This paper will discuss Kant’s second formulation of the moral law in the Groundwork, the “formula of humanity as end in itself”:
FH: “Act so that you use humanity, in your own person as well as in that of another, always also as an end and never only as a means” (G, 429). 1
Most people associate Kantian ethics most closely not with FH but with the “formula of universal law”:
FUL: “Act only according to a maxim by which you can at the same time will that it should be a universal law” (G, 421).
But there are several reasons for emphasizing FH. First, there are notorious problems, arising from the criticism of Hegel and others, with FUL as a procedure for moral reasoning. For those who are sympathetic to the criticisms of FUL (and I am among them), it is worth looking closely at FH to see if we cannot do better by Kantian ethics if we associate it with FH. Secondly, the idea of human dignity which grounds FH is in any case the Kantian principle which arguably has the greatest resonance with our culture’s moral consciousness. This idea even has the greatest universal appeal, since it seems to ground the human rights in terms of which decent people everywhere frame both their protests against obvious wrongs and their ideals of a better world. Thirdly, it is to FH which Kant himself nearly always appeals in the Metaphysics of Morals when he derives particular ethical duties.2 FH is also the only formulation of the moral principle Kant ever invokes in relation to the right of human beings, which makes FH uniquely pertinent to issues of political right and international law, including the final end of perpetual peace (G, 430; MM, 6:236, 238, 239, 270, 278, 295). Finally, FH has much to teach us about Kantian ethical theory itself. For (in contrast to common misunderstandings of that theory) it presents ethics as based on an end, whose character, however, clearly distinguishes Kant’s theory from all forms of consequentialism.
FH strikes many people as high-sounding but too abstract to yield determinate moral prescriptions, and hence just as “formal” and empty as FUL. People wonder what it means to claim that human beings are “ends in themselves,” and whether Kant has any good argument for according them this special moral status (whatever it may turn out to be). These worries are not entirely without substance, but I think they are due partly to a failure to understand some of the value conceptions employed in FH and the meaning of claims Kant employs in its defense. These are the matters on which I hope to shed some light.
In this paper I will try to do three things. First, I will try to say something about what FH means, by discussing the different value conceptions involved in it and their relation both to FH and to one another. Second, I will examine Kant’s derivation of FH in the Groundwork, his argument that humanity does have the value FH attributes to it, of being an end in itself. Finally, I will try to say something about how FH is to be interpreted and applied.
FH is introduced by way of an inquiry into the possibility of a will’s conforming itself to a categorical imperative or practical law. It is supposed to address a question to which many of Kant’s critics think he has no good answer: namely, why—for what reason or motive, or in the name of what value—should we obey categorical imperatives? FH formulates the moral law in terms of the value which provides the kind of ground or reason which could motivate a rational being to act on a categorical imperative. In other words, FH provides the content of what Kant described in the First Section of the Groundwork as the motive of duty—an idea which many have found objectionably cold, empty and forbidding. It may make that idea more attractive to realize that for Kant, to act from duty means to act out of esteem for the worth of humanity in someone’s person as an end in itself. We then see that when Kant regards the cold-hearted man who helps others from duty as more deserving of esteem than the warm-hearted person who helps them from sympathy, he is not saying that it is bad to care about people and good only to follow abstract rules. Instead, he is saying that it is better to care about people because we recognize them as beings with worth and dignity, who have a genuine claim on our concern, than because we just happen to like them or because in our present mood helping them makes us feel good.
Kant distinguishes two types of grounds on which a will may act: subjective grounds, based on empirical desire for an object, and an objective ground, which is also an end, but one given by reason alone and valid for all rational beings. A subjective ground is called an “incentive” (Triebfeder), while an objective ground is called a “motive” (Bewegungsgrund) (G, 427-428). The most striking thing Kant says here is that a motive must always be an end (Zweck). This might seem to contradict what he says elsewhere. In the Critique of Practical Reason, for instance, Kant distinguishes between “formal principles” of the will, whose determining ground consists solely in the “legislative form” of the agent’s maxim, and “material principles,” whose determining ground is the end of the action. Formal principles alone, Kant says, can be practical laws; material principles are all empirical, fall under the principle of one’s own happiness, and are opposed to the determination of the will by reason alone (CPracR, 5:21-22).
Yet on closer inspection there is no real disagreement between the two texts. The Groundwork draws the same distinction between “formal” and “material” principles, characterizing material principles as based on “incentives,” and formal principles as based on motives (G, 427). When the second Critique says that the ground of a formal principle is the “legislative form of the maxim,” that is “the mere form of giving universal law” (5:27), the Groundwork merely tells us that this form, when it motivates us, constitutes a certain distinctive kind of end. If it surprises us that Kant should have characterized the legislative form or the motive duty as an end, we should investigate further to see why he does so.
Light is shed on this question by something Kant argued twenty years earlier in his prize essay, Inquiry Concerning the Distinctness of the Principles of Natural Theology and Morals (1764).
No specifically determinate obligation flows from {any formal grounds of obligation] unless they are combined with indemonstrable material principles of practical cognition. . . . They cannot be called obligations as long as they are not subordinated to an end which is necessary in itself (2:298-299).
To those who think of Kant’s ethical theory as “deontological” in the sense that it is a theory which regards moral principles as binding independent of any end served in following them, it should be enlightening to find Kant explicitly rejecting any such position, and to realize that it is this rejection which lies behind the Groundwork’s argument that a rational will can be motivated to obey a categorical imperative only by a distinctive kind of end.
Kant begins with a hypothetical account of the kind of value he ascribes to humanity or rational nature.
Suppose, however, that there were something whose existence in itself had an absolute worth, which, as end in itself could be the ground of determinate laws; then in it and it alone would lie the ground of a possible categorical imperative, i.e. a practical law (G, 4:428).
Here Kant invokes three distinct value conceptions.
First, there is the conception of an end in itself or objective end, whose worth, as the ground of categorical imperatives, is unconditional, independent of desire and valid for all rational beings, a “motive” constraining the will through its own faculty of reason. The opposite of an end in itself is therefore a relative end, whose worth is relative to, conditioned by and dependent on the subjective constitution of each particular rational being, and hence varying contingently from one such being to another. This sort of end would be only an “incentive” and could ground only hypothetical imperatives (G, 428).
Second, Kant introduces the concept of an existent end, or, as Kant also calls it, a self-sufficient end (G, 437). Kant opposes this sort of end to an end to be affected, that is, some thing or state of affairs which does not yet exist, but is to be brought about through an agent’s causality (G, 437). An existent end, by contrast, is something that already exists, and whose “existence is in itself an end,” having worth as something to be esteemed, preserved and furthered. This concept perhaps requires a bit of discussion, because people are often tempted to think that the concept of an end is nothing but the concept of a not yet existent object or state of affairs to be brought about. Until this prejudice is overcome, the idea of an existent end may appear nonsensical, self-contradictory, or at least peculiar and suspicious. But it is nothing of the kind.
An end is anything for the sake of which we act (or refrain from acting). Or, what I think amounts to the same thing, an end is something whose value provides some kind of terminus in a chain of reasons for an action. These descriptions fit ends to be effected, since when we build a house, our action is for the sake of bringing the house into being, and the house’s value to us is the reason why we build. But they also fit existing things or states of affairs, for whose sake we act. My self-preservation as an existent end, because I cross the street at the crosswalk (rather than charging out into traffic) for the sake of preserving my life, and the value I place on my life is my reason for crossing the street in the way I do; of people who drive recklessly or put themselves in needless danger we say that they place less value on their lives than most of us do.
Other kinds of existent ends often play a role in our actions. When people bow their heads or doff their hats to their country’s flag or to a religious object, they may have no end to be effected, except perhaps the successful performance of the gesture of veneration itself. But they certainly do act for an end, namely, the value of the revered object. It is for the sake of this value that they perform the act of respect or reverence. Finally, when I build a house for myself and my family I build not only for the sake of the value of the house but also for my sake and that of my family; in fact, I value the house as our dwelling only because I place value on myself, my spouse and my children. The existence of the house is an end to be effected, but we are the existent ends for whose sake the producible end is pursued.
Kant’s theory maintains that to act morally is always to act for the sake of the value of humanity in a person. FH is based, in other words, on the worth of humanity regarded as an existent end, that is, as an object of respect, esteem or veneration. From the standpoint of FH, all conduct is regarded fundamentally from the standpoint of what it expresses about the agent’s valuation of humanity. Morally good conduct is that which expresses respect for humanity as an existent end, while bad conduct is bad because it expresses disrespect (or lack of respect) for this value.
But what is it to value, respect or esteem the value of humanity? We will be misled here if we think of “valuing,” “respecting” and “esteeming” merely as subjective attitudes or states of mind, and interpret FH as commanding only that we put ourselves in such inner states. If I make a promise to you which I don’t intend to keep, then I am treating humanity in your person with disrespect, no matter what subjective attitude I may have or what inner state I may be in. On the other hand, in dealing honestly with you, I treat you with respect in that dealing even if I do so only from self-interested motives and will cease treating you with respect as soon as it ceases to serve my self-interest. The actions commanded by FH are those which express respect for humanity, whatever the subject’s feelings toward humanity may be. Of course Kantian ethics is not indifferent to what motivates the respect or disrespect shown to humanity. An action which expresses a respect the agent does not feel accords with duty, but it has no moral worth because it is not produced by the motive of duty, which is the worth of humanity as an existent end in itself.
Kant famously insists that practical philosophy must not begin with an end or object of reason and attempt to derive the moral law from that, but must instead begin with the law or principle of reason and through it determine the object or end of practical reason, or the good (CPracR, 5:58). He insists that determination of the will by its objects always results in heteronomy of the will (G, 440). In such claims, he appears always to have in mind an end to be produced (e.g. the highest good or summum bonum), as something to be “obtained” through conduct which follows the law (CPracR, 5:108). He is mainly concerned to reject the view that moral principles may be based on the representation of a future object or state of affairs, such as happiness or well-being, the representation of which gives pleasure and stimulates the will to actualize the object of the representation (CPracR, 5:22-26, 58-62). Kant does not mean to deny that conformity to the law itself must be motivated by a value; on the contrary, he claims that only a certain existent end is capable of determining the will to follow the moral law (G, 427). This end, the worth of humanity as end in itself, does not undermine autonomy simply because, as we shall see presently, its value is necessarily connected with autonomy. The worth of humanity as an end in itself (in our own person) may even be seen as the value which grounds autonomy itself, from the standpoint of our motivation to act autonomously.3
The third value conception we need to consider is that of an end with absolute worth or (as Kant also says) dignity, something whose value cannot be compared to, traded off against, or compensated for or replaced by any other value (G, 434). This is to be contrasted with an end with only relative worth or price, whose value can be measured against the value of something else and can be rationally sacrificed to obtain something else of equivalent or greater worth. It is this absolute worth in our own person, and specifically in our rational capacity to set ends and make laws, which provides the motive for obeying the laws we give ourselves through reason. That is the sense in which even autonomy itself is grounded on the worth of humanity as an end in itself.
Now let us consider these three value conceptions in relation to each other and to the value attributed to humanity in FH. Kant’s position is that there is only one thing, namely humanity or rational nature, which satisfies the concept of end in itself; and humanity is in fact also an existent (or self-sufficient) end having absolute worth or dignity. As far as these three value conceptions themselves are concerned, however, there is no reason the three value properties would have to be found together. There is no contradiction in supposing that an end in itself might be an end to be produced: that is, that there should be some thing or future state of affairs which we have an objective reason to produce independently of any contingent desire for it. An end in itself might also be an end with only relative worth or price. In that case, rational deliberation would be able to weigh objective ends against the ends which depend on our desires, and a certain amount of what is objectively valuable might be rationally traded away to get a greater amount of what is subjectively valuable. Conversely, we sometimes do things for the sake of non-human animals, regarding them as existent ends, even though they are not (at least in Kant’s view) ends in themselves and do not have absolute worth. Moreover, an end having absolute worth could be both a relative end and an end to be effected. This would happen, for instance, if there were an object of inclination to whose satisfaction we gave absolute priority over everything else in our scheme of values.
Therefore, Kant’s attribution of all three value properties to humanity or rational nature is not a conceptual claim about these value properties. It is instead a substantive value claim about that which in fact possesses the property of being an objective end or end in itself, namely humanity. The claim is that humanity is an end in itself whose objective value takes the form of its being an existent or self-sufficient end which has dignity or absolute worth.
What does Kant mean by ‘humanity’ when he says it is an end in itself? The term refers to the capacity to set ends and choose means to them (G, 437). It is being used interchangeably with ‘rational nature.’ For this reason, FH in no way contradicts Kant’s frequent assertions that the moral law is independent of empirical anthropology. FH also does not privilege the members of our own species over other possible rational beings, and involves no preference toward members of our own species just because it is ours (and therefore is not a form of “speciesism,” in Peter Singer’s sense). “Humanity” is one of the three “original predispositions” belonging to our nature, falling midway between ‘animality’ and (3) ‘personality’ (Rel, 6:26). The predisposition to humanity includes the “technical predisposition,” which includes all our learned skills and deliberative abilities used for arbitrary ends, and the “pragmatic predisposition,” which is the basis of our ability to compare our contingent ends and organize them into a systematic whole, which is called ‘happiness’ (Anth, 7:322-324; CJ, 5:426-427). Humanity, therefore, refers to the capacity to set ends in general, to select means to them, and to weigh and combine them into a conception of our overall well-being.
“Personality” is the rational capacity to respect the moral law and to act having duty or the moral law as a sole sufficient motive of the will; it is the basis of moral accountability, and what makes it possible for us to have a good will (Rel, 6:27-28). Personality therefore seems to be a ‘higher’ predisposition than ‘humanity.’ For this reason, it may puzzle us that Kant regards ‘humanity’ rather than ‘personality’ as the end in itself. But Kant does not regard personality as distinct from humanity. The rational capacity to set and organize ends necessarily involves the capacity to make comparative judgments of value and therefore an (at least implicit) awareness of standards of value. The foundation of all such standards, if Kant is right, is the dignity of rational nature itself. Therefore, any being which has humanity will have the capacity to recognize the worth of humanity, and will therefore have personality as well. In fact, Kant specifically links the dignity (absolute worth) of rational nature (as distinct from its status as an end in itself) to its capacity to be morally self-legislative, thus to personality rather than to humanity (G, 437-440). So the designation of humanity as an end in itself is not meant to imply that personality has any lesser value than humanity.
The point of characterizing the end in itself as humanity rather than as personality is simply to emphasize that this end is rational nature, in all its functions and not merely in its moral function. This follows directly from Kant’s argument that the end in itself must ground categorical imperatives. For since such imperatives must be necessarily binding on all rational beings, the end which grounds them cannot have merely contingent or even doubtful existence, as it would if it were present only in the good will or the virtuous person. A morality grounded on the principle that only virtuous people have worth could not issue categorical imperatives capable of determining the will. 4
Let us now turn to Kant’s derivation of FH, which is also his argument that humanity is the sole end in itself. This argument has four steps, which I will separate and number for the sake of perspicuity.
The ground of [the moral principle] is: Rational nature exists as end in itself.
[1] This is how the human being necessarily represents its own existence; to this extent, therefore, it is a subjective principle of human actions.
But [2] every other rational being also represents its existence consequent to precisely the same rational ground which is valid for me;
[3] therefore, it is at the same time an objective principle, from which, as a supreme practical ground, all laws of the will must be able to be derived.
[4] The practical imperative will therefore be the following: Act so that you use humanity in your own person, as well as in the person of every other, always at the same time as end, never merely as a means” (G, 429).
In this argument, (1) and (2) may appear to be independent premises, though as we shall see, (2) is best understood as grounded on (1). In (3) Kant infers that rational nature as an end in itself provides an objective principle, and (4) formulates that principle. The main difficulties with the argument concern the interpretation of (1). (1) evidently harmonizes with Kant’s view that propositions of ultimate good are indemonstrable and can be argued for only by appealing to what is already somehow acknowledged or presupposed to be of ultimate value. Thus (1) involves Kant’s anticipation of Mill’s idea that principles of ultimate value are indemonstrable, though they can be argued for rationally and even ‘proven’ in a looser sense of ‘proof,’ by showing that what the principle takes to be valuable is already, “in theory and in practice, acknowledged to be an end.”5
Beyond this, however, (1) is difficult to interpret. At first reading it looks like merely a universal empirical generalization about the worth people subjectively attribute to their existence. So regarded, it would clearly be false since in fact people sometimes do think of their own existence as lacking in value. Even if it were a true empirical generalization, it could not sustain Kant’s claim that every rational being necessarily represents its existence as an end in itself. Nor, finally, could (1) on this reading be put together with premise (2) to yield the argument’s conclusion, since a mere subjective representation of my existence as an end in itself would not provide any rational ground for representing the existence of every other rational being in the same way. It is only the existence of such a ground which could justify the objective principle whose existence is inferred in (3) and finally formulated in (4).
But these considerations may make it possible for us to reconstruct how (1) would have to be taken if Kant’s argument is to work. (1) would have to mean that necessarily, rational beings subjectively represent their existence in some way which brings to light an objective ground for regarding them as ends in themselves.6 (1) therefore means that whenever a human being sets an end for itself through reason, it thereby necessarily represents itself (if only obscurely or inexplicitly) as having some property which makes it an end in itself, thus providing all rational beings with an objective ground for respecting its absolute worth. In the Groundwork Kant identifies this feature as the capacity to set ends, and says that the worth of rational nature as an end in itself can be inferred from the fact that it is the subject of all possible ends.
Rational nature is distinguished from others in that it sets an end for itself. . . . It is that which must never be acted against and which must never be valued merely as a means but in every volition also as an end. Now this end can never be other than the subject of all possible ends themselves (G, 437). 7
Christine Korsgaard is on the right track in suggesting that Kant begins from the value we place on the ends we set, and infers (by means of a “regress on conditions”) that this value is grounded in the rational nature of the being who set the end. She claims that rational nature therefore possesses a “value-conferring status” in relation to the ends it sets. Because humanity or rational nature is the source of all such value, it is regarded as absolutely and unconditionally valuable, and an end in itself.8 Crucial to this line of reasoning is the idea that when we set an end we attribute to it objective value or goodness. Goodness, whether moral nor nonmoral, is objective value in the sense that the good is represented by reason as practically necessary (as either means or end) and hence as an object of will for all rational beings (G, 412). The fact that every end we set is represented by us as good therefore entails that it is represented by us as having a rational claim on the will of every rational being. The idea here is not that all goodness is only subjective (because it depends on the choice of the being who sets the end and considers it good), but on the contrary, that rational choice of ends is the act through which objective goodness enters the world, and the fact that rational nature serves as this source of all value is supposed to imply that it has objective value as an end in itself.
It might prove illuminating to consider Kant’s conception of objective value in light of J. L. Mackie’s contention that “objective goodness” is a “queer” property because it has to be “objectively prescriptive. ”9 Mackie is operating within the options standardly available within twentieth century analytical metaethics. He is supposing that either goodness exists only relative to desire (it is only “subjective”) or else it is real out there in the world (as something “objective”). In the latter case, goodness is either something which necessarily provides reasons for action (if we hold what is called “internalism”) or else such reasons must be added contingently to it (as “externalism” maintains). In the latter case, the good gives us reasons to act only if along with it there is contingently present in us some desire or attitude disposing us toward the good (which once again reduces the authority of the good to something merely subjective). The good has genuine practical authority, therefore, only if internalism is accepted. But then real goodness must be a property out there in the world which somehow of itself tells us what we must do—it must be Mackie’s “queer” property of being “objectively prescriptive”.
The options canvassed so far, however, do not include any Kant would find attractive. Kant maintains that goodness is indeed objectively prescriptive, but not because there is some queer property out there in the world which imposes itself prescriptively on the will (thus violating its autonomy). Kant instead locates the good’s objective pre-scriptivity in precisely the sort of thing which might naturally have been thought to prescribe: namely, the rational will itself. It says that the objectivity of a will’s prescriptions comes from its rational capacity to make objective laws and set ends having objective value. The source of all such value is therefore nothing except the value of the rational will itself. Rational volition is what makes it the case that other things are good (i.e. that they make objective claims on human beings). Rational volition can do this only if it is presupposed that the worth of rational nature is that which makes such claims necessarily and unconditionally. In other words, the autonomy of the will can be reconciled with the objectivity of value only if the sole object of value which is not legislated to be such by the will is the absolute value of the legislating will itself as the single necessary ground of all objective value.
Kant’s argument for FH depends on an inference from the premise that the objective goodness of all ends is grounded in rational nature to the conclusion that rational nature itself is objectively good or an end in itself. We might balk at this inference on the ground that if Y is something valuable and X is its source, it does not in general follow that X is something valuable, still less that it is objectively valuable or an end in itself. Sometimes in such cases we don’t consider X itself good at all, but merely tolerate it as a necessary evil required to procure Y. So even if we grant that the rational will is the source of all objective values, it does not seem to follow that it is valuable, still less that it is absolutely valuable.
But this objection misses the point, because in Kant’s argument rational nature is not being viewed as the source of good things (of their existence, for example), but rather as the source of their goodness itself, of the very fact that they are good, indeed, of the fact that anything at all is good. The right parallel here is rather with our attitude toward an authority as a source of recommendations or commands. It makes sense for us to take someone’s advice or prescription as authoritative only to the extent that we respect and esteem the authority itself as their ground. Hence we have reason to regard as good the ends we ourselves set only to the extent that we (at least implicitly) respect and esteem our own rational nature as that which sets them.
More generally, if something is good only through being made the object of rational choice, then the capacity to make such choices is the sole authority over goodness, or as Kant puts it, “the subject of all possible ends.” If rational nature is the prescriptive source of all objective goodness, then it must be the most fundamental object of esteem or respect, since if it is not respected as objectively good, then nothing else can be regarded as objectively good. Rational nature is therefore the only thing which could answer to the concept of an objective end or end in itself.
In the Second Section of the Groundwork, Kant has provisionally assumed that there is a categorical imperative. He has argued that if there is such an imperative, then the possibility of its determining the will depends on there being an objective end or end in itself. Thus if, as we have just seen, rational nature is the only thing which could answer to the conception of such an end, then Kant is entitled to conclude that rational nature, and it alone, is an end in itself.
This will have to suffice as a discussion of Kant’s argument for FH. I now turn to the question of interpreting and applying FH in moral deliberation. It was said earlier that to many people FH seems a highly abstract principle from which determinate moral conclusions would be hard to derive. Now it is time to consider the reasons why FH gives this impression, and to what extent it is accurate.
FH involves a kind of moral appeal which is unfamiliar in philosophical theories of morality. Consequentialist theories represent the fulfillment of our duties as bringing about desirable states of affairs; deontological ones (including Kant’s own theory, to the extent that we consider only formulas such as FUL) represent it as obedience to an obligatory rule or commandment. FH, however, bases duties on the worth of something—on the dignity of rational nature as an end in itself. This is not a rule or principle to be obeyed, but a value which grounds moral rules, providing an objective rational motive for obeying them. But it is not a value in the sense of a desired object to be brought about through obeying moral principles. It is something existing, whose value is to be respected in our actions. What FH demands of our actions is not that they pursue some good result or obey some obligatory rule but rather that they express due respect for the worth of humanity. Since FH identifies the end in itself which is the objective determining ground of the will which acts from duty, it even presents the expression of respect for humanity as the fundamental reason why we should conform to moral laws and pursue moral ends.
In ordinary life people often do things in order to show respect or reverence for someone or something. Often they behave politely toward others, thank them or congratulate them in order to manifest respect, gratitude or esteem. Earlier I mentioned saluting the flag or bowing one’s head as ways of honoring one’s country or venerating a religious object. Yet if we think only of these examples, the motives involved in expressing respect or esteem may seem suited only to matters of etiquette or ritual, and therefore an unlikely ground for a comprehensive rational morality. But if we think that expressive reasons for acting are restricted in this way, we are badly misled. For in fact every action which is done for a reason, as distinct from being merely a response to an impulse, is based on regarding something as valuable. The performance of such an action is always an expression of respect or esteem for that value, and this expressive reason for performing the action is even the ground of all the other reasons we have for performing it.
From this standpoint even our furtherance of a producible end is fundamentally a piece of expressive behavior, since it demonstrates or exhibits how much we value the object pursued. We care about someone’s good only to the extent that we value that good, and we value it only insofar as we think of the person as worth caring about. This is true even about our concern for our own good. When I choose to satisfy a desire (as distinct from merely reacting to an instinctive impulse) I do so because I judge the desire to be worth satisfying, and this in turn expresses a judgment about my own worth. When people consider themselves worthless, they tend not to care very much about their ends, and even their own happiness may not matter very much to them. This pathology expresses a piece of consequent reasoning, though if Kant is right its premise is necessarily false, since their humanity always has absolute worth.
Conversely, the basic reason why it is wrong to harm people, or to fail to help them, is that we thereby show that we place too little value on them. Consider California’s Proposition 187, which intends to deny education, health care and other basic social services to undocumented immigrants. Ignoring for the moment issues of legality and right, what is so morally repugnant about the attitude of those who proposed, supported and voted for this reprehensible measure?10 It is not merely that they have done something which will cause human suffering. For the thought of such suffering, by itself, makes us only sad, not indignant. Likewise, what makes their conduct blamable is not their “lack of compassion.” For in many cases they may feel sincerely sorry for those whom Proposition 187 will hurt. Their contention is only that they have a right to deny social services to immigrants, and that, regrettably, the state can no longer provide these services to undocumented immigrants.
However, even if we grant that everything they say here is true, there is still something outrageous about their conduct, which no decent person can fail to perceive. (Even they themselves cannot have failed to perceive it. This is why such a measure could pass only in a political season in which a sizeable proportion of the voting electorate was seized by a fanatical paroxysm of fear, hate and blind rancor, which turned them with aggressive perversity against every impulse in themselves belonging to reason, decency or humanity.) Proposition 187 says, in effect, that the State of California does not consider undocumented immigrants valuable enough as human beings that those around them are required to meet their basic needs for education and medical treatment. No doubt the supporters of Proposition 187 meant to say to these people that they should go back where they came from. And perhaps this message, nasty as it sounds, is one which they not only have the right to convey but which it might be morally permissible for them to want to convey. However, the problem with Proposition 187 is that it conveys this message in such a way as to express contempt for the humanity of the immigrants. No one can fail to be aware that humanity in their person is not being treated as an end in itself.11
But exactly how do we determine that Proposition 187 expresses this contempt for humanity? More generally, how do we know when any sort of conduct treats, or fails to treat, humanity as an end in itself? Let us consider a couple of Kant’s own examples. He alleges that suicide is always wrong because it involves disposing of a person (a being with absolute worth), merely as a means to some discretionary end, such as maintaining a tolerable condition up to the end of life (G, 429). Kant regards suicide as “degrading to the humanity in one’s own person” (MM, 6:422-423). Our duty not to make promises we do not intend to keep is based on the fact that the making of such a promise uses those we deceive merely as means, “without considering that, as rational beings, they must always at the same time be esteemed as ends” (G, 430). When we violate these narrow or perfect duties, we have done something which fails to show respect or esteem for humanity or rational nature. Our wide or imperfect duties are based on the fact that proper esteem for rational nature requires us to bring our actions into “agreement” with the existent end of humanity, by setting ends which harmonize with those of rational nature, even if the precise manner and extent of their pursuit is optional. Kant uses this consideration to argue that we must make it our end to develop our talents (since rational nature is honored by putting these at its disposal for use in pursuing its optional ends) and to help others when they need it (since we honor their capacity to set these ends by helping to further them) (G, 430; cf. MM 6:445).
Every argument from FH depends on an intermediate premise, logically independent of FH itself, which tells us what the action expresses or fails to express concerning the worth of humanity in someone’s person. Kant’s argument about suicide, for example, is:
One cannot reach the conclusion (3) from (1) without (2), or some similar premise connecting FH with suicide. But (2) is logically independent of FH, as we can easily see from the fact that it is possible to reject (3) simply because one rejects (2). For example, some people think that it is demeaning to human dignity for a person afflicted with a horrid and debilitating disease to live on in an agonized and subhuman condition. They might view the suicide of someone faced with this prospect as an act which positively respects the worth of humanity in one’s own person. To accept FH, therefore, clearly does not commit one to accepting all the conclusions (about the immorality of suicide, for instance, or lying) that Kant himself would want to derive from it. For two people who equally accept FH may disagree over propositions such as (2), which purport to tell us what certain kinds of actions express regarding the worth of humanity.
Fewer of us will quarrel with Kant’s conclusion in the case of the false promise. For coercive or deceptive actions often obviously violate FH because they adopt a means to an end which positively prevents the other person from rationally sharing the end. The use of such a means deliberately circumvents or frustrates the rational agency of the person coerced or deceived in a way which displays obvious contempt or disrespect for the value of that agency.
(2′) plays the same role in this argument as (2) did in the argument about suicide. It connects the nature of the action to FH by establishing that action as one which fails to respect humanity as an end in itself.
We may be tempted in the case of (2′), as we probably were not in the case of (2), to think that it follows directly from FH itself or is even part of the very meaning of FH. But this temptation should be resisted. For although (2) may be more controversial than (2′), it is still possible without contradiction to affirm FH while denying (2’). There are actions which frustrate or restrict someone’s agency: for example, having an end that cannot be shared by the patient and nevertheless do not fail to respect the patient’s humanity.12 There are significant controversies, for instance, over the circumstances under which paternalistic interference with someone’s agency might be permissible in the name of the worth of humanity, in order to prevent them from harming their own rational capacities or doing things which they will later regard as dishonoring their humanity. It is implausible to assume that the right answer on these issues can always be found right in FH itself. It is those who assume this who naturally infer that the meaning of FH is so flexible and disputable that no moral conclusions could be rationally drawn from it except those imported into it by an interpretation so vulnerable to controversy that FH itself is useless in the context of moral debate. People who understand FH in this way cannot be blamed for regarding it as high-sounding but virtually empty of content.
A more plausible way to look at the matter, however, is to insist that the meaning of FH is quite clear and determinate, because the concept of humanity as an existent end in itself possessed of dignity is clear and determinate. The application of FH is sometimes unclear or controversial, however, because every use of FH in moral deliberation requires an intermediate premise, logically independent of FH, which does the work of (2) or (2′), and in many cases there are legitimate questions about which acts express respect or disrespect for humanity. Controversies surrounding these intermediate premises constitute a legitimate reason for viewing FH as incapable, by itself, of settling the legitimate moral issues to which we might expect to apply it.
But it is all too easy to take such suspicions farther than is warranted. It would be entirely mistaken to claim, for example, that FH is empty or formalistic, settling nothing as regards the content of morality but leaving all the substantive moral issues still to be decided. For, in the first place, it is FH alone that serves as a moral principle in the above arguments for (3) and (3′). If (2) and (2′) are independently necessary to derive these conclusions, it is even more obvious that the conclusions cannot be derived without FH, which alone gives moral significance to the fact that certain actions treat humanity with respect or disrespect.
Secondly, that FH is a contentful principle is attested by the fact that it is itself far from being undisputed. It says that rational nature is the highest value, that it is an absolute value for which none other can be substituted, and hence implies that all rational beings are of equal value, whether they are wise or foolish, useful or useless, even whether they are morally good or bad. FH is denied by all those who think rationality does not make human beings worth more than animals, or who believe some rational beings are worth more than others, or who hold the value of rational nature to be less than absolute. A principle as controversial as FH cannot be regarded as trivial or empty of moral content.
Finally, although every application of FH requires an intermediate premise logically independent of it, it is nevertheless true that FH itself is not entirely neutral regarding such premises. For what any such premise tells us is something about what it is to value rational nature, and that value is just what FH asserts. A premise such as (2’) is easier to defend than a premise such as (2), because the direct frustration of a rational being’s agency is more intimately related to rational nature than a rational being’s termination of its life-processes, and hence a more direct and obvious expression of disrespect for rational nature. (2) in turn is easier to defend than the analogous premises in some of Kant’s other arguments from FH, such as those which depend on the claim that we always dishonor our humanity when we enjoy sexual pleasure for its own sake (MM, 6:424-425).
Our concept of rational nature itself is a corrigible, hence variable one. That is why there is also variability, from person to person, or society to society, in what we take to show respect for rational nature. But we do have a certain amount of knowledge about rational nature, and this guides our judgments about what is valuable about it, and hence how respect for this value should be expressed. There is no algorithm or decision procedure leading from such knowledge and such judgments to intermediate premises which do the work of (2) and (2′), but because rational capacities themselves are something real about which knowledge is possible, the defense of such premises need not be merely a matter of incommensurable intuitions or irrational prejudices.
Therefore, there is after all a grain of truth in the temptation to treat intermediate premises such as (2) and (2′) as part of the very meaning of FH. For FH itself, if taken seriously, can help to shape how people think, feel and perceive the various ways humanity is treated. Since FH holds that all rational beings have absolute worth, it implies that they all have equal worth. Of course it doesn’t follow from this that all rational beings deserve identical treatment in any particular respect. Premises specifying how respect for human dignity is to be expressed, and which aspects of people are relevant to this, can be used to reconcile FH with all sorts of social, political and economic inequality. Kant himself regards FH as compatible with some inequalities which would today be regarded as abominable or at least controversial, such as the exclusion of the vast majority of society from the right to vote or hold political office, and the authority of the male heads of households over wives, children and servants within families.
Yet when a social order treats some people better and some worse in ways that they themselves regard as essential to their self-worth, there is a presumption, based on FH itself, that this social order fails to respect the humanity of those who receive worse treatment. As more people come to regard political participation as an essential function of their rational agency, it becomes harder to regard exclusion from it as anything but an expression of disrespect for the humanity of those excluded. Any such egalitarian presumption might be rebutted, of course, by showing that greater equality in one area could be achieved only at the cost of more fundamental failures to respect humanity in other areas—as defenders of economic inequality do, for example, when they argue that greater economic equality would require unacceptable restrictions on external freedom. But if we take FH seriously, we ought to be discontented with trade-offs when they concern matters essential to human self-worth and willing to adjust our views about what respect for humanity requires in order to give the equal absolute worth of every human being its due. For example, when untrammeled external freedoms in the economic sphere lead to large inequalities in wealth and status between human beings which are degrading to those placed in an inferior position, this should make us reluctant to regard the respect for economic liberty as vital to respect for humanity, and willing to see this liberty curtailed in the interests of human dignity. In short, there is in FH itself a powerful tendency toward the relentless expansion of human equality in all areas of life that matter to human self-worth.
In this paper, Kant’s writings will be cited according to the Akademie Ausgabe by volume:page number.
Kant glosses Ulpian’s principle of natural right, honeste vive, as “asserting one’s worth as a human being in relation to others, a duty expressed by the saying: ‘Do not make yourself a mere means for others but be at the same time an end for them’” (MM, 6:237). The innate right to freedom is said to “belong to every human being in virtue of his humanity” (MM, 6:237). There are fourteen ethical duties explicitly enumerated by Kant. Of these, only (1) the duty of beneficence to others is grounded on FUL (MM, 6:389, 451, 453). Nine of the remaining thirteen are explicitly based on FH, and the other four are based on it by implication. The emphasis on FH is strongest in the case of duties to oneself. (2) The duty against suicide is based on the fact that disposing of oneself as a mere means to some discretionary end is debasing humanity in one’s person (MM, 6:423); (3) the duty against carnal self-degradation on the fact that it “violates humanity in one’s own person” (MM 6:425); (4) that against drunkenness on the fact that the drunkard is “like a mere animal,” and cannot be “treated as a human being” (MM, 6:427). (5) Lying “violates the dignity of humanity in one’s own person” (MM, 6:429) and (6) the self-respect opposed to servility is a duty “with reference to the dignity of humanity within us” (MM, 6:436). (7) The human being’s duty to develop our natural perfection is one “he owes himself (as a rational being)” because it is “bound up with the end of humanity in our own person” (MM, 6:444, 392). (8) Violation of the duty of gratitude is “a rejection of one’s own humanity” (MM, 6:454) based on “pride in the dignity of humanity in one’s own person” (MM, 6:459), while (9) the duty to sympathize with others holds insofar as “the human being is regarded not merely as a rational being but as an animal endowed with reason” (MM, 6:456). (10) All duties of respect to others are grounded on “the dignity in other human beings” (MM, 6:462). There is no explicit appeal to any formula in the case of four duties: (11) our duty to ourselves not to be avaricious (MM 6:432), (12) our duty as self-judge (MM, 6:437-440), (13) our duty to increase our moral perfection (MM 6:446-447), or (14) our duty to ourselves regarding non-rational beings (MM, 6:442-443). But (12)-(14) are all duties relating to our acting from the motive of duty, which (as we have just seen) is explicitly grounded on our dignity as rational beings; and (11) our duty to avoid avarice is defended on the ground that it impairs our rational nature in respect of the use of money (self-impairment is also used as the basis for (4).) Reference to FH also grounds Kant’s discussion of five of the six enumerated vices opposed to duties to others: envy, ingratitude, arrogance, defamation, ridicule (MM, 6:458-461, 465-467); no explicit appeal to any formula occurs regarding the vice of malice (MM, 6:460).
I am grateful to Thomas Hill, Jr. for pressing me to clarify this point.
The same considerations tell us how Kant’s theory should deal with the fact that humanity or rational nature apparently comes in degrees and also that there are borderline cases of it. A categorical imperative or practical law either has an objective ground or it does not, so the existence of an end in itself cannot be a matter of degree. Every being with the capacity to set ends according to reason is therefore equally an end in itself, irrespective of how well or badly it may exercise this capacity. By the same token, Kant’s theory implies that if some members of our biological species lack ‘humanity’ in this sense, whether temporarily or permanently, then not every member of our species is an end in itself. It seems evident that young children and people whose rational capacities are severely impaired simply do not have the capacity to set ends according to reason. On Kant’s theory, therefore, they have to count as nonpersons and are not ends in themselves. By itself, however, this does not entail that they should not be treated, as far as possible, as we treat persons, but only requires that additional arguments need to be given for treating them as persons. Most such human beings either have been rational agents at one time or can be expected to become rational agents in the normal course of things. Respect for the worth of humanity does not entail precisely the same respect for beings whose humanity is merely potential, or actual only in the past, or temporarily absent, but it may have some implications just the same. Even human beings who have never been and will never be rational agents are often bound to actual persons by the same emotional ties that other persons are, and the concern that others have for them may be indiscernible from that which they have for other persons. Kant’s view implies that such ties and concern are not morally required, but where it exists, the theory entails that it has moral consequences, since respect for the humanity of those who have this concern requires us to share their morally permissible ends. It is a complex question how we should treat various beings (human or non-human, born or unborn) who lack ‘humanity’ in the strict Kantian sense but for whom moral claims can still be made in the name of humanity. Kantian moral theory sheds light on such problems by placing them in a theoretical context, but it would be a misunderstanding to think that it can or should pretend to a simple or neat solution of them. The Kantian position, admittedly controversial, is that such moral problems should be settled solely by reference to the absolute worth of humanity or rational nature as an end in itself. Mere sentience, for example, considered apart from the worth of rational nature, is not a basis for moral status. But it does not follow that there could not be indirect reasons, grounded in the value of rational nature, for giving moral weight to the pleasures and pains of non-rational creatures (MM, 6:442-443).
John Stuart Mill, Utilitarianism, ed. George Sher (Indianapolis: Hackett, 1979), pp. 4-5, 34-40.
That Kant believes something like this is indicated in his essay Conjectural Beginning of Human History, where he provides an imaginary account of how human beings first left the path of animal instinct and began to use reason to set ends. One result of this, Kant says, is that the human being “came to understand, albeit only obscurely, that he is the true end of nature,” and made “the claim to be an end in itself, estimated as such by every other, and to be used by none merely as a means to other ends”. (MM, 8:114).
A similar thought is expressed in the Critique of Judgment where the status of rational beings as the ultimate end of nature is inferred from their capacity to conceive of ends and organize them into a system: “The human being is the ultimate end of creation here on earth because he is the only being on earth who can form a concept of ends and use reason to turn an aggregate of purposively structured things into a system of ends” (CPracR, 5:427).
Korsgaard, “Kant’s Formula of Humanity,” Kant-Studien 77 (1986), pp. 194-197.
John L. Mackie, Ethics: Inventing Right and Wrong (New York: Penguin, 1977), pp. 26-27.
We need to distinguish the question whether Proposition 187 is constitutional (that is, in conformity with the Constitution of the State of California and with the United States Constitution), from the further question whether it is consistent with natural right. But even supposing (what is far from evident) that Proposition 187 is not objectionable on either of these grounds, we may still ask whether those who support it and vote for it are performing actions which are morally objectionable because they thereby fail to treat humanity in the person of the undocumented immigrants as an end in itself having dignity or absolute worth. It is only this last question that I am concerned with in the present discussion. It is worth focusing on this sort of question because in political discussions in states with a liberal ethos there is often a peculiar tendency to infer from the premise that we have the right to do something (for example, to agitate for a certain change in laws, or to vote for it) that it is morally permissible for us to do these things—as though anyone who points out to us that we are violating a moral duty in using our political rights in a certain way is trying to deprive us of those rights. The point that seems to me obvious is that even granting that the voters of California were acting within their rights in passing Proposition 187, they were behaving shamefully and blamably in doing so. Further, it is an extremely harmful feature of our political culture that it should confuse the rightful with the morally permissible in this way. For on the one hand, it encourages people to think that when they morally disapprove of something (some type of sexual behavior, for example), then what they disapprove of is a fit object of legal coercion, or at least that there can be no right (subject to legal protection) to engage in it. Conversely, if they held themselves morally accountable for their political behavior (for the way they vote, for instance) then measures such as Proposition 187 would stand no chance of passage and the issue of whether it is constitutional or something the state has a right to do would never arise. Kant might sometimes seem to be taking issue with this last point, when he suggests that moral beneficence might be superfluous from the standpoint of human welfare, and that all that is needed for maximal human well-being is that people merely respect one another’s rights (G, 4:423; MM, 6:452, 458). But it is noteworthy that in such contexts Kant also occasionally insists that existing differences in wealth, for example, are contrary to right, resulting from a “general injustice” in the social system of distribution (27:416, 20:140-141). (I am grateful to Professor Roman Hruschka for pressing me to clarify this point.)
Another source of offense is that those who would deny education and social services to immigrants ignore the positive contribution they make to society (the work they do, the taxes they pay, and so on). This one-sided portrayal of immigrants is also a denial of their humanity, since in effect it views them solely as objects to be exploited and not as people whose activities have value and whose needs ought to be satisfied. But from the standpoint of FH, it is objectionable to argue that immigrants should be eligible for social services only because they make an economic contribution to society, as if others have reason to satisfy their needs only if they pay for it. For the dignity of humanity is not conditional on one’s ability to pay. Kant would point out that each of us owes it to the worth of humanity in our person not to be dependent solely on the charity of others, since such dependence is humiliating and degrading, and amounts to a form of servility (MM, 6:436). But from the converse standpoint, of those who have it in their power to meet the needs of others, it is wrong to withhold help from those dependent on it or to treat them in a humiliating fashion because they need it. On the contrary, benefactors must try to show those they help that they are honored that the recipients should accept it—which they always are, since the recipients are beings having absolute worth (MM, 6:453).
Kant himself would cite an act of just punishment as an example of this, since he maintains that it is impossible to will one’s own punishment and yet that just punishment respects the dignity of humanity in the person of the criminal (CPracR, 5:37-38; MM, 6:331-337).