, therefore, a morally worthy action treating humanity as an end — in the sense of a limiting condition — would be determined by reason’s interest in the unity of the system of means to ends. And this is an interest that no natural instinct or inclination could support.
Guyer’s charge is that there is a gap in Kant’s argument because it does not preclude the prospect that “the fundamental principle of morality could be the substantive requirement to act only on maxims that would bring about [humanity] rather than just the formal requirement to act only on maxims that should also be universal laws.” The reply to this charge is that the premises of Kant’s argument do indeed preclude that prospect, if “humanity” is seen as an end that morally good actions are expected to bring about. For then we could have an instinct or inclination that would be better suited for this than reason; and nature would have given us this instinct or inclination instead of making reason practical. On the other hand, if we see “humanity” as only a limiting condition on human conduct, then Guyer has presented a much weaker basis for criticism. His charge reduces to the point that Kant’s derivation of the formula of universal law does not rule out the prospect that the fundamental principle of morality might be the substantive requirement to act only in ways that treat humanity as an end in itself. But this, of course, is roughly the so-called “formula of humanity,” which Kant subsequently presents as equivalent to the formula of universal law. And we cannot say that there is a gap in the logic of his derivation if it fails to rule out a conclusion logically equivalent to the conclusion he draws.
Not everyone agrees that the formula of universal law and the formula of humanity are equivalent, however. We can return to this point later, near the end, in order to offer some observations on their relation.
4. Denying the Second Alleged Gap
The charges of a gap in Kant’s reasoning around step III, leveled by Kerstein and Guyer, have been answered for the most part. It remains now to see how step III in the argument is supposed to support step IV, and how that step, in turn, would support the formula of universal law as expressed in step V.
Step III tells us that morally worthy action depends upon a “representation of the law in itself, which can of course occur only in a rational being.” From here Kant draws the conclusion, in step IV, that “Since I have deprived the will of every impulse that could arise for it from obeying some law, nothing is left but the conformity of actions as such with universal law. …” Commentators and critics have often taken this to mean that once contingent and particularizing inclinations are ruled out as a basis for the supreme principle of morality — by the conclusion in step III — nothing remains capable of expressing the contents of that principle but laws that would be valid for all rational agents. But how, then, would the formula of universal law follow from the idea of actions’ conformity to universally valid laws? That is the question posed by Aune in his charge of a gap in Kant’s derivation. Our knowing that in order to be right or moral our actions are expected to conform to laws valid for all does not imply our knowing the contents of any such laws — still less would it imply our knowing that we ought to act only on maxims we can will as universal laws.
But I think the charge of a gap in Kant’s reasoning here can be traced to a misunderstanding of what he means by actions’ “conformity to universal law.” I think he means by this not conformity to particular laws valid for all rational agents, but conformity to, or coherence with, a unified system of principles of conduct. In other words, I think Kant’s phrase “conformity to universal law” ought to be interpreted more generally than critics often interpret it. To show this here I’ll draw an analogy, which Kant himself draws also, between actions’ conformity to law and the conformity to law of objects in nature. This analogy will show how Kant thinks of “conformity to universal law” as the connection or coherence of the parts of a systematic whole. And once this point is clarified, I think it will be easier to see how the formula of universal law is supposed to follow from the idea of actions’ conformity to universal law.
We should first note that in Kant’s statement of what I am calling step IV of the derivation he included the seemingly inconsequential qualifier “as such” — “nothing is left but the conformity of actions as such with universal law.” The role of that qualifier in Kant’s sentence is ambiguous. In the translation I have quoted it is taken as qualifying “actions.”Footnote 19 But in a number of other translations it is presented as qualifying “universal law” — “nothing is left but the conformity of actions to universal law as such.”Footnote 20 I think the position of the qualifier makes a subtle difference. For on a plausible reading of what Kant might mean by “conformity of actions to universal law as such,” he is not thinking of actions as conforming to whatever particular laws may be valid for all moral agents, as Aune and others have supposed in their reading of step IV. He is instead thinking of actions’ conformity to law in some more general sense.Footnote 21 The following considerations, based on Kant’s theoretical philosophy, seem to elucidate his derivation of the formula of universal law from the concept of actions’ conformity to law as such.
Two years prior to publication of the Groundwork, in his Prolegomena to Any Future Metaphysics, Kant had offered a pair of definitions of “nature,” distinguished by matter and form. “Nature considered materialiter is the sum total of all objects of experience,” he wrote; and, “The formal in nature … is therefore the conformity to law of all objects of experience.” These definitions are not intended to refer to things in themselves, of course. They refer to, as Kant wrote, “things as objects of a possible experience, and the sum total of such objects is properly what we here call nature.”Footnote 22
In this text Kant poses for himself the challenge of answering the following question, expressed in two ways he considers to be equivalent: “How is it possible to cognize a priori the necessary conformity to law of things as objects of experience, or: How is it possible in general to cognize a priori the necessary conformity to law of experience itself with regard to all of its objects?” (91/4: 296). Here Kant is not asking how it is possible to know the particular laws of nature to which objects of experience conform. He is asking about something presupposed by objects’ conformity to particular laws. He is asking about nature’s lawfulness as such. “There are many laws of nature that we can know only through experience,” he writes, “but lawfulness in the connections of appearances, i.e., nature in general [or in the formal sense] we cannot come to know through any experience, because experience itself has need of such laws, which lie a priori at the basis of its possibility” (111/4: 318–9). Note how the distinction employed here, between objects’ conformity to laws known through experience, and their conformity to the lawfulness of nature known a priori, parallels the distinction we have just seen in Kant’s practical philosophy — between actions’ conformity to particular laws valid for all rational agents, and their conformity to universal law as such.
As a summary answer to his two questions about our a priori knowledge of nature’s conformity to law, Kant explains that “we are not acquainted with nature except as the sum total of appearances, i.e., of the representations in us, and so we cannot get the laws of their connection from anywhere else except the principles of their connection in us, i.e., from the conditions of necessary unification in one consciousness. …” (111/4: 319). In other words, his answer is that natural objects’ conformity to universal law as such would be their coherence with the systematic connection of all natural objects — a connection grounded in the unity of perceptual consciousness. By analogy, therefore, in practical philosophy we can understand an action’s conformity to universal law as such as its coherence with the systematic connection of all actions, based on the unity of practical consciousness, or of rational willing.Footnote 23
To flesh out this analogy further, we can instructively compare Kant’s thoughts on human actions with what he calls “empirical judgments.” He writes that empirical judgments would be either subjectively valid judgments of perception, or objectively valid judgments of experience:
All of our [empirical] judgments are at first mere judgments of perception, they hold only for us, i.e., for our subject, and only afterwards do we give them a new relation, namely to an object, and intend that the judgment should also be valid at all times for us and for everyone else; for if a judgment agrees with an object, then all judgments of the same object also agree with one another, and hence the objective validity of a judgment of experience signifies nothing other than its necessary universal validity. (91/4: 298)
The relevant idea here is that in judging “experience” we are committed to our judgment’s cohering with all other judgments about its object — our own and everyone else’s. We are committed, that is, to our judgment’s conformity to nature’s lawfulness as such. When we judge this way we aver something expected to be implied by a particular law of nature. Yet if that supposed law would not cohere with universal law as such — with nature in the formal sense — then our judgment must be regarded as false. So from the requirement of the conformity of natural objects to universal law it would follow as a rule for judging experience that, You ought to judge only in such a way that the law implying your judgment can be a universal law of nature — i.e., can conform to the systematic unity of nature.
Actions of the rational will can be instructively compared to empirical judgments. Choosing to act is like making a judgment, where the maxim of the action corresponds to the content of a judgment. If we see the maxims on which we act as having nothing more than private, subjective validity, then they are like judgments of perception. They are not in that case held to any standard of evaluation — “Why would do you do that?” “That’s just the way I am.” But in order to give our maxims objective validity we must think of them on the model of judgments of experience, and subject them to the standard of coherence with all maxims. Our maxims will satisfy this standard if and only if they can become universal laws in the system of practical principles. But if they cannot, they are like false judgments of experience. Hence, it can follow as an objective law for actions that, You ought to act only in such a way that the maxim of your action can become a universal law (of nature).Footnote 24
So we can see more clearly how in the last step of the argument of his derivation Kant moved without any gap in his reasoning from the idea that actions must conform to universal law as such to the formula of universal law. For just as, in theoretical philosophy, empirical judgments conform to universal law when the particular natural laws that would imply them cohere with the systematic unity of nature, so also, in practical philosophy, actions conform to universal law when their maxims can be willed to become universal laws.
5. Four Possible Objections
To the preceding defense of Kant’s derivation it is possible to anticipate at least four objections, which pose problems of unequal difficulty. First, some may be critical of our reliance here on the teleological principle about nature’s allowing reason a “practical use.” Second, a related worry may be that in emphasizing teleology in this way we commit Kant to a surreptitious leap from “is” to “ought.” Third, since some critics claiming to find gaps in Kant’s derivation have focused on versions of his argument found in Groundwork II, and in the second Critique, it might seem to them that we have here evaded difficulties with Kant’s argument just by shifting the focus of the debate. Fourth, and last, a concern capable of developing into a line of objection was mentioned briefly above, at the end of section III. It was that something seems wrong with Kant’s derivation if its premises could also support a conclusion different from the one he actually draws — a conclusion presenting morality’s principle in terms of “treating humanity as an end.” This should be a pressing concern if, as some now presume, the formula of universal law and the formula of humanity are not equivalent. In what follows we can address these anticipated objections in turn.
First, readers of Kant’s Groundwork are sometimes surprised to find there, in the early paragraphs of the first section, an argument based on teleology. Had not Hume, who awakened Kant from dogmatism, thoroughly discredited teleological explanation? And had not the first Critique firmly established Kant’s doctrine that human knowledge is limited always by possible experience, so that speculation about ends or purposes in nature must take us beyond what can be known? How, then, can Kant seriously offer an argument presupposing purposes and intelligent design among the faculties? And how can we, in our post-Darwinian age, be expected to take such an argument seriously?
The teleological premise presupposed in Kant’s derivation of the formula of universal law could well be false. But this would not affect the validity of that derivation, which is all I have been trying to establish here. Still, some might suggest that Kant must have recognized the weakness of his initial remarks about natural purposiveness in the mental faculties. So it may be objected here that from his initial and probably “throw-away” argument presuming nature’s intelligent design of human psychology, he would not have borrowed any dubious assumptions about practical reason’s purpose vis-à-vis inclination. Hence, he would not have argued as I have represented the argument of his derivation above.
I think it will suffice to reply to this line of objection by briefly characterizing Kant’s attitude toward teleological thinking. The fact is that he had not been persuaded by any Humean arguments to reject teleological explanation. Nor would even the later biology of natural selection have altered the core of his views on natural purposes. Evidence for these claims can be found not only in Kant’s “Critique of Teleological Judgment,” but also in the Appendix to the “Transcendental Dialectic” in the first Critique. For in these places Kant demonstrates his conviction that teleological thinking is a perfectly legitimate and useful scientific methodology. Only, it is crucial to bear in mind that in his view, teleological principles are not constitutive but regulative. As regulative principles they present nature — including human nature — as if adapted by design to a unified system of purposes. And, Kant thought, there is no reason to see this “as-if” purposiveness as inconsistent with the blind mechanism of natural causation. He even thought that assumptions about the ordered system of natural purposes are useful, as far as they may go, for scientific progress in uncovering nature’s mechanical causes. So it is a mistake to think that in deriving the formula of universal law Kant would intentionally have avoided considering human nature teleologically. For he would see nothing faulty in teleological thinking about human psychology as regulative, whenever this would extend our insights into that other system of regulative principles, morality.
Thus, in summary, we can offer the following response to the first possible line of objection. It should suffice to say that though the argument of Kant’s derivation may make implicit use of a teleological presupposition, this is no ground for skepticism about its soundness, much less its validity. Moreover, if we are prepared to acknowledge that human actions taking place in a natural world of mechanical causes can have moral worth — another presupposition of Kant’s derivation — we should not hesitate to think of nature as if purposively ordered. This is especially so when human psychology modelled as purposely ordered can help us understand the conditions for morally worthy actions, including the contents of the action-guiding conceptions behind them.
Yet, second, wouldn’t the dependence of Kant’s derivation upon a teleological presupposition commit him to deriving a moral “ought” from a psychological “is”? In fact, it does not. For at the outset of the derivation the “ought” of moral duty is already taken for granted. The teleological principle about what the role of practical reason “is,” vis-à-vis inclination, serves merely to establish the content of the “ought” already presupposed as binding. The question of its bindingness for us is something Kant puts off until the third section of the Groundwork. There he does not argue that we “ought” to act in certain ways because of the way human nature “is.” He argues instead that the presumption of freedom gives rise to the moral “ought.” So the reply to the second possible line of objection is that, rather than committing Kant to deriving an “ought” from an “is,” his teleological presupposition helps determine the content of the principle expressing an already assumed “ought,” the principle of duty.
Third, there is the possible line of objection that gaps in Kant’s derivation may be more readily apparent in subsequent versions of his argument, in Groundwork II and in the beginning of the second Critique. In these places he attempts to derive the formula of universal law from the very concept of a categorical imperative, or of a practical law. His inference there is that since the validity of morality’s command can depend on no material objects to be brought about as consequences of our actions, that command must be formal. It must command nothing other than that our maxims be capable of becoming universal laws, in a unified system of principles.
In their complaints about this later strain of argument critics have made the following assumption. They have assumed that the reason Kant supposed morality would not command us to act in order to bring about any object is that our having an inclination toward that object would be required to make the command valid for us. And since inclinations are contingent — that is, not universal and necessary — neither, then, could morality’s imperative be universal and necessary. It could only be “hypothetical.” Yet it is not obvious, these critics say, that an inclination would be required in order to motivate our bringing about an object that morality commands us to bring about. Guyer has suggested, as we saw above, that morality could command that we bring about some necessary object of reason. And for this, practical reason itself would presumably supply the motive. So, seen in this way, Kant’s argument in these places does seem problematic. If reason could motivate us (necessarily) to act in order to bring about some object, the formula of universal law seems not to be the argument’s only valid conclusion. From just the idea of a practical law or categorical imperative it might follow, also, that morality commands us to bring about the existence of some object of reason.
But the question to pursue here is whether, apart from the teleological presupposition to which Kant alludes in his derivation in Groundwork I, resources are available in the contexts of his later arguments that would answer this line of criticism. I think this is so; and I think some critics have focused too narrowly on only one aspect of inclinational motivation — its contingency. Another problematic aspect of this type of motivation that is often lost sight of is that acting rationally in order to bring about an object of inclination always requires knowledge of causes and effects. And this requirement, also, would pose a problem for the prospect that a necessary object of reason might serve as the basis for morality. Here is how.
We do not all share the same inclinations. But we are all interested, through rational self-love, in our own happiness. Hence it is plausible to say that, for each of us, though our interest in achieving happiness is inclination-based, it is not contingent. So if in Kant’s mind there is a problem with the idea that morality might command us to bring about our own happiness, it must not have been a problem of motivational contingency. It was rather a problem of ignorance. The problem with an inclination-based, prudential morality, as Kant sees it, would be that we do not readily know what to do in order to make ourselves happy, or as happy as we can be. In the text preceding his derivation of the formula of universal law in Groundwork II he writes that “imperatives of prudence cannot, to speak precisely, command at all, that is, present actions objectively as practically necessary. …” But this is not because the inclination-based motive to prudence is contingent; for to Kant’s mind it was not. It is rather because, “the problem of determining surely and universally which action would promote the happiness of a rational being is completely insoluble” (29/4: 420; cf. also Practical Reason, 5: 21). The same will be true also for any other object that morality might command us to bring about. For obeying morality’s command to bring about any object would require our knowing how to act in order to do so. Moreover, we may also need to know how to refrain from acting to bring it about now, in order to enable ourselves or others to do so more effectively in the future.
It seems to follow, therefore, that not even a necessary object of reason — the existence of which it would be rational to attempt to bring about — could be the basis for a categorical imperative. And hence Kant’s derivations subsequent to Groundwork I, though they presuppose no teleological principles, nevertheless rule out the prospect of morality’s being based on any such object. So, as with the argument analyzed above, the formula of universal law seems to be the only valid conclusion of these subsequent derivations.
Fourth, and finally, we should return to a point raised above, at the end of section III. Guyer has suggested that “humanity” would be a necessary object of reason — something that morality could command us to bring about, instead of commanding us to act only on maxims that can become universal laws. In addressing this prospect initially I acknowledged that Kant did regard “humanity” as a rational end. But I showed how he supposed that humanity would be an object of reason in a special way — a way making the principle of action derivable from it logically equivalent to the formula of universal law. Thus, I claimed, Kant’s derivation of that formula can still be valid, because the principle of humanity could always serve as an alternative conclusion.
Yet some claim to have found that in their application these principles are not equivalent. Consider the famous example of the would-be murderer inquiring after the whereabouts of his intended victim. It has been said that in a case like this the formula of universal law permits telling a protective lie, while the formula of humanity prohibits the lie.Footnote 25 Here I do not want to offer any resolution of this notorious problem for Kantian ethics. All I want to show here is how the same considerations that Kant appealed to in support of the formula of universal law can be extended to support a plausible interpretation of the formula of humanity. Showing this should help show how the two formulae can be logically equivalent, and so why the interpretive presumption of their equivalence should prevail in their applications to problems and cases.
In explaining above how Kant derived the formula of universal law, I emphasized the analogy with conformity to universal law among objects of nature. I compared natural laws that would imply judgments of experience with universal laws of action extrapolated from maxims. A judgment should be rejected as false if the law from which it would be derived could not belong to the systematic unity of nature — just as an action should be rejected as wrong if its maxim could not become a universal law in the system of practical principles. I think it fruitful to interpret Kant’s concept of a maxim as if universalized maxims of action would be comparable to natural laws, in a system of nature considered teleologically. This is consistent with his sometimes expressing the formula of universal law in terms of natural law — as in, “I ought never to act in such a way that I could not also will my maxim to become a universal law of nature.” A maxim indicates a course of action to be taken, in a circumstance, and for an end. When the maxim is regarded as a natural law, the action in the circumstance can be seen as the cause, and the end to be achieved can be seen as its effect. It then becomes easier to see how actions and their maxims might belong to a natural, teleological order of means to ends. And the unity of such a system would imply a final end, with every action in the system serving as a means, however remote, to that end in itself.
In order now to see how the formula of universal law and the formula of treating humanity as an end in itself would be equivalent, we may start with the following observation. Any action that would treat the end in itself merely as a means to something else cannot conform to the systematic unity of a natural, causal order of actions as means to ends. Consequently, the maxim of the action could not conform to universal law as such in that system; which is to say that it could not become a universal law. So this shows how any action contrary to the principle of humanity as an end in itself — any action treating humanity merely as a means — would also be contrary to the formula of universal law. But would the converse also hold? Would every action contrary to the formula of universal law be contrary also to the formula of humanity? If we equate Kant’s concept of “humanity” with freedom, it seems relatively easy to see how it would. Any maxim of action that cannot become a universal law is one that, if acted upon anyway, would restrict the freedom of some agent(s) in the system of actions as means to ends. Thus, with the occurrence of an action on that maxim there would be at least one agent who would not enjoy freedom of action equal to that enjoyed by the agent acting.Footnote 26 The effectiveness of the latter’s action would depend upon the restriction of the former’s freedom. And this seems to be the core of what is meant by treating the humanity of the former merely as a means. So here we see how it can be said in response to the fourth possible objection that the two formulae should be interpreted as equivalent — any violation of either seems also to be a violation of its counterpart.
6. Conclusion
In the preceding I have attempted to defend Kant’s derivation against different charges of gaps in his reasoning. For this I have emphasized some presuppositions of the premises of his argument that are accessible in its immediate context. One was that the argument is intended to derive the principle of action that would guide an agent in morally worthy action. Another was that only reason can be expected to determine the performance of such an action. This second presupposition is supported by a teleological principle appearing several pages earlier. A related, third presupposition of the argument was Kant’s assumption that morally worthy action would spring from reason’s native interest in systematic unity. This assumption shows how he was entitled to derive the formula of universal law from the idea of actions’ conformity to universal law as such, with the help of an analogy drawn from his philosophy of nature. Just as the laws implying judgments of experience must conform to universal law in order for those judgments to be objectively correct, so agents’ maxims must conform to universal law in order for their actions to be morally correct.
What we can now see more clearly, after having demonstrated the validity of Kant’s derivation of the formula of universal law, is the core of his conception of morality. Morality aims not at bringing about or maximizing the existence of some object, whether of inclination or of reason. Nor does it require mere conformity of action to laws valid for everyone, regardless of what these laws may be. Instead, in Kant’s view, morality is a systematic order of principles of free actions, on the model of the system of laws of nature. And just as we can know a priori that objects of experience conform to the system of laws of nature, so also can we know a priori that actions ought to conform to a system of laws of freedom.
