The 2nd CI is, I would argue, the most influential piece of Kant’s moral philosophy. The idea that persons should never be treated as just a means to an end is a recurring trope in the sermons and speeches of Martin Luther King, Jr. It is the undergirding principle of the notion of “informed consent” that drives so much of the ethical decision-making in contemporary medical practice. As our ethical and legal thinking about sex has moved away from the patriarchal conception of it—in which the chief wrong of rape is that it violates a man’s proprietary claim on the sexual use of “his” woman (and in the case of an unpledged virgin, the future husband’s right to exclusive use)—what has replaced it are Kantian conceptions embodied in such ideas as “sexual objectification” and the primacy of consent in distinguishing between acceptable and unacceptable sexual acts.
This is just a partial list, but it gives a sense of how deeply influential Kant has been in shaping our modern understanding of morality. So let's look at the 2nd CI. In Kant’s words, it runs as follows:
Act so that you treat humanity, whether in your own person or in that of another, always as an end and never as a means only.The first thing to keep in mind here is what Kant takes “humanity” to designate. For him, that term is simply short-hand for a being with a rational nature. While there is more to us than rationality, our rationality is what Kant has in mind when he speaks of our "humanity." And Kant understands reason as a faculty that discerns what can be consistently generalized. On the level of “pure reason,” it sees experience in terms of general or universal categories and recognizes what is consistent with (or contradicts) what is given in experience. On the level of “practical reason,” it legislates principles of behavior in terms of this same standard of consistent universalizability.
In order to be rational on the practical level, Kant thinks we must act in accord with principles or rules of action that can in theory apply to more than the particular case at hand. That is, we need to follow rules that abstract from the particulars of who is doing the act, to whom, where, etc.. The rule should be formulated without any particular referents at all (including, especially, reference to oneself). But, insofar as it is a rule of action, it is directed to a being that is capable of acting on such rules—in other words, a rational agent.
Rational principles must thus have the form, “Under conditions of this sort, a rational agent is to perform an action of such-and-such kind.” When my reason legislates rules for me, it thus doesn’t lay down a rule for me alone, but for all rational agents. If I cannot, without some contradiction in my will, endorse the rule as universally binding on all rational agents, then I am treating myself as a special case—and as such, the rule is not a rational rule in the indicated sense. Reason is the faculty that sees the general in the particular, that abstracts from this isolated experience, this person, this situation, and finds a pattern that can apply to all relevantly similar cases.
All of this, of course, is what Kant says in developing his First Formulation of the Categorical Imperative (1st CI). In a sense, the 2nd CI builds on the foundation laid in the first—insofar as it relies on the conception of rational agency defined in the course of explicating tghe 1st CI. What the 2nd CI says, in effect, is that rational agency is to be valued for its own sake, “as an end,” and that it is therefore illegitimate ever to treat a rational agent merely as a means to an end.
So: What is being said here? And why does Kant think this is an inescapable principle of reason? As to the first question, the key involves understanding what it means to treat rational agency as an end. What the 2nd CI rules out isn’t the use of rational agents as a means to an end, but rather their mere use. That is, you can use someone to further your own objectives so long as you also, at the same time, treat their rational agency as an end.
In Kant’s case, here, the means/end distinction picks out two ways of valuing something. When you treat some entity E as a means, you’re aiming at promoting or achieving some goal G, and you are making use of E in order to further this aim. To aim at G is to attach value to it. Either you desire G already and so are valuing it as the object of desire, or you have made a decision to treat G as worthy of pursuit.
But making G the goal of your activity isn’t enough for G to be valued as an end. That happens only if the attainment of G is what you’re after in such a way that you are, if you will, “content” with G coming about (or being preserved, etc.) whether or not anything else comes of it. This would not be the case if you wanted G to come about solely because you saw G as a good way to promote G*, and G* is what you’re really after.
In that case, if G came about (or were preserved) but G* did not come about as a result, G’s existence would have no value for you. Its value for you would be wholly parasitic on the value of G*, such that without G conducing to G* you could care less about G. In that case, you value G merely as a means.
Likewise, that something—say E—is valued as a means does rule out it being valued as an end. Even if I make use of E to achieve G, I might still be treating E as an end if I value E apart from its role in promoting G (or anything else). The key is that in pursing G, I do so in a way that also promotes E. If all I will is the attainment of G, and the promotion or preservation of E is a matter of indifference to me so long as G is attained, I am treating E merely as a means. If what I will is both the attainment of G and the preservation of E for its own sake (and not merely for the sake of being able to use it again to achieve other goals), and I make use of E in the pursuit of G in such a way that E is promoted by this use, then I may be treating E as a means but I am not treating it merely as a means. I am also respecting E as an end.
So, when Kant says that we must treat the humanity in ourselves and others as an end and not merely as a means, he is saying that the rational agency of persons must be valued for for its own sake, such that if we ever make use of a person for some further end of our own, we are committed to doing so only in such a way that we at the same time preserve/promote/respect their rational agency.
Now Kant connects rational agency in an essential way to autonomy. Autonomy is about being self-governed—being free to act on laws that we make for ourselves, rather than being determined by external laws (what Kant calls “heteronomy”). If we are determined in our choices by the laws of physics as they play out in the complex organic systems we call our brains, then we are subject to the rule of an external power. And so, when we act on desires that come to us through the interplay of biological predispositions and environmental conditioning (when we act on “inclinations”), we are being heteronomous rather than autonomous. It is only when we act on the “maxims” given to us by the exercise of our reason that we are being autonomous. To be moral is to be autonomous, and as such morality presupposes the possibility of being governed by rational laws rather than by our inclinations. It presupposes, in other words, the capacity of the will to be confronted with the urgings of inclination and the dictates of reason, and to choose reason for reason’s sake.
This, for Kant, is freedom of the will: the ability to choose between autonomy and heteronomy, between the universal moral laws legislated by our own reason and the particular desires produced in us by heredity and environment. Rational agency is thus a synthesis of two things: our reason does the job of establishing a law for us to live by, and our will chooses to be subject to that law. But in human beings, both of these capacities are limited. Sometimes we reason poorly. And we all suffer from “weakness of will.” Some desires are (or so it seems) just too strong to resist, even when we think that acting on them is wrong.
Now suppose I know something about human beings—namely, that when presented with certain stimuli (a gun to the head, say) their will becomes virtually impotent to do anything but follow the survival instinct. In other words, suppose I know that in general, the capacity for rational autonomy is limited because certain inclinations are just too powerful. And suppose I take advantage of this general limitation to get people to do things I want them to do—like open a safe so I can make off with the money inside. If I take advantage of a limit in the human capacity for rational agency by creating conditions under which I know that this capacity vanishes; and if, furthermore, I do so for the sake of achieving some further end (getting rich), then I am treating a person as a means to an end in a way that fails to value their rational agency. Rather than being regarded as something to be treasured and promoted, they rational agency is being treated as something to be bypassed as a pesky impediment to getting what I want.
Sometimes, rather than taking advantage of general limitations in the scope of human rational agency, I take advantage of a more specific or personal limitation. You might have been raised in such a way that you have a powerful guilt complex that essentially takes charge of your behavior when triggered. And so I trigger it deliberately in order to manipulate your behavior to get what I want. Such psychological manipulation, which takes advantage of an individual person’s “weakness of will”—treating such weakness as an opportunity to be exploited rather than as a limitation in rational agency to be overcome—fails to show respect for rational agency for its own sake. If I get you to do what I want through such manipulation, I am again treating you merely as a means.
Or suppose I decieve you--convincing you that I like you, say, when I really think you're a disgusting dweeb--in order to motivate you to do something that would serve my interests, such as drive me to the mall. In that case, I have attempted to arrange things such that what you think you're doing--driving a friend to the mall--is different from what you are actually doing (driving someone who loathes you). But in that case, you are not in a position to choose to do what you are in fact doing should you agree to drive me. I have deliberately sought to impede your capacity to get what you choose. Such general interference in choice-making also prevents rational deliberation about the maxim of one's action. I am at trying, therefore, to do something that would, if successful, block your capacity to act as a rational agent...for the sake of getting to the mall.
There are many more examples, but this is sufficient to get a sense of what Kant’s 2nd CI is saying. In effect, in order to treat humanity as an end, I must avoid making use of someone in a way that fails to respect their rational agency. Which means, in general, that when I make use of another person I must first ensure that their rational capacities are fully functional, that I am doing nothing to block or impede or sidestep their ability to choose to follow the dictates of their reason, etc. Put simply, at a very minimum I must secure their fully informed consent without taking advantage of (let alone cultivating) their weakness of will.
The question that remains is why Kant thinks that obedience to the 2nd CI is a demand of reason. Think of it this way. To be rational in one’s behavior is, according to Kant’s understanding of things, to choose to follow laws of reason rather than inclinations. To make such a choice is to value being rational for its own sake (as opposed to valuing it merely as a way of achieving one’s inclinations). But reason finds or endorses the general, not the particular. To act rationally is thus to value my reason as an instance of rationality, and hence to value (for their own sakes) all instances of rationality. To be rational is therefore to value rationality wherever I find it—for its own sake, or as an end. Thus, to be rational requires that I treat the rational agency of all persons as an end to be respected, and hence that I never treat a person merely as a means to my ends.
It is true, of course, that if I follow an inclination I am implicitly valuing inclinations as such. But insofar as I indulge my inclinations I am favoring inclination over reason. And while reason is a faculty that insists on consistency, inclination is not. It is irrational to be inconsistent. It is not “non-indulgent” to be inconsistent. So, if by an act of will I choose inclination over reason, there is nothing in that act of will which in any way forces or demands that I value the inclinations of others (or my own inclinations, for that matter) in a consistent way. But if by an act of will I choose reason, then failing to respect the rationality of others amounts to my not choosing reason after all. I cannot actually choose reason unless I consistently value the rational agency of all human beings as an end. That’s what it means to choose reason.
The upshot is this: Reason, by its nature, demands that reason be valued in a way that inclination, by its nature, does not demand that inclination be valued. To be rational calls for something in my relation to others: it calls for an attitude of respect for their rational agency. Following my inclinations calls for nothing at all, not even that I attach any worth to the inclinations I’m following.
There are two ways in which Kant’s theory, thus understood, raises concerns. First, it seems to narrow the scope of morality too far. If the nature of reason only demands respect for reason, does this mean that I can do as will with non-rational creatures? Kant says no, but his reason rests on the indirect effects of our behavior towards nonhuman animals. Put bluntly, if I abuse dogs I’m more likely to abuse people. But this just seems unsatisfying. It seems to me that if there is something wrong with abusing dogs (and I think there is), it’s because of how the abuse affects the dogs.
But perhaps this problem can be overcome by something like the following line of thinking: If I am in the domain of reason and thus called upon to respect reason itself, the demand for consistency "spills onto" inclinations. It would be inconsistent for me to care deeply about my own physical comfort and yet care nothing for the physical comfort of other beings capable of similar experiences. While my inclinations care nothing for such consistency, reason does. And so, once I’m in the domain of reason, the inclinations of other beings, rational or not, become something I should care about on pain of inconsistency.
But this point raises a deeper concern: Why care about reason in the first place? Doesn’t Kant just assume that the domain of reason is the domain we should operate out of?
One way to answer this question is as follows: To ask why we should care about reason is to ask for reasons. And when we ask for reasons, we are placing ourselves in the domain of reason—even if we accept inclinations as reasons. After all, if I offer inclinations as reasons, there is nothing in the domain of inclination that requires that I care one whit about the inclinations that have been thus invoked. Inclinations can be reasons only if we are in the domain of reason. To ask “Why be rational at all?” is thus to put oneself in the domain of reason—whereupon the question answers itself.
This line of argument is not beyond criticism, but I think a deeper concern is that the argument misconstrues the point of asking why we should care about reason. Maybe the real point of this question is not to ask for a reason why we should respect rationality, but to invite us to adopt the perspective of the person who is utterly indifferent to reason. Is there anything we can say that would speak to such a person? Is there any sense in which it is true of that person that they should care about reason even if they don't, and so should do what reason demands? And if not, should that bother us?