In utilitarianism, what makes an action right or wrong is outside the action; it is the consequences of the action. By contrast, deontological theories put the emphasis on the internal character of the act itself. (1) What makes an action right or wrong for deontologists is the principle inherent in the action. If an action is done from a sense of duty, if the principle of the action can be universalized, then the action is right. For example, if I tell the truth, not just because it is convenient for me to do so but because I recognize that I must respect the other person, then I act from duty and my action is right. If I tell the truth because I fear getting caught or because I believe I will be rewarded for doing so, then my act is not morally worthy.

Deontologists say that individuals are valuable in themselves, not because of their social value. Utilitarianism is criticized because it appears to tolerate sacrificing some people for the sake of others. With the emphasis on maximizing overall happiness, there are no absolute prohibitions on how we treat others. By contrast, deontological theories assert that there are some actions that are always wrong, no matter what the consequences. The best example of this is killing. Even though we can imagine situations in which intentionally killing one person may save the lives of many others, deontologists insist that intentional killing is always wrong. Killing is wrong (even in extreme situations) because it means using the person as a means and does not treat the human being as valuable in and of himself. (Deontologists do often recognize self-defense and other special circumstances as excusing killing, but these are cases when the killing is not exactly intentional.)

At the heart of deontological theory is an idea about what it means to be a person, and this is connected to the idea of moral agency. Charles Fried puts the point as follows:

the substantive contents of the norms of right and wrong express the value of persons, of respect for personality. What we may not do to each other, the things which are wrong, are precisely those forms of personal interaction which deny to our victim the status of a freely choosing, rationally valuing, specially efficacious person, the special status of moral personality. (2)

According to deontologists, the utilitarians go wrong when they fix on happiness as the highest good. Deontologists point out that this cannot be the highest good for humans because if this was what we were meant to achieve, we would have been better designed without minds. That is, if our function as human beings was simply to be happy, blind instinct would have suited us better. The fact that we are rational beings, capable of reasoning about what we want to do and then deciding and acting, suggests that our function must be something other than mere happiness. Humans differ from all other beings in the world insofar as we have the capacity for rationality. The behavior of other things is determined simply by laws of nature. Plants turn toward the sun because of photosynthesis--they don't think and decide which way they will turn. Physical objects fall by the law of gravity. Water boils when it reaches a certain temperature. In contrast, human beings are not entirely determined by laws of nature: We have the capacity to legislate for ourselves; we decide how we will behave.

So Kant identifies a fundamental feature of human beings-our capacity for rational decision making. Each of us has this capacity; each of us can make choices, choices about what we will do and what kind of persons we will become. This means that no one else can make these choices for us, and that each of us must recognize this capacity in the other.

Notice that it makes good sense that our rationality is connected with morality, for we could not be moral beings at all unless we had this rational capacity. We do not think of plants or fish or dogs and cats as moral beings precisely because they do not have the capacity to reason about their actions. We are moral beings because we are rational beings, that is, because we have the capacity to give ourselves rules and follow them. We are capable of determining our own behavior, in a way that other beings are not.

Where utilitarians note that all humans seek happiness, deontologists emphasize that humans are creatures with goals who engage in activities directed toward achieving these and they use their rationality to formulate their goals and figure out what kind of life to live. In a sense, deontologists pull back from fixing on any particular value as structuring morality and instead ground morality in the capacity of each individual to organize his or her own life, make choices, and engage in activities to realize his or her self-chosen life plans. What morality requires is that we respect all of these beings as valuable in themselves, and refrain from seeing them or valuing them only insofar as they fit into our own life plans.

Although deontological theories can be formulated in a number of ways, one formulation is particularly important to mention. This is a rule Kant referred to as the categorical imperative. There are actually three versions of it, and the second version, as discussed earlier, goes as follows: Never treat another human being merely as a means but always as an end. This general rule is derived from the idea that persons are moral beings because they are rational, efficacious beings. Because we each have the capacity to think and decide and act for ourselves, we should each be treated with respect, that is with recognition of this capacity.

It is important to note the "merely" in the categorical imperative. Deontologists do not insist that we never use other persons, only that we never "merely" use them. For example, if I own a company and hire employees to work in my company, I might be thought of as using those employees as a means to my end (that is, the success of my business). This, however, is not wrong if the employees agree to work for me and if I pay them a fair wage. I thereby respect their ability to choose for themselves and I respect the value of their labor. What would be wrong would be to take them as slaves and make them work for me, or to pay them so little that they must borrow from me and must remain always in my debt. This would be exploitation. This would show disregard for the value of each person as a "freely choosing, rationally valuing, specially efficacious person."

Case Illustration

Consider a case involving computers. Suppose a professor of philosophy at a major research university, undertakes research on moral attitudes of toward sex and sexual behavior among high school students. Among other things, she interviews hundreds of high school students concerning their attitudes and behavior. She knows that the students will never give her information unless she guarantees them confidentiality, so before doing the interviews, she promises each student that no one (other than her) will have access to the raw interview data and that all publishable results will be reported in statistical form. Thus, it will be impossible to identify information from individual students.

Suppose, however, that it is now time to analyze the interview data, and she realizes that it will be much easier to put the data into a computer and use the computer to do the analysis. She will have to code the data so that names do not appear in the database and she will have to make an effort to secure the data. She has hired graduate students to assist her and she wonders whether

she should let the graduate students code and process the raw data.

At first glance it would seem that from a consequentialist point of view, the professor should weigh the good that will come from the research, and from doing it quickly (on a computer), against the possible harm to herself and her subjects if information is leaked. The research may provide important information to people working with youth and may help her career to prosper. Still, the advantage of doing it quickly may be slight. She must worry about the effect of a leak of information on one of the students. Also, since she has explicitly promised confidentiality to the student-subjects, she has to worry about the effects on her credibility as a social researcher and on social science research in general if she breaks her promise. That is, her subjects and many others may be reluctant in the future to trust her and other social scientists if she breaks the promise and they find out.

Thus, it would seem from a consequentialist point of view that the professor should not violate her promise of confidentiality. Fortunately, there are ways to code data before putting it into the computer or turning it over to her graduate students. She must, however, do the coding herself and keep the key to individual names strictly to herself.

This is how a consequentialist might analyze the situation. A deontologist would probably not come to a very different conclusion, but the reasoning would be quite different. The philosopher is doing a study that will advance human knowledge and, no doubt, further her career. There is nothing wrong with this as long as it does not violate the categorical imperative. The question here is, is she treating her subjects merely as means to knowledge and her own advancement, or is she truly recognizing those subjects as ends in themselves? The categorical imperative requires that the philosopher seek the permission of each subject before she gathers data on their sex lives. In seeking the permission of each subject, she respects each as an individual who has her or his own desires, needs, and plans, and the ability to make her or his own choices about what to do or not to do. If, however, the philosopher were to ignore her promise of confidentiality, she would not be treating each subject as an end. After all, each student made a choice based on her pledge of confidentiality, and she must acknowledge and respect that choice. Thus, out of respect for the subjects, the philosopher must code the data herself and maintain confidentiality.

The two theories do not, then, come to very different conclusions in this case. However, the analysis is very different; that is, the reasons given for coming to a conclusion are very different. In other cases, these theories lead to dramatically different conclusions.

Adapted from Chapter 4, Computer Ethics, by Deborah Johnson.

End Notes

1 The term "deontology' derives from the Greek words deon (duty) and logos (science). Etymologically, then, deontology means the science of duty. According to the Encyclopedia of Philosophy, its current usage is more specific, referring to an ethical theory which holds that "at least some acts are morally obligatory regardless of their consequences for human weal or woe" (Paul Edwards, ed., The Encyclopedia of Philosophy (New York: Macmillan, 1967).

2 Charles Fried, Right and Wrong (Cambridge, Mass.: Harvard University Press, 1978), pp 28-29

Ethical Analysis
Updated 27 February 2002