Categorical and Hypothetical Imperatives

        If you know anything about Immanuel Kant's (1724-1804) views on ethics, you have probably heard the term "categorical imperative." One of the principal insights of Kant's moral philosophy is that morality applies categorically, without any "ifs ands or buts," so to speak. The demands of morality are such that they apply to all rational creatures without exception."¹ One cannot wiggle out of the rational demands on morality simply because one does not have certain inclinations (or desires). That is the difference between moral "ought" judgments, which Kant calls "categorical imperatives," and other "ought" judgments, which Kant labels "hypothetical imperatives."

        An example of a hypothetical imperative would be that expressed by "You ought to take an umbrella" said to someone about to take a walk on a rainy day. That imperative only applies to one if one has an interest in keeping dry. "Brush your teeth after every meal" is also a hypothetical imperative, since it has rational application only to those beings who wish to avoid tooth decay or bad breath. Other people, if there be any others, have no reason to act in accordance with such a requirement. Even some action that is required for happiness itself (as, for example, perhaps avoiding exposure to mustard gas would be) is, according to Kant, only hypothetically imperative. It is something we must do only because we want to be happy. (Kant believed every human being wants to be happy.) The requirement is still conditional, in that it rests on our wants or inclinations.

        Moral imperatives, by contrast, are categorical, Kant tells us. They apply to us regardless of what inclinations or wants we have. In fact, they apply to all rational beings. Even if some rational extraterrestrial cared nothing about human suffering and had no sympathy for anyone else, it would still be bound by the demands of morality.

        There is one kind of rational being to whom categorical imperatives do not apply--at least not as imperatives. That is what Kant calls "holy wills." These are beings who out of necessity do the moral thing. They always of necessity do what is right. It therefore, Kant maintains, makes no sense to say of them that they "ought" to do something. "Ought" judgments are only directed to beings who might not do what they ought to. No normal human beings have holy wills.² Most who believe in God probably believe that God has a holy will in this sense. Angels, if they were free to rebel against God (and if that was morally wrong), would not.

Consequences Morally Don't Matter

        Kant believed that what matters morally is our will. Whether we actually accomplish what we try to is not relevant to how good we are. It is not what we do that counts--not what we succeed in doing--but what we intend to do (and how hard we try). Whether we did the right thing, thus, does not depend on the consequences, or results, of our action. Our action could have horrible results and still be the morally right thing to do. While many moral theories have a consequentialist component to them,³ Kant's is about as far from consequentialism as one can get.⁴

Acting from duty versus acting merely in accordance with duty

        But for Kant more matters morally than just what someone intends to do. It matters why the person intends to do it. This matters not to whether the person did the right thing in doing something, but to whether there was any "moral worth" in the person doing the action. Not all actions that are in accordance with duty, Kant says, are actions that are done from duty. To put it colloquially, sometimes we do the right thing for the wrong reason. Suppose it would be morally wrong for you to copy someone else's answers during a philosophy test. If you refrain from copying during the exam, but refrain because you are afraid you might get caught and punished, then you are doing the right thing--but you are doing it out of fear, not out of duty. Your action then has no "moral worth," according to Kant. Although Kant does not put it just this way, we might say that you get no moral credit for such actions.⁵ Suppose it is morally right, and something it is our duty to do, to visit widows and orphans. Suppose further that you in fact do this because you enjoy visiting with these people. Kant would say that you are acting merely in accordance with duty, that you are not acting from duty. You are acting from inclination. Now if you were to do the right thing simply because it was the right thing to do (and not because you desired to do it, quite independently, or in order to reap praise or avoid blame from other people), then you would be acting "from" duty and your action would have moral worth, according to Kant.

        So there is a big difference, according to Kant, between doing something because you enjoy doing things of that sort, or because it is in your self-interest to do so, and doing something simply because it is your moral duty to do so. It is only actions in this final group, actions done simply out of respect for the moral law, that have moral worth, that count toward the moral goodness of a person.

        Nonetheless, a person who does the right thing for some reason other than it being the right thing (say, out of fear, or because of inclination (desire), or from self-interest) still has done the right thing. She or he has acted in accordance with duty. But such a person, Kant might say, deserves no pat on the back. Presumably such a person would have done the same thing even if it had been wrong. It is almost as though it is simply good luck that the person happens to want to do what is right.

Formulations of the Categorical Imperative

        So what is the right thing to do? What is our duty? This is the question the categorical imperative is an answer to. Kant insists that there is only one categorical imperative. This remark is perhaps best understood as the claim that there is only one fundamental categorical imperative, since derivative moral rules would still, it would seem, be imperatives and be categorical. Even so, Kant provides several very different "formulations" of the "one" fundamental categorical imperative. He says:

A few lines later he writes:

Now while these might seem similar enough, Kant complicates matters by providing other formulations that seem downright different. He writes:

There are other formulations too, but I will ignore them here. While most accounts of Kant's ethics focus on the "universal law" formulations (the first two here), I will concentrate first on this last formulation, because I think it provides a clearer picture of Kant's views about what our moral duties are.

Treating People as Means and Treating People as Ends

        What Kant's imperative says, in this formulation, is that we must always treat people as ends. We need to figure out what that means. First, however, allow me to make one emendation to the formulation as it stands. Kant talks of treating "humanity" as an end. This makes it look as if he is talking of human beings and of their human nature, but, to be strictly consistent with his earlier point about the scope of morality, he should be talking instead of the rational nature of rational beings. So if Kant had heard of E.T., the extra terrestrial, he would have insisted that E.T. be treated as an end and not as a means only. In particular, given that E.T. was a rational being, Kant would have objected to scientists treating him as they would a mutant dog and cutting him up simply to further scientific knowledge.

        In order to understand what it means to treat someone as an end, it is perhaps best to determine first what it means to treat someone as a means. This is conceptually unrelated, I think, to the idea of being "mean" to a person. To treat someone as a means is, in effect, to use that person. It is to act toward the person in a way designed to get something or achieve a certain result. Doing is treating the person as a means no matter how noble that result may be. In particular, even if our goal is the happiness of other people, using someone simply as a means to achieve that goal is wrong. So it would be wrong to lock up a dangerous criminal if our sole reason for doing so were to keep other people safe from harm, just as much as it would be wrong to sacrifice a virgin (by throwing her in the volcano) in order to placate the gods.

        We often do something as a means to something else. I put money in a Coke machine in order to get a Coke. I press on the accelerator in order to get my car to go. I turn up the thermostat in order to make it warmer in the house. These are all cases of doing something as a means to some further end or goal.

        Sometimes we even do things to people in order to achieve a further goal. A man may rape a woman in order to enhance his own feelings of power and virility. An athlete may injure a competitor in order to increase her own chances of winning a competition. Such actions are clearly wrong.

        But there are also more benign cases of using people as means. I may ask a person for directions in order to get to my destination. I may pay a clerk in order to own a magazine. I may act friendly toward a person in order to get a positive recommendation from her. These are all cases of treating people as means. The more you think about it, the more you realize we all treat people as means practically every day.

        Does Kant say we should not treat people as means? No, he does not. He says we should not treat them merely as means. And what is it to treat someone merely as a means? It is to treat the person as a means without at the same time treating that person as an "end."

        So now we must confront the question: What is it to treat a person as an end? If "end" means "goal," as it often does when we speak of "means to an end," then we might think that it is to treat someone as a goal. Yet it is hard to figure out what that means. Actually, when speaking of people as ends Kant often uses the phrase "end in itself." The goals we have are our ends, but they are not, presumably, ends in themselves. They are ends for us. Let us then look to what Kant says about ends in themselves:

What this seems to be saying is that for me to treat someone as an end in herself or himself, is for me to treat that person's ends as I treat my ends. It is to respect the needs and interests of another as being as legitimate as my own.

        Let us apply this to common instances of human interaction to see what it amounts to. Suppose I give the farmer at the market two dollars and she gives me a bag of potatoes in return. As we have already seen, I am treating the farmer as a means to an end. Do I treat her also as an end? Well, it depends on my behavior toward her. What are the ends of the farmer? Well, she wants to earn some money. She also presumably wants to be treated with respect and to have a pleasant day. If in my interaction with the farmer I work toward these goals as well as my own goals of getting potatoes and perhaps not spending more time or money in doing so than I need to, then I treat her as an end in herself. I might make small talk with her, comment on how fine her potatoes look, offer to pay her an acceptable price, etc. If, on the other hand, I treat the farmer as I would a vending machine, and simply push the money toward her and grab the sack, without regard to her interests, then I am treating her as means only.⁹

        Obviously there are more serious instances of violating the categorical imperative. Rape, theft, murder and blackmail all generally involving treating a person merely as a means to an end and not treating the person as something inherently valuable. These are things we should not do. Kant's "Persons as Ends" formulation of the categorical imperative is an attempt to get at the core reason why these actions are wrong.

Kinds of Moral Duties

        Treating people kindly seems bland enough, but one might wonder whether Kant's formulation of the categorical imperative provides any further guidance in our daily affairs. Kant thought it does, and he worked through four examples of immoral action attempting to show in each case how his various formulations of the categorical imperative lead to the conclusion that these actions are morally prohibited.

        He does not choose four examples randomly, but categorizes duties into four groups by making two distinctions that cut across each other. (See the diagram below.) He distinguishes duties towards oneself from duties toward others, and he distinguishes what he calls "perfect" duties from "imperfect" duties.¹⁰ The first distinction is the clearer one, so let us start with that.

       Duties Toward Self and Others

        Kant felt that in addition to duties that we have to other people, we have duties we owe ourselves. We have, for example, a duty not to commit suicide. Whether or not we would harm others (e.g., our family) or sin against God by taking our own life, Kant seems to think suicide is wrong because it is violating a duty we have towards ourselves. Another example Kant offers of a duty towards ourselves is the duty to develop our talents. If you let yourself go to pot, so to speak, and make no effort to develop your potential (whether it be in art, sports, music, philosophy, or human relations), you are violating a duty you have toward yourself.

        We also have duties toward others. It is wrong to steal. It is wrong to tell lies. It is wrong to kill people. Those actions all involve violation of duties we have toward others.

        Duties toward ourselves and duties toward others are both still our duties, i.e., duties we have. The difference lies not in who has the duty, but generally in who the intended beneficiary of the duty is.

       Perfect and Imperfect Duties

        The distinction Kant uses between what he calls perfect and imperfect duties is perhaps not as clear.¹¹ Officially, imperfect duties are those that admit of exception and perfect duties those that do not. What this means is that imperfect duties are ones in which we have a choice about how (not whether) to fulfill them. We could also think of imperfect duties as ones which one could never completely (i.e., "perfectly") fulfill. Consider, for example, the duty to help people in need. Kant regards this as an imperfect duty. We indeed, Kant says, have a duty to help people in need, but that does not mean that we have to help every person with every need. Choices must be made. It is left for us to decide how we are going to fulfill this duty. Some people give lots of money to charity. Others volunteer lots of their time. In each case choices are made. One cannot give all of one's money to all charities. Even if you were to give all of your money away, there are charities and causes that you would not have given to. If we give all our money to United Way, we cannot also give it all to the American Heart Association or to Oxfam. If we spend all our free time tutoring disadvantaged children in Memphis, we cannot use that time also to build homes for the homeless in Detroit. We do not, morally, have a choice of whether to fulfill an imperfect duty, but we have a choice of how to fulfill it.

        Perfect duties, in contrast, admit of no exception and can be fulfilled completely. I have a duty, let us suppose, not to kill people. There is no choice on my part, left open by morality, concerning how to fulfill that duty. I can perfectly fulfill that duty by never killing anyone.

        Think of it this way:¹² At the end of our life here when we are reflecting on whether we have behaved morally, there may be some questions that can be answered by a "yes" or a "no." Did you ever lie? Did you ever kill? Did you ever steal?¹³ There are other questions that have instead to be answered in terms of how much. The question, for instance, is not whether you helped others or not, but how much you helped others. These latter questions concern imperfect duties, the former perfect duties.

Kant's Four Examples

        To illustrate the application of the categorical imperative and its implications for the moral requirements we are under, Kant provides four examples of moral duties and shows how each of these are supported by the categorical imperative. He chooses one example of a perfect duty to self, one of an imperfect duty to self, one of a perfect duty to others, and one of an imperfect duty to others. (I will address these in an order different from that Kant uses.) The idea is to show how in each case the fundamental categorical imperative generates more specific duties.

        Kant's example of a perfect duty to others is the requirement that we must not promise something to someone if we have no intention of doing what we promise. You may remember Wimpy from the old Popeye cartoons. His famous and frequent line was "I'll gladly pay you Tuesday for a hamburger today." I have not gone back to study these cartoons, but my impression always was that Wimpy had no intention of paying the debt on Tuesday--or any other time--and was merely making this statement in order to get a hamburger today. Of this sort of example Kant says "the man who intends to make a false promise will immediately see that he intends to make use of another man merely as a means to an end which the latter does not likewise hold."¹⁴ This is a clear case of using someone merely as a means. The person to whom Wimpy directs the request is being used (or so Wimpy intends) to get a hamburger. If Wimpy has no intention to repay, then the interests of the other person (not to be out the price of a hamburger, for instance) are being ignored and Wimpy is not treating the other as an end in himself. So Wimpy is treating the person not as an end but as a means only. His behavior is morally wrong, according to Kant's categorical imperative.

        Consider now imperfect duties to others. Kant's example is this:

Kant says that ignoring the needs of others is wrong. We morally ought to help others in need. This purported duty toward others is an imperfect one, as I've explained. It is in regard to this case that Kant writes: "the ends of any subject who is an end in himself [as all rational beings are] must as far as possible be my ends also ..."¹⁶ To ignore the needs of others is to act as though their needs do not count as worthy of being met. And this, according to Kant, is not to treat them fully as ends in themselves. This case does not seem to involve the issue of treating people as means, but the injunction to treat them as ends rules out, Kant claims, such self-centered behavior.

        Let us turn now to duties to self. Kant's example of a perfect duty to self is the moral obligation we have not to kill ourselves. Suicide, Kant insists, is morally wrong. Unfortunately, Kant's arguments here may not be as compelling. He imagines someone who sees only pain and suffering ahead and considers ending his life. Of that person he writes: "If he destroys himself in order to escape from a difficult situation, then he is making use of his person [i.e., himself] merely as a means so as to maintain a tolerable condition till the end of his life."¹⁷ Kant is saying here that those who commit suicide are using themselves as a means. Recall that in formulating the categorical imperative Kant insisted that it was wrong to treat personhood "whether in your own person or in the person of another" as a means only. The categorical imperative, thus, according to Kant, rules out suicide.

        Finally let us consider Kant's example of an imperfect duty to self. Kant insists that we must not neglect the cultivation of our talents. If there are ways in which we could improve ourselves (becoming better singers or poets or mathematicians), we are morally obligated to do so. Now, again, this is an imperfect duty, and as such it leaves us some leeway to determine which of our talents we will develop. We cannot put all our time into practicing the oboe while also putting all our time into developing our philosophical skills. Note also that Kant regards this duty to cultivate our talents as a duty to ourselves. The fact that others might benefit from our talents is not to the issue.¹⁸ How does this involve failing to treat personhood as an end in itself? Kant writes: "it is not enough that the action does not conflict with humanity in our own person as an end in itself; the action must also harmonize with this end."¹⁹ The idea is that to treat someone as an end, one must work to further that persons ends and indeed to further that person. So, to treat oneself as an end, one must work to further one's own personhood. Now these are not all the moral duties we have. Kant simply chose a single example from each of the four varieties of duty for illustration. The examples do serve, however, to clarify the categorical imperative and to help us see how it can be applied in concrete cases.

Maxims

        I return now to Kant's earlier formulations of the categorical imperative. According to them we must act only on those maxims which we can at the same time will to be universal laws. There is nothing here about ends and means. Still a lot needs explaining. Let me begin with the notion of a "maxim."

        Kant seemed to think that all our intentional actions are done from maxims. Now by "maxim" he does not mean some "proverbial saying," as your dictionary might suggest. He means a personal principle of action. Sometimes, at least, when we act, we are acting from some self-imposed rule or principle. When I get behind the driver's seat of a car, I refuse to begin driving until everyone in the car is wearing a seat belt. This is my principle: "Don't drive unless everyone is belted in." I also don't have breakfast in the morning until I have taken a shower. These principles Kant calls "maxims." And Kant apparently thought that whenever we consciously choose to do something, we are acting from some personal maxim.

        Now Kant's categorical imperative, at least in its initial formulations, specifies which maxims must not be acted from. Thus it does not directly apply to our actions. The categorical imperative in these formulations does not directly say what actions we must do and what actions we must not do. It does tell us what maxims we must not act from. One apparent consequence of this is that two people may do the "same thing," and one of them be doing right while the other is doing wrong, because they are acting from different maxims. (The mugger puts a knife into a man from the maxim "Kill anyone who doesn't hand over his wallet." The surgeon puts a knife into a person from the maxim "Remove cancers from people when they have given permission to do so and when doing so will improve their lives.") Another consequence, even apart from this, is that this formulation of the categorical imperative does not tell us what we must do; it at most tells us what we must not do.

Universalizing

        To be morally permissible, the categorical imperative tells us, a maxim must be such that an agent can will it to be a universal law. For me to act morally from a certain maxim, it must be possible for me to will (i.e., want, and indeed insist) that everyone act from that maxim. Kant is serious here about talking simply of the possibility of so willing.²⁰ Suppose Flo acts from a personal principle that whenever she meets someone wearing glasses, she steps (or at least attempts to) on his or her toes. Can she will this maxim to be a universal law? First imagine what it would be like if everyone were to do this. It might be a slightly crazy world and people might get miffed at having their toes stepped on. Perhaps contact lenses would be more popular. It would certainly be a possible world. I see no reason why Flo could not will this to be.

        Clearly Kant thinks that some maxims cannot be willed to be universal. Imagine Pete, who decides not to do his own homework, but to copy the homework from another student. Can he will his maxim to be a universal law? The law would be: everyone is to copy homework answers from someone else. But now it can be seen that this cannot be universal law. It is not just that this would be a bad state of affairs. It cannot be a state of affairs at all. It cannot be that everyone copies all answers.²¹ So, Pete cannot will that his maxim be a universal law. So, according to Kant, it would be wrong for Pete to act from this maxim.

        For another example consider counterfeiting money. Suppose Bill, who is skilled at duplicating currency, decides that instead of paying "real" cash for things, he will instead print up and pay with counterfeit money. Can he universalize his maxim? Can he will that everyone pay with counterfeit money? I suspect Kant would answer in the negative.

        You might want to go back to the four examples Kant offered and see whether in each of those cases there is some maxim which cannot be willed to be a universal law. Kant thinks so, and he discusses the four examples in an attempt to make that clear, but I will bypass that discussion here.²²

        Some ethical philosophers are convinced that Kant's insistence that maxims, to be morally permissible, must be able to be universalized, gets at the core of what morality is. It is saying in effect that we may not (morally) treat ourselves as an exception. It will not do to insist that everyone else refrain from stealing, but allow ourselves to do it. The essence of morality, it might be said, is to treat each person equally. In utilitarianism this is found in the notion of each person's happiness counting just as much as any other's.²³ In the "end" formulation of Kant's categorical imperative this notion of impartiality shows itself in how we must treat each person as an end and never as a means only. Thus we cannot give ourselves (or any specific others) preferential treatment in this regard. Finally, the impartiality notion comes up in the Kantian universalizability requirement. We must not act from maxims which we would not allow everyone else also to act from.

Objections to Kantianism

       Universalization

        Most objections to Kant's ethical theory focus on the requirement of universalizability. It is not often argued that this condition is wrong,²⁴ but that it is far too weak a requirement to be of much use in helping us determine our moral duties.

        Take almost any action generally regarded as clearly morally wrong. Even take an example of an action Kant takes to be wrong. In each case one can imagine some maxim which could be universally willed, which would allow the action to be done with moral impunity on Kant's theory. So recall Bill, the counterfeiter. Suppose he acts from the maxim "Let talented counterfeiters print up counterfeit money and pay with that instead with `real' cash." Since not everyone is a talented counterfeiter (but Bill is), this maxim could be willed to be a universal law of nature without contradiction.²⁵ Similarly, Pete could will that the one copy answers from one's neighbor when one doesn't know the answer but one's neighbor has written one down. This, it seems, could be willed to be a universal law of nature.

        So Kant's theory allows these immoral actions to slip through as morally permissible. And one could analogously generate maxims permitting lying, killing, and stealing. If so, Kant's theory is bankrupt as a moral theory.

        I'm not sure how Kant would reply to this charge, but I suspect he would insist that the actual maxims people act from are simply not this devious. The maxims people act from when they do immoral actions are more straightforward and are contradictory if willed to be universal laws of nature.

        But while this charge makes out Kantian theory to be too weak--to allow too much immoral action--the criterion of universalizability can also be objected to on the grounds of being too strong. Suppose Karen chooses not to have sex on the grounds that she does not wish to bring more people into this world.²⁶ Her maxim for this action "don't have sexual intercourse if that will bring about additional human beings" is not universalizable. After all, if everyone (throughout all time) acted from this maxim, there wouldn't be any people in the world to act from the maxim.²⁷ And while Catholics may welcome this argument against all forms of birth control, the argument cuts more deeply than they would like, for the celibate priest is just as guilty, according to this argument, of acting on this maxim that one must not procreate.²⁸ Do we all have a duty not to avoid procreation? If so, what are we doing sitting around reading this?

        One gets the sense that perhaps this universalizability business is just too malleable, too open to distortion.

        Still this line of objection leaves open the other main formulation of the categorical imperative, that involving treating rational beings as ends. That, it seems to me, by itself provides clear moral injunctions which are not susceptible to these arguments against the universalizability criterion.²⁹ So Kant's theory, at least in that expression of it, provides moral guidance. Whether Kant has any argument to offer for whether we have any grounds for accepting the categorical imperative in that formulation is an issue I'll leave aside in this introductory work.

       Dilemmas

        Another objection to Kantianism concerns the possibility of moral dilemmas. A moral dilemma, as defined by moral philosophers, is a situation in which an agent is morally obligated to do some action, and also morally obligated to do some other action, but in which the agent cannot do both. Perhaps I've made two promises, am obligated to fulfill each, and yet I now find out it is simply impossible for me to do so. (I promised to be at my brother's wedding in New York and also to be with my wife as she gives birth in Arkansas and now find that these are happening at the same time.)

        For Kant moral duties are exceptionless, categorical, demands. He therefore insists that there cannot be conflict between them (since then reason would be demanding something of us which we could not do). And yet it seems to many that there simply are cases of moral dilemmas of this sort. So, Kantianism is wrong.

        The literature on moral dilemmas is vast.³⁰ This is not the place to address the issue, but let me point out that (a) it is not obvious that there are any real moral dilemmas (some things which seem to be our duty, simply aren't), and (b) if Kantianism is guilty of disallowing moral dilemmas, so are many other moral theories (such as most forms of utilitarianism), so Kant's theory is in no worse shape than its competitors on this front.