James W. Cornman, Keith Lehrer, George S. Pappas, Philosophical Problems and Arguments: An Introduction, 3d edition, 1982. [1st ed., 1968; 2d ed., 1974]

Chapter SIX
The Problem of Justifying an Ethical Standard


James W. Cornman

There is one kind of problem that continually confronts most people. At one time or another we are faced with deciding what we ought to do. We also often wonder whether we have done the right thing, and we accuse others, as well as ourselves, of not doing what ought to be done. In many of these cases we are making moral or ethical judgments, judging the moral worth of actions we or others have done or are thinking of doing. Think back about some of your past actions. Probably you can find some actions you think you shouldn't have done. Perhaps it was lying about your age to get served liquor in a bar, or taking a glance at the test paper next to you in an examination, or indefinitely "borrowing" a library book without signing for it. Even now you may be thinking about some course of action in the future, whether to use the fraternity files for a course paper, whether to bury yourself in your work and avoid participating in social action, or whether to ignore an oft-proclaimed principle of your own in order to avoid some physical hardship. Where there is a person who thinks about what she and others have done and are doing, rather than acting without thinking, there we find a person who is faced with making a moral judgment. And with any judgment, when we make it we like to think we made the correct judgment or at least that we are justified in the judgment we make.

How can we justify our moral judgments? When we decide what we ought to do, we would like to base our decisions on sound reasons, although, as in many other areas of human endeavor, we often decide without thinking. Usually when we try to defend our moral decisions and actions we do so by reference to some moral rule or standard, such as "Thou shalt not kill" or "Lying and cheating are wrong." That is, we often justify a claim that a particular action is right or wrong by reference to some ethical rule or standard which applies to the action. It is obvious, however, that we cannot show an action to be right or wrong by appeal to a standard unless we have appealed to the correct standard. For example, the attempt to absolve a white person of the killing of a black person by appealing to the standard that no white person ought to be convicted of a crime when the victim is black, may convince some people, but it fails to justify the act morally, because the standard appealed to is incorrect. On the other hand, attempting to eliminate acts of capital punishment by appealing to the standard that no person, or group of people, has the right to take the life of another person surely has some force to it. Those who defend capital punishment usually will not attack the standard but try to show that it must be modified to account for certain exceptions. An important part of justifying a particular moral decision, then, is basing it upon the correct ethical standard.

If we can find some way to justify a standard or a group of standards, then the only other particularly moral task we have left -- probably the most difficult task of all -- is the task of applying the standards throughout our lives. The second task faces us all, including philosophers, who are in no better positions to achieve success than anyone else. Philosophers are particularly suited to the first task, however, because they are centrally interested in and uniquely trained for critical investigations of the arguments people propose to justify their actions and beliefs. In this chapter we shall examine the various leading theories which propose and defend particular moral standards, and we shall attempt a philosophical examination of each, with the hope that we can draw a justified conclusion about correct ethical standards.


Before we move on to consider ethical theories (that is, theories that propose ethical standards) two points should be emphasized. The first is that we are interested in a standard that can be used to prescribe and evaluate particular courses of action, that is, a standard that can be used to prescribe what we ought to do and evaluate what we have done. We are, then, not interested in a standard that is to be used to evaluate morally the people who perform actions, but rather in a standard for evaluating the actions people perform. We certainly use both kinds of standards, for we not only decide that what someone did was right or wrong, but we also praise or blame the person for doing it and sometimes judge him to be moral or immoral. Both kinds of standards are important, but they are different. It seems essential in morally evaluating a person for what he does, that we consider his motives, his beliefs, and the particular circumstances surrounding his decision to act, but it is not clear that any of these are relevant to the evaluation of his action. For example, many people have claimed that it was wrong to drop the first atomic bomb on Hiroshima, and they have consequently blamed President Truman for ordering the bomb to be dropped. However, these are two quite separate issues. We might argue that it was morally wrong to drop the first bomb on a city because a less populated site might have been equally effective. Here we decide the issue without considering President Truman's motives, beliefs, and the pressures under which he made the decision. But in deciding whether to blame the President we must consider his motive, his beliefs about the war and whether they were reasonable, and the external and internal forces playing upon the one person who had to make the decision. It may be, then, that the action he took was wrong, but he should not be blamed for it. Similarly, someone might do something which, contrary to what he intended, turns out to be right. In such a case the action may be right but the person blameworthy. Consequently, we should remember to distinguish between these two kinds of standards, because we are considering only standards for evaluating moral actions and because failure to distinguish the two has often led to unjust accusations of blame and unnecessary feelings of guilt. There are many actions which are wrong but which reflect no blame or guilt on the doer. Understanding rather than blame is often appropriate.


The second point concerns the means we shall use to evaluate critically the various ethical theories. In general we shall proceed as we did in Chapter Four, in which we considered various mind-body theories. That is, we shall try to elaborate each position clearly, consider the problems confronting each, and then decide which position is least troubled by serious objections. We must, then, elaborate and evaluate the most serious objections to each theory. We shall find, for example, that the standards proposed by some theories do not apply to all situations, that other standards result in unresolvable moral conflicts when applied to certain situations, and that still others prescribe morally repugnant courses of action in certain situations. This last point is very important and deserves further comment.

We shall claim that someone has some reason to reject a standard which is clearly contrary to what, in an uncritical way, he or she feels certain is correct. There are a number of things that must be said in clarificatiori of this idea. First, it is not enough for a person to be sure about whether or not what the standard prescribes is correct; he must be quite sure, or certain, that what the standard prescribes is not correct. Secondly, this sort of situation may occur in a number of different ways. For example, a given ethical standard might dictate that a specific action is wrong, when a person feels quite certain that the action is right. Of course, the reverse of this may also occur. In addition, a standard might dictate that a specific action is obligatory when a person feels certain that the action is morally prohibited. Similarly, a standard might tell us that an action is morally permissible, that is, neither obligatory nor prohibited, when a person feels certain, that the action is obligatory, or feels certain that it is forbidden. It is clear, too, that other conflicts of this sort between what an ethical standard prescribes and what a person feels certain is correct in a specific situation might arise. The term 'incorrect' was just used to cover each of these possibilities.

Imagine that a person tries to test an ethical standard by seeing whether it agrees what it prescribes with what he or she feels certain is morally correct. Suppose, too, that this person finds that there is considerable agreement on the matter. From this alone it will not follow that the ethical standard is acceptable for that person. There may be all sorts of other things wrong with the standard. It does not even follow that the person has some reason to accept the standard. The problem is that a person might have inconsistent moral beliefs. Few people have self-consciously examined the whole spectrum of their moral opinions and decisions, and it is quite likely that many people are inconsistent. They decide differently at different times, even under similar circumstances, and especially when the action involves someone they love or hate. Where someone finds that he has inconsistent beliefs, then, even if he believes one of them quite strongly, he should not use it to test a standard. Consequently, a person should rely on his own intuitive opinions of what is right and wrong, obligatory, permissible, or forbidden, only when he feels quite certain of those opinions, and none of his other other beliefs are incompatible with those opinions.

It can be objected, however, that it is a mistake to rely on this intuitive test of ethical standards, because people's ethical opinions, even their strongly held opinions, differ extremely widely about particular cases. For example, many Jews find to be obviously morally repugnant certain actions which many Nazis found quite acceptable. Also there clearly are equally deeply felt disagreements between many pacifists and many military rulers. It surely is mistaken, according to this objection, to rely on a method of evaluation that allows Nazis and some of the most callous military rulers to be justified in holding a standard because they do not find that it prescribes anything morally repugnant, when so many other people find it clearly abhorrent.

There is some force in this objection, though not quite as much as one might initially expect. Reliance on intuitive opinions of different people as just described will not by itself lead to different standards being justified for different people. At most, what will follow is that the way is left open for different people to each have some reason for accepting different standards, Nevertheless, it must be conceded that reliance, in part, on different people's intuitive opinions does allow that one person has some reason to accept ethical standard S1 and another person has some reason to accept S2, even when S1 and S2 are not only different but also in conflict with one another. We think, however, that this much relativity of reasons (for different ethical standards) is inevitable. The critical questions are whether the proposed method or test sanctions justification for a person of an ethical standard that is clearly mistaken, and whether the test allows for different standards to be justified for different people. There are three reasons why we believe it is plausible to expect that the proposed method will have neither of these consequences.

The first is that this intuitive test is but one of several tests or conditions for a satisfactory ethical standard. Many standards that will meet this test for a particular person will fail because they do not meet the other conditions. Part of what we will do, as we examine various proposed ethical standards throughout the chapter, is to try to uncover these important other conditions for a satisfactory standard. We will do this primarily by examining the reasons we found for rejecting unsatisfactory proposals. By the time we have finished, it is hoped that we will have found not only a satisfactory theory, but also the conditions and tests it has met in proving to be satisfactory.

The second reason is that we do not even expect widespread divergences among standards that meet the intuitive test. A standard is not shown to meet this test for someone if he finds that he is not bothered by any of the actions it prescribed that most people find morally repugnant. He must find it meets this test in a wide variety of other cases as well. For example, many Nazis would find it morally repugnant for anyone to put loyal Nazis in gas chambers. But in certain conditions in different countries, such actions might well be prescribed by the same standard that he otherwise finds acceptable. Thus a person is not to select in a biased way the cases he uses to test a standard. He must examine a wide range of possible as well as actual cases to see if it prescribes anything he feels quite certain is mistaken. It is only once he has done this that he can justify that a standard meets this particular test. And once he has done this, a person will find that many fewer standards meet the test than he might have expected.

A third and final reason is that we predict that there will be widespread agreement among different people that certain actions are morally repugnant, say, or morally right. The Nazi example used earlier on p. 264 may be used again here. We confidently predict that many people, indeed the great majority of people, will agree that Nazi torture and execution of millions of innocent people was morally wrong. Hence, if a given ethical standard were to allow such actions as permissible, the case for rejecting such a standard would be based, at least partially, on the fact that the great majority of people are quite certain that such actions are morally repugnant. Needless to say, here, we are not claiming that a decision to reject or accept an ethical standard should be based on majority rule.

We may sum up this discussion of our method by saying that we shall be relying in part on two rules or tests which may be expressed as follows:

  1. If a person feels certain that a specific action is morally incorrect, and this belief is not inconsistent with any of his other beliefs, and there is an ethical standard which dictates that this action is morally correct, then the person has some reason to reject the ethical standard.
  2. If a person feels certain that a large number of actions are morally correct, and none of these beliefs is inconsistent with any of his other beliefs, and he has not been biased in choosing these actions for consideration, and he finds that an ethical standard agrees in all these cases with his beliefs, then the person has some reason to accept the ethical standard.
These are both complex rules but, we think, they are tolerably clear. Notice that they speak of having some reason to accept or reject an ethical standard. Thus, it is not being claimed that if an ethical standard fails to measure up in the way described in (1), or does measure up in the manner described in (2), then one has conclusive evidence against or in favor of the standard. One would have simply one piece of relevant evidence, positive or negative depending on the particular case.


Much of our ethical training and learning takes place in a religious context. Indeed ethics seems to be an essential part of religion. In both the Old and New Testament and in most other religious documents, such as the Koran, there are ethical teachings. In the Old Testament the Ten Commandments are central and in the New Testament we have, among others, the teachings of the Sermon on the Mount. It is natural, then, to associate ethics and morality with religion, so that it is also natural to look to religion for the ethical standards we can use to prescribe and evaluate our actions. And if we think back about the discussion of God in Chapter Five, we might derive a standard from the discussion of the goodness of the supreme being. That is, we might propose that the correct ethical standard is the following:

Whatever God wills is what ought to be done.
If this is the correct ethical standard then whenever we are deciding what ought to be done or what ought to have been done we should base our decision on what God wills. What we must do to make this standard applicable to specific situations is to find some way to discover what God would will in the situation. There are two ways to discover this. First, God might reveal his will to us by directly communicating with us, or he might reveal his will to someone else who relays it to the rest of us. For most of us, if God's will is revealed to us at all, it is only indirectly, with someone else as the intermediary. Consequently, if we are to apply the theological standard on the basis of God's will as indirectly revealed, we must be able to justify some particular claim about what God wills, for example, the Ten Commandments. But we have already seen in Chapter Five how difficult it is to provide grounds for thinking that any revelation of God's will has occurred, let alone show that any one particular claim is correct.1

Objection: Must Justify Religious Claims by Ethical Claims

Let us assume someone claims that what God wills is that people obey the Ten Commandments. We now have an ethical standard that we can apply to particular situations. But how are we to justify the claim that this is the correct criterion? We cannot do so merely by claiming that God revealed the commandments to Moses, because we must justify the claim that it was God who gave them to Moses. Consider what we would do if we read that Moses had returned with such commandments as "Make love to thy neighbor's wife," "Steal thy neighbor's goods," and "Take advantage of thy parents." We would decide that whatever was revealed to Moses, it was not the will of God, because these are immoral commandments. We do not justify that something is moral by showing that it expresses God's will, because the only available way to evaluate conflicting claims about what God wills is by finding which one is in accordance with what is moral. Thus we must use ethical claims to justify religious claims rather than ground ethics upon the claims of some religion.2

This is not to deny that religion is for many people the psychological basis of ethics. It may be, then, that religion has an important psychological relationship to ethics. Nor is this to deny that God has willed or prescribed certain moral commandments. Everything just discussed can be accepted by someone who believes that whatever God wills should be done. Furthermore, nothing said so far gives any reason to think that the Ten Commandments do not express the revealed word of God. They might. If they do, they will also express at least part of a correct ethical standard. The only claim made here is that we cannot justify that they or any other ethical standards are correct by appealing to the pronouncements of some particular religion, because we must justify that these pronouncements express the revealed word of God by showing that they are the correct moral pronouncements. Because our task here is to find and justify some ethical standard, we cannot stop with the pronouncements of some religion even if they are correct. We must find some way to show that they are correct, and this we cannot do by appealing to the religion itself.

We are in no better position if we try to ground an ethical standard on direct revelation. It may be that some day you will have a religious experience in which certain commands are issued to you. You may, as others have after similar experiences, uncritically accept these as revealing the word of God and proclaim them to all. But we are here interested not in what you might do, but in whether you would be justified in claiming that you had heard the word of God. That you have received these commands in a very strange and unique way is not enough. There are many cases of a person who has followed his "voices" and has committed ghastly crimes. In such cases we usually think that the "voices" are the result of psychological disturbances. Furthermore, it is possible that not only God, but also the devil reveals his will to human beings. Consequently, you could justify your claim that you had heard the word of God only if you could provide a reason to think that the word expressed a command of God rather than one of the devil. You cannot do this by appealing to your religious experience. Thus in the case of direct revelation as in the case of indirect revelation, justification of an ethical claim cannot be based on a religious claim.

The conclusion we have reached is that although a religion may help us psychologically to decide what to do, it cannot help us justify what we decide to do. The justification of our ethical standards and thereby our actions proceeds independently of appeals to religion. Because of this, the existence of God is irrelevant to the justification of ethical standards. Thus those who find they can no longer believe in God are not forced to the conclusion that nothing is right or wrong. There is nothing inconsistent in holding some particular ethical standard, and also believing that God does not exist. As brought out earlier, and as we shall see, the critical evaluation and justification of ethical standards is carried on without reference to religion.

Nevertheless there is a widespread view that if there is no God, then nothing is moral or immoral, nothing is right or wrong. This is the view that if anything is right and anything is wrong, it has been decreed so by God. Notice that this is a different claim from the one we previously examined. The previous claim is

If something is willed by God, then it ought to be done,
but the present claim is
If something is the right thing to do (ought to be done), then it is willed (proclaimed or commanded) by God.
Although the first statement is acceptable the second is surely debatable. For one thing, there is no reason to think that all right actions are willed or commanded by God, because there is no reason to think that God issues commands covering every moral situation. It may be that if in a certain situation a particular action is right, then if God were to command some action in that situation he would command that action. But this claim does not lead to the conclusion that if there is no God, then nothing is right or wrong. Secondly, it is at least possible that we shall be able to justify some ethical standard as correct, and because, as we have seen, such a justification does not require reference to God, there is no reason to think that the correct ethical standard must issue from God. However, no matter what causes them to do it many people talk of the breakdown of morality and the destruction of ethical standards, which they often blame on the decline of religion. The result of this, they claim, is that morality is relative, so that nothing is right or wrong, and what it is right for me to do is merely what I want to do. Although it is not unusual to hear such a claim, the claim itself is quite unusual, because it embodies three distinctly different ethical positions -- ethical relativism, ethical nihilism, and ethical egoism. In one way or another these positions are prominent among the views about ethics expressed today. Consequently, each deserves individual attention here.


Ethical relativism seems to be expressed by the frequent claim:

What's right for you is not always right for me.
And because this claim seems to be true many people become convinced of ethical relativism. But this is an ambiguous claim and those interpretations of it which are readily acceptable are not those which imply ethical relativism. One interpretation of 'What's right for you is not always right for me' is the following:
The right action for you is not always the right action for me.
This interpretation is often true because two people are often quite different, but it does not imply ethical relativism. For example, if you are a good swimmer and I cannot swim, then in the same situation where each sees a child drowning, it is right for you to swim to help the child, but right for me to run off for help. But although what each of us ought to do in this same situation differs, it is still true that both of us should do our best to help the child. There is nothing relative about this.

Action Relativism Versus Standard Relativism

What we must do to avoid confusion here is to distinguish between the relativism of ethical actions and the relativism of ethical standards.

Action Relativism: Actions are in some situations right and in some situations wrong.

Standard Relativism: Ethical standards are in some situations correct and in some situations incorrect.

We have seen a case of action relativism in the example of the drowning child, but this was not a case of standard relativism. Both you and I applied the same standard, that we ought to do our best to help the child. Thus there can be relativism of right actions without relativism of ethical standards. Consequently, as presently interpreted,
What's right for you is not always right for me.
expresses an action relativism, but because ethical relativism concerns standard relativism and because action relativism does not imply standard relativism, this often true claim does not imply ethical relativism. What may be confusing is that there are some ethical standards which claim that certain acts are always wrong, such as "Thou shalt not kill," so it may seem that what we can call standard.absolutism implies action absolutism. But there are many other cases such as "Honor thy father and mother" where no specific actions are forbidden. Thus those who rebel against action absolutism are not forced to ethical relativism, because the correct ethical standard may allow that whether a specific action is right or wrong depends upon the specific circumstances in which it is done. How we honor our parents at some time depends upon them, us, and the particular circumstances.

Another interpretation of 'What's right for your is not always right for me' is the following:

What you think, is right is not always what I think is right.
On this interpretation the claim is surely true. But all it expresses then is that you and I sometimes disagree about what we think is right and this is quite compatible with standard absolutism. Thus, this interpretation does not lead to ethical relativism. In order to get to ethical relativism we need an interpretation which will make ethical standards relative. Another interpretation which comes closer and is often thought to lead to ethical relativism is the following:
The standard you are justified in accepting as correct is not always the standard I am justified in accepting.
This interpretation, although once again true, does not lead to ethical relativism, because someone can be justified in accepting something as correct when it is not correct. For example, we would certainly agree that someone who believed that the velocity of objects can be indefinitely increased was justified in his belief before Einstein propounded his theory. But we would also claim that although he was justified, his belief is incorrect. Furthermore, as we have seen, it is possible that our method of evaluating ethical standards, which uses each person's strongly held moral convictions as one of several tests, will result in some relativity in the justification of ethical standards. But as with scientific hypotheses, although different people might be justified in accepting different standards, any one of them, indeed all of them, might nevertheless be mistaken. Thus although there is a relativity of justification, this does not imply a relativity of correctness. Even though which beliefs are justified differs as people's knowledge changes, this does not affect which are the true or correct beliefs. Relativity of when a person is justified in accepting a belief or standard does not help the ethical relativist, who requires a relativity of correct standards. The kind of interpretation we need for ethical relativism is one such as the following:
The correct ethical standard for you is not always the correct ethical standard for me.
This interpretation implies ethical relativism because it implies that an ethical standard is correct relative to some situations and incorrect relative to others. However, on this interpretation, the claim is no longer obviously true. We must consider what reasons there might be for accepting it.

Definition of Ethical Relativism

Let us first define ethical relativism as follows:

Ethical Relativism: Different ethical standards are correct for different groups of people.
This definition is stated broadly to allow for various different kinds of ethical relativism. For example, one species of ethical relativism which some sociologists and anthropologists are tempted to accept is cultural relativism which is the theory that whether an ethical standard is correct depends upon the culture or society of the person concerned. There is also class relativism which, having its roots in Marxism, makes the correct standard relative to the economic class of the person. There is also a relativism which appeals to historians, historical relativism, and which makes the correct standards relative to the particular times at which the person lived. None of these species of ethical relativism is any better justified than the general theory. Therefore if we find a reason to reject this general theory we will be justified in rejecting each of its specific versions as well.

Let us begin the examination of ethical relativism by considering two of the main arguments used to justify it, the argument from differing ethical judgments and the argument from different ethical standards.

The Argument from Differing Ethical Judgments

One of the most widely accepted facts relevant to ethics is that there is, has been, and probably always will be widespread disagreement about what is right and what is wrong. It is not merely that the judgments of people of one culture differ greatly from the judgments of people of another culture, nor is it merely that the judgments of people at one stage of history are quite different from those of people at some earlier or later time. We find widely divergent ethical judgments within one culture and at the same time. Surely, this objection states, if over centuries and throughout the world people have continually made widely divergent and often contradictory moral judgments, then it must be that the ethical standards of people differ from place to place and time to time relative to the situations in which the people live. Therefore, according to this argument, correct standards are relative to the situations of the people who apply the standards. That is, we must conclude that ethical relativism is true.

Let us outline this argument so that we can critically evaluate it. It can be stated as follows:

  1. The ethical judgments people make differ greatly, depending on where and when they live.
  2. If the ethical judgments people make differ greatly, then the ethical standards people use differ greatly.
  3. The ethical standards people use differ greatly.
  4. Ethical relativism is true.
There are two objections that can be raised against this argument. First, although premise (1) is acceptable, there is reason to doubt the truth of the second premise. We have already seen that action relativism does not imply standard relativism, and there is also little reason to think that a judgment relativism implies a standard relativism. Indeed, some anthropologists and sociologists who agree with (1) are not at all sure about (2). Many quite divergent judgments can be explained by pointing out that the people concerned have different beliefs about what the facts are rather than different ethical standards. For example, in one society, the people had the custom of killing their parents when they began to grow old. In Western cultures such an act is considered completely immoral. Most of us judge that killing one's parents is wrong, because we employ the standard that we should honor our parents. It certainly seems that we can conclude that the people of this society had no such standard, but this would be wrong. These people believed that each of us spends his afterlife in the physical state in which he dies. Thus to allow someone to grow old and decrepit would not be honoring him. These people did what they thought best for their parents, and thus honored them by helping them obtain immortality in an enjoyable physical states.3 In this example, they and we seem to use the same ethical standard, but because we disagree about the facts of afterlife the judgments we make differ greatly. In such a way many differences of actual judgments can be explained without postulating different ethical standards. Some anthropologists hope to find that certain ethical standards are universally believed to be correct. If this is so, it would certainly make premise (2) highly dubious.

However, even if the divergence of ethical standards is not as great as some people claim, the evidence presently available supports the claim that people often have different beliefs about which ethical standards are correct. Consequently, we can defend (3) interpreted as follows:

3a. The ethical standards believed to be correct by people often differ.
Thus because we can accept (3a), we can also accept (4) if the inference from (3d) to (4) is valid. However, as it stands, the inference is invalid because (3a) is a statement only about what people believe to be correct and (4) is a statement about what is in fact correct. This is the second objection to the argument from differing ethical judgments -- it is invalid because the inference from (3a) to (4) is invalid. What we must do is find a premise which, with (3a), will allow us to infer (4). This takes us to the second argument for ethical relativism.

The Argument from Different Ethical Standards

It may seem to some that although the inference from (3a) to (4) is, strictly speaking, invalid, it resembles the inference from 'Socrates is a person' to 'Socrates is mortal' -- what is missing is an obvious truth such as 'All human beings are mortal.' However, the dubious part of such an enthymematic argument is very often that missing premise. Let us examine the present case by constructing the argument as follows:

3a. The ethical standards believed to be correct by people often differ.
5. If the ethical standards that people believe to be correct often differ, then the ethical standards which are correct often differ for these different people.

4. The ethical standards which are correct are often different for different people, that is, ethical relativism is true.

What we have done is add statement (5) as the missing premise to make the argument valid. Consequently, because we have seen that (3a) is acceptable, we should accept (4) if we can justify (5). Let us consider (5), which is a sentence of the form:
5a. If the x's which people believe to be correct often differ, then the x which is correct often differs for these different people.
When we consider (5a) we see that there are many sentences of that form which are plainly false. For example, many people differ in their beliefs about the world around us but this does not imply that in each case a different belief is correct. If I believe that the correct number of planets is eight and you believe that the correct number is ten, it is not that one number is correct for me and another correct for you. In this case both you and I are wrong, both of our beliefs are incorrect, because there is one and only one correct number of planets and that number is nine. In general, sentences of the for (5a) are false. Furthermore, there is no reason to think that beliefs about ethical standards are relevantly different from those beliefs for which (5a) is false. We have, therefore, reason to conclude that (5) is false.

Because both arguments supporting ethical relativism are unsound, we have found no reason to accept it. Moreover, because it is clearly contrary to our ordinary conception of morality, there is some reason to reject it. When we claim that lying, cheating, and killing are wrong, we do not claim that these prohibitions are derived from standards which correctly apply to some of us but not to others. We think that an ethical standard is either correct or incorrect for one and for all, and because we have found no reason to deny this we can continue to accept it.

Ethical Relativism Defended: A Modified Argument

The case against ethical relativism as it applies to ethical standards certainly seems quite strong and compelling. But we may have overlooked some things that can be said in its favor. In particular, recall our earlier discussion of the method we shall use in this chapter where two principles or tests were endorsed, one concerning reasons for rejecting an ethical standard and the other concerning reasons for accepting a standard. Perhaps we can argue, in line with the latter of these tests, that different people have some grounds for accepting different ethical standards, and thus produce a new argument for ethical relativism. Such an argument might be the following:

3b. The ethical standards that different people often feel certain are correct often differ from one another.

5b. If the ethical standards that different people often feel certain are correct often differ from one another, then these different people have some reasons for accepting different ethical standards.


6. Different people have some reasons for accepting different ethical standards.

Of course, even if statement (6) is established, we would not have established ethical relativism. But the argument could be continued with just one more premise, namely,
7. If different people have some reasons for accepting different ethical standards, then it is reasonable to believe that ethical relativism is correct.


8. It is reasonable to believe that ethical relativism is correct.

There is reason to think that, when properly understood, this argument is sound. The expression 'reasonable to believe' that is used in (7) and (8) must be taken to mean something like 'there is some reason in its favor.' When (7) is understood along these lines, then it is quite plausible. Certainly premise (3b) is true. Premise (5b) might be suspect, however, since after all, people feel certain about all manner of things that are utterly absurd. However, (5b) can be bolstered if we understand it to mean that different people find that different ethical standards agree with what, in an unbiased way, they feel certain are correct actions, and in no case are the beliefs these people have about these actions inconsistent with their other beliefs. Then, according to principle or test (2), stated earlier (see p. 265), each of these people has some reason for these different standards. In other words, we may suppose that (5b) is plausible provided that the reason such people feel certain about different ethical standards is because they have found, in accord with test (2), that these standards are in line with, or agreement with, actions that these same people feel certain are correct. Hence, since everything in this argument is plausible, we have unearthed some positive support for ethical relativism.

Before we conclude that ethical relativism has been established, however, we should pay careful attention to just what this argument shows. Its conclusion asserts that ethical relativism is reasonable to believe and, as we just saw, this means that there is some reason in favor of ethical relativism. Such a conclusion, though, is not at all the same as claiming that ethical relativism is justified. To understand why we need only note that while there may be some reason in favor of ethical relativism, this is perfectly consistent with there also being some reason or reasons against it. And, indeed, this is precisely what we have just noted; ethical relativism runs against our ordinary conception of morality, what we ordinarily suppose about morality in our nonphilosophical moments. Such a factor is by no means a decisive consideration against ethical relativism. However, it is a negative factor that is certainly relevant. And, when this negative factor is balanced with the positive factor noted in the argument just discussed, we see that the negative factor offsets the positive. Hence, ethical relativism has not been justified, despite the acknowledged fact that some reason has been provided in its favor.

The way seems open to search for an ethical standard we can justify as the correct standard for all. Our search, however, may be frustrated in another way. What reason have we for thinking that there is any correct standard of right and wrong either for some or for all? It is true that when we make moral judgments we act on the assumption that there is such a standard, but perhaps nothing is right and nothing is wrong; perhaps there is no correct ethical standard. If this is true, it is foolish to struggle to justify that some standard is correct. We should thus examine the claim of ethical nihilism before embarking on a critical examination of particular ethical standards.


We can define ethical nihilism quite simply as follows:

Ethical Nihilism: There is nothing morally right and nothing morally wrong.

If this position is correct, then because no actions are either right or wrong, it follows at nothing we do is moral and nothing immoral, that everything is permitted and nothing either morally forbidden or obligatory. It also follows that there are no correct ethical standards, because if there were, then the actions they required would be morally obligatory, and the actions they forbad would be morally prohibited. Such a view is quite contrary to our ordinary beliefs. Most of us feel quite certain that some actions are right and some are wrong. Consequently, unless there are cogent reasons for accepting ethical nihilism, we can reject it as we did ethical relativism. Generally, the debate about ethical nihilism does not center directly on the question of whether any particular actions are morally right or wrong, because the issue can best be discussed by reference to ethical standards. If there is good reason to think that some ethical standard is correct, there is good reason to reject ethical nihilism. If there is good reason to think that no ethical standard is correct, there is some reason to doubt at any actions are right or wrong.

It is important to note here that it is possible that actions are right or wrong but that no ethical standard is correct. We might somehow "sense" that particular actions are right just as we visually sense that objects are red. As with seeing, such moral "sensing" would not depend on the existence of any standard. Nevertheless, the ethical nihilist argues his point by trying to show that there are no correct ethical standards, because this is the best way he can defend his position. He generally depends on two main arguments, one of them resembling the second argument for ethical relativism because it is taken from the disagreement about correct ethical standards, and the second derived from the lack of a justification of any ethical standards.

The Argument from Different Ethical Standards

Some people claim that it is wrong to infer ethical relativism from the widespread and enduring divergence of beliefs about which ethical standards are correct. Such a disagreement, which has persisted for centuries everywhere, instead testifies that there really are not any correct ethical standards. Such an argument, as you can probably see, is no better than the corresponding argument for ethical relativism. We can show this by presenting the argument as follows:

  1. The ethical standards which people believe to be correct differ throughout the world and time.
  2. If the ethical standards which people believe to be correct differ throughout the world and time, there are no correct ethical standards.
  3. There are no correct ethical standards.
As might be guessed, premise (2), which is like premise (5) in the corresponding argument for ethical relativism, is highly suspect. Premise (2) is of the form:
(2a). If the x's which people believe to be correct differ, then there is no correct x.
And many sentences of this form are quite clearly false. For example, there are many divergent beliefs about life on distant stars, but this does not imply that none of these beliefs is correct. Some beliefs about life on stars are correct and some are not. One often finds a wide range of different beliefs on a difficult topic, most of which are wrong, but some of which are right. We can, therefore, reject this argument for ethical nihilism as no better than the corresponding argument for ethical relativism.

The Argument from the Lack of Justification

It becomes apparent when critically evaluating the various leading candidates for the correct ethical standard that none has overcome problems important enough to keep it from being justified. Thus, this argument goes, because no ethical standards are justified, all are unjustified and therefore none is the correct standard. We can state the argument as follows:

  1. No ethical standards are justified.
  2. All ethical standards are unjustified.
  3. If all ethical standards are unjustified, then no ethical standards are correct.
  4. No ethical standards are correct (and thus ethical nihilism is reasonable).
There is a certain plausibility to this argument because we can gather evidence to support (1), the immediate inference from (1) to (2) is certainly valid, and (3) appears true. That is, if no possible ethical standard can be justified to be correct so that all are unjustified, it certainly seems reasonable to conclude that none of them is correct. Thus because both premises (1) and (3) seem acceptable and the argument is valid, it seems that we should accept the conclusion. But think about it some more. We have accepted (1) on the grounds that none of the candidates has yet overcome problems, so that none has yet been justified. However, when we supported (3) we did so by talking about what is implied if no candidates can ever be justified. There is a difference between 'not yet justified' and 'cannot be justified,' for the first is compatible with a future justification but the second precludes any possibility of justification. The argument, then, seems to involve an equivocation of the word 'unjustified,' because (2) seems to require one sense of unjustified' and (3) a different sense. Thus the argument invalid as it stands. We can bring out the equivocation and make the argument valid by replacing premise (3) with two other premise's, namely,

3a. If all ethical standards are not yet justified, then all ethical standards are unjustifiable (cannot be justified).


3b. If all ethical standards are unjustifiable (cannot be justified), then no ethical standards are correct.
When we do this it is easy to see that (3a) is false and there is also some doubt about (3b).

In general it is false to state that if we have not yet justified some of a group of alternative claims, then no such claims can be justified. No particular claim about whether there is life on distant stars can now be justified, that is, there is not enough evidence to strongly support any particular claim. But this does not imply that it is not possible that someday one claim will be justified. Thus in this example, as in ethics, if no position has been justified, we need not conclude that no position can be justified. We should reject premise (3a) and with it the argument containing it.

Although rejecting (3b) as well as (3a) is not necessary for our purposes, it is worthwhile to point out that (3b) derives from the claim that there are no correct but unjustifiable assertions. That is, it derives from the position that if an utterance makes true assertion it is at least possible to justify it. But as we saw in Chapter Five it is very hard to establish such a claim. Even if the utterance 'God exists' is made compatible with every possible state of affairs, it has not been shown that the utterance is not a true assertion. Premise (3b), then, although by no means as dubious as (3a), is by no means obviously acceptable either. In any event, because (3a) is dubious we have reason to reject the argument from lack of justification.

We have found no reason to think that ethical nihilism is true, so we have no reason to think that nothing we or anyone else does is morally wrong. Furthermore, because certain actions clearly seem to be wrong and others right, we have some reason to reject ethical nihilism. Consequently, we can cast ethical nihilism aside with ethical relativism.


seems we can begin our attempt to justify some ethical standard as correct. Before we do so, however, there is one other view relevant to our interests which deserves mention, especially because it is often confused with ethical nihilism. It is the view that no ethical standards can be known to be correct, because no opinion is more reasonable than all the others and consequently, no ethical standard is justifiable as the correct one. We are, then, wasting our time trying. This is ethical skepticism, which is different from ethical nihilism in that it merely claims that no standard is known to be correct, or is justifiable as correct, rather than that no standard actually is correct. It is, then, a weaker claim than ethical nihilism. Moreover, it is by no means unreasonable. In fact, it might be claimed that we have good inductive evidence to support ethical skepticism because it seems that, as of now, none of all the ethical standards proposed throughout history has been justified. Such evidence, if correct, makes ethical skepticism more reasonable than not, and if we must accept this evidence and find no counterbalancing evidence, then it seems that the correct position for us to take is that of ethical skepticism. However, if we are to justify such a position we must examine the evidence ourselves. That is, we must critically evaluate the various candidates for a correct ethical standard. Thus although ethical skepticism may be the correct position, we cannot justify it until we have completed the task we have set before us.


Sometimes a person claims that no one has the right to tell him what to do, because he can do whatever he wants. Such a claim sounds like a statement of both ethical nihilism and egoism, and if he adds that anyone else can do what he wants also, it sounds like ethical relativism as well. But we must be careful to separate these three different claims, because taken together they are incompatible. The claim that each person can do what he wants is not a form of ethical relativism because it is a claim that applies to one and all. If the claim is that it is correct for me but not for anyone else to do what he wants, then it would be a form of relativism. But this is not what we are discussing. The claim might be an assertion of ethical nihilism if what is meant is that we are all permitted to do whatever we want because nothing is right and nothing is wrong. But this claim, in denying that there is a correct standard, is incompatible with both ethical relativism, which states that there are several correct standards, and ethical egoism, which claims there is just one, namely:

Ethical Egoism: Each person ought to act to maximize his own good or well-being.
Consequently, although we have cast doubt on both ethical relativism and nihilism, nothing we have said yet casts any doubt on ethical egoism.

Strictly speaking, if someone is expressing an egoistic ethical theory when he says that he can do whatever he wants to do, it is more likely that he is asserting the species of ethical egoism known as egoistic hedonism, because he is talking about what he wants or desires. This brand of egoism often equates what is good with pleasure or happiness:

Egoistic Hedonism: What each person ought to do is to act to maximize his own pleasure or happiness.
Let us interpret this statement of egoistic hedonism as the following:
A person ought to perform an action in a situation if and only if he does it in order to maximize his own pleasure or happiness.
This will give us a standard we can use not only to decide what we ought to do, but also to decide what we have no obligation to do. On this interpretation if I do something to maximize my pleasure then I ought to do it; if it is not something I do to maximize my pleasure then it is not true that I ought to do it (which means that it is morally permissible for me not to do it). Notice that the claim is that this interpretation provides a standard for what we are not obligated to do rather than a standard for what we are obligated not to do. This is an important difference because, for example, although we are not obligated to tie our left shoelaces before the right, this does not mean that we are obligated not to tie the left before the right. The first tells us that we have no moral obligations about the order of tying shoelaces, whereas the second states that we have a moral obligation not to tie them in a certain order, that is, we are forbidden to tie the left lace before the right. Using the standard of hedonistic egoism we find out what we ought not do, what is morally forbidden or prohibited, by finding which actions are the contradictories of those actions we ought to do. For example, if I ought to tell the truth, then I am forbidden to fail to tell the truth, that is, I ought not lie. With these distinctions in mind let us examine the most widespread species of egoism, egoistic hedonism.

Before we begin a critical evaluation of this theory we must be sure of what it implies and what it does not imply, because certain objections to the theory have arisen from misunderstanding it. This is one kind of hedonistic theory and thus proclaims that pleasure is the one thing good in itself. That is, it proclaims that whereas certain things may be good as means to certain other things, pleasure is the one thing good as an end, the one thing to be sought for its own sake. Other things are to be sought only if they are means to pleasure. Thus, medicine is not good as an end, but it is good as a means because it leads to pleasure by helping to cure us of diseases. Pleasure, therefore, is what has been called the summum bonum, or the highest good. Some people have objected to equating the summum bonum with pleasure, because they equate pleasure with such bodily pleasures as those provided by sex, food, and drink. But a hedonist is not committed to such a position, because he can recognize what have been called the "higher" pleasures, such as aesthetic pleasures and the pleasures of contemplation. invention, and artistic creation. Consequently, a hedonist can aim at these "higher" pleasures and thus justify performing the activities which are the means to achieving them.

An egoistic hedonist is interested in doing what maximizes his own pleasure. Many people picture such a person as one who every minute is seeking immediate thrills and excitement without a thought for the future. This, however, is wrong because the amount of pleasure someone gets from an act depends not only on the present pleasures he receives, but also on the longer-range consequences of the act. A hedonist need not be shortsighted, because he can realize that by abstaining from pleasures now he might maximize the pleasure he gets throughout his life. A hedonist who refused a rabies shot after being bitten by a rabid dog because of the present pain of the shots would be a poor hedonist indeed, because the future pain of the disease far outweighs the pain of the shots. A hedonist, therefore, need not be one who lives for the moment and seeks sensual pleasures. He can aim at the intellectual pleasures by carefully planning his daily life, with his sights set on some future goals. Egoistic hedonism, then, is not what it might first appear to be. Nevertheless, it does seem to be contrary to our usual conception of morality, because it seems to condone a kind of selfishness. Consequently, unless there are good reasons for accepting egoistic hedonism, it is a theory that we should reject. J

However, most people who profess egoistic hedonism are not disturbed by the challenge to defend their position, because they base it on a certain theory about the psychological abilities and limitations of human beings, which they think is clearly true. Although there are several versions of this theory, many of them state that whether a person has the ability to do a certain action in a particular situation depends on which of his desires is strongest at that time. And, furthermore, in any situation, a person's strongest desire is always to increase his own pleasure or happiness as much as possible. Thus in any situation a person acts to maximize his own happiness, regardless of anything else. This theory is psychological egoism, and can be defined as follows:

Psychological Egoism: A human being is psychologically able to perform an action if and only if he does it in order to maximize his own pleasure or happiness.4
Notice how psychological egoism differs from ethical egoistic hedonism. The first states conditions for what we have the ability to do, whereas the second states conditions for what we have a moral obligation to do. They are, then, importantly different. The first is purely a factual statement, but the second expresses an ethical standard. Nevertheless, the factual, psychological claim is thought to provide reason for accepting the ethical claim.

The Argument from Psychological Egoism

The argument from psychological egoism can be stated as follows: The only actions a person is psychologically able to perform are those in accordance with his strongest desire, or in other words, those which he does to maximize his own pleasure. But surely we are under an obligation to do something only if we are capable of doing it. That is, we ought to do something only if we can do it. Therefore, the only things we ought to do are things which we do to maximize our own pleasure. Egoistic hedonism is true.

Let us examine this argument by putting it in the following form:

  1. A person has an obligation to do an action only if he is able to do it.
  2. A person is able to do something only if he does it to maximize his own pleasure or happiness.
  3. A person has an obligation to do an action only if he does it to maximize his own pleasure or happiness.
  4. Ethical egoistic hedonism is true.
The first thing to notice about this argument is that the inference from (3) to (4) is invalid because egoistic hedonism states not only that someone has an obligation to do something only if he does it to maximize his own pleasure, but also that he ought to do it if it is something he does to maximize his own pleasure. The second part of the claim is important for our purposes because we are searching for a justifiable standard or criterion for deciding what we ought to do, which means that we want something that is a sufficient rather than a necessary condition for moral obligation. Thus egoistic hedonism cannot be established merely by an appeal to the hedonistic psychological theory. However, because the inference from (1) and (2) to (3) is valid, an egoistic hedonist can, if (1) and (2) are true, establish that we have no obligation to do anything unless we do it to maximize our own pleasure. That is, what this argument can establish, if sound, is that any ethical standard that requires someone to perform actions that he would not do to maximize his own pleasure is an incorrect standard. Consequently, any correct ethical standard would have to prescribe only actions which someone does to maximize his own pleasure, whether or not it also states that he ought to do those actions because they are the actions he does to maximize his own pleasure. The argument, then, although it does not establish egoistic hedonism, does provide us with a way of evaluating those ethical standards which compete with egoistic hedonism. And because it seems clear that most other standards will sometimes prescribe actions which someone would not do to maximize his own pleasure, this argument, if sound, provides a powerful means of eliminating alternative ethical standards, perhaps to the point where only egoistic hedonism remains unscathed. It is important, then, to examine this argument.

Premise (1), often expressed as the dictum that "ought implies can," is a generally accepted principle. Generally we agree that no one has an obligation to do something if it is impossible for him to do it, for example, if he has some disability. Thus in the example of the drowning child used earlier in Chapter Six, I have no obligation to jump into the water to save the child if I cannot swim. And if people blame me for not swimming out to the child instead of running after help, I can absolve myself of blame by telling them I could not swim. Thus I had an obligation to swim to the child only if I could swim. Premise (1), then, is acceptable.

Objection to Psychological Egoism: People Sometimes Act Benevolently

The crucial part of the argument is obviously premise (2), the one derived from psychological egoism. Let us examine it. It might be claimed, however, that a philosopher has no business critically evaluating premise (2), because it is a claim within the domain of the empirical science of psychology. But although it is generally true that philosophers are not competent to evaluate empirical scientific claims, we shall find grounds for thinking that if, as claimed, psychological egoism is an empirical claim, then its falsity is so apparent that no special training is needed to show it false.

We are assuming that like any competing psycholgical theory, the theory we have called psychological egoism is an empirical scientific theory. As such it should have one feature in common with other empirical theories, that is, it should be empirically falsifiable. There should, then, be some empirically ascertainable situation which if it occurred would falsify the theory. What we seem to need to test psychological egoism is a case where someone did not act in order to maximize his own pleasure or happiness. We can use, therefore, a case where someone acted to sacrifice his own happiness for the happiness of another, or, it would seem, any case of someone acting altruistically or benevolently. But surely cases of people acting benevolently are not uncommon. We read of parents working many extra hours to help educate their children, of people donating a kidney to help a person dying for lack of one, of missionaries who risk their lives to bring help and knowledge to backward peoples. In these and many other cases we surely seem to have people acting benevolently for others instead of acting for themselves. Thus it seems that we can conclude not only that psychological egoism is falsifiable, but also that it has been quite easily falsified. The argument we have used goes as follows:

  1. If psychological egoism is true, then each person always acts to maximize his own happiness.
  2. If each person always acts to maximize his own happiness, then no one acts benevolently.
  3. Some people act benevolently.
  4. Psychological egoism is false.

Reply: People Always Act Out of Self-love

Defenders of psychological egoism can be expected to reply with a counterargument also based on premise (6) but with the conclusion that (7) is false, for the reason that people always act out of self-love or self-interest even when what they do helps others. The point is that, although it is true that people often do benevolent acts (that is, acts which actually help others), they do not act benevolently (that is, act for the sake of the others they help). People always act for their own happiness even when what they do helps others. This -- one of the chief arguments to show that, in spite of the way things seem, no one acts benevolently -- can be stated as follows:

  1. If each person always acts out of self-love (that is, to maximize his own happiness), then he never acts benevolently (that is, for the sake of others).
  1. People always act out of self-love.
  2. No person ever acts benevolently [and (7) is false].

Butler's Argument: Acting Benevolently and Acting Out of Self-love Are Compatible

The previous argument has been widely discussed, but the classical refutation of it occurs in the works of Bishop Joseph Butler.5 The brunt of Butler's attack is aimed at premise (6), the premise common to this and the previous argument. Thus if his attack is sound both arguments fail. He denies that acting out of self-love or self-interest is incompatible with acting benevolently, and argues that even if we always act out of self-love we may still sometimes act benevolently. Butler begins by noting that acting out of self-love can be described as acting to satisfy the desire for our own happiness. But such a desire is not a specific or particular desire such as the desire for a chocolate ice cream cone. It is a general desire, like the desire we hope to find in our politicians to promote the general welfare of the people. And just as someone can act to satisfy the desire for the general welfare only by doing specific acts, such as passing specific laws, so we can act to satisfy the (general) desire for our own happiness only by doing specific acts, such as buying a chocolate ice cream cone. There is, then, no, one specific act which satisfies the general desire for our own happiness so that when we act we always act to satisfy some specific desire whether or not we also act to satisfy the general desire for our own happiness.

We can distinguish two kinds of specific desires and thus two kinds of specific acts. FIRST, there are self-directed desires -- such as desires of hunger, sex, pride, and those desires relevant to our own health and knowledge. In each of these cases we satisfy a specific desire by doing something for (or against) ourselves. And although all selfish acts, for instance, taking candy from a baby, are self-directed, it does not seem that all self-directed acts, for instance, buying candy for oneself in normal circumstances, are selfish.

Second, there are other-directed desires, such as desires to help or harm someone else. In these cases we do something specific for or against someone else rather than for or against ourselves. When we act out of self-love we often do so by acting to satisfy specific self-directed desires, as when we eat, make love, practice tennis, take pills, and sometimes read books. Sometimes, indeed, we act selfishly when we do these things. However, we also often act out of self-love by acting to satisfy other-directed and thereby clearly unselfish desires, as when we get satisfaction from helping someone, teaching someone, or even harming him. In the first cases we are acting specifically for ourselves, but in the second we are not, because in the first we are acting to satisfy specific self-directed desires, but in the second we are acting to satisfy specific other-directed desires. Consequently, we can conclude that although acting to satisfy specific self-directed desires is inconsistent with acting benevolently, because acting benevolently is acting specifically for someone else, acting out of self-love and acting benevolently are not inconsistent, because acting to satisfy specific other-directed desires is consistent with acting to satisfy the general desire for our own happiness. We can, then, act for ourselves in general at the same time we act for (or against) others specifically. Acting benevolently is not precluded by self-love, and premise (6) is thus false.

At this point a defender of psychological egoism might go so far as to assert dogmatically that we never do act to satisfy other-directed desires -- that is, he might claim that we always act specifically for or against ourselves and thus never act benevolently, although of course we sometimes do benevolent acts. But dogmatism is of no help because it does seem that people do, all too seldom perhaps, act benevolently. Therefore unless there is a sound argument to the contrary we should accept premise (7) as true. The psychological egoist need not take such a desperate course, however, because, as we have seen the previous objection to premise (6) destroys not only the argument against (7), but also the argument we used to prove that psychological egoism is false. We were wrong, consequently, to think that we could use examples of acting benevolently to refute psychological egoism.

Final Objection: People Do Not Always Act Out of Self-love

Our mistake was to stress examples of acting benevolently. We need, rather, to point out clear examples of people acting contrary to their own happiness, because taking such examples with premise (5):

If psychological egoism is true, then each person always acts to maximize his own happiness.
we can clearly show psychological egoism to be false. Once again it seems we have a straightforward empirical question before us, and once again the answer seems to be that there are cases where it is most implausible to claim that a person is satisfying a desire for his own happiness. People sometimes seem to sacrifice their own happiness when they act out of a sense of duty to do what they think is right. Many patriotic soldiers are convinced by intimate experiences that war is hell, yet they volunteer for dangerous missions which surely seem to be contrary to their own well-being and happiness. There have also been cases of people bent on revenge who plainly do not care what happens to them just as long as they destroy someone else. In such cases, it seems clear that we should conclude that these people are acting to satisfy other-directed desires rather than their desires for their own happiness. Consequently, once again it seems that we should conclude that psychological egoism is false.

At this point the almost invariable reply is that people act out of self-love even in those cases in which it seems evident that they do not act out of self-love. Once this move has been made, psychological egoism becomes like two other often heard claims (that is, people always act to satisfy their strongest desire, and people always act to reduce tension), in that even the most obvious contrary cases are not taken to count against the claim. That is, the theory has been made immune to empirical falsification. It is not merely claimed that there have been no actual cases which would falsify the theory, but rather that no case we could possibly conceive would falsify it. Psychological egoism is often held in this dogmatic way, but when it is, it is no longer an empirical scientific theory open to checking by observation. It has become consistent with anything we might observe and thus no longer a theory of the empirical science of psychology. Consequently, it is no longer the kind of theory that can be justified by the findings of psychology, so that science will not provide a reason for accepting it.

Should we reject this modified form of psychological egoism as we have the earlier version? Unlike the previous version, it really does not matter what we do. Recall that the previous version, if acceptable, was to be used as a test of what persons can do and thus as a means for rejecting ethical standards which prescribe actions incompatible with it. It was, then, importantly relevant to ethics. But the present version has been made compatible with any behavior we could observe, and thus it cannot be used to test the actions prescribed by standards that compete with egoistic hedonism. Psychological egoism has been weakened to such a degree that it cannot help us decide among ethical standards. Thus any ethical theorist can accept it. We can, if we like, also accept this version of psychological egoism, but if we do, we can immediately ignore it because it has no relevance to ethics, or for that matter, to psychology. ~

We have found that the ethical egoist cannot rely on psychological egoistic hedonism to help him justify his position. If the theory is empirical it surely seems to be false, and if it is nonempirical it does not help provide reasons to choose ethical egoism instead of some other ethical theory. Consequently, the ethical egoist must find some other way to justify his theory. The only other remotely plausible argument is based on the fact that when someone asks us why we did something we often answer him by saying it was because we wanted to and that settles it. For example, if you ask me why I went to the movies instead of studying last night I might answer that I felt like going to the movies or that I did not want to study, and when I reply this way, the question is answered. Now, the argument states, because such a question is a request for justification of my action, I have justified my action by answering the question. Thus, I have provided a good reason for what I did by reference to what I wanted, so that my doing what I want is justified and therefore is what I ought to do. Because this argument, which we can call the argument from good reasons, embodies two mistakes that are relevant to our current interests, we should examine the argument with some care.

The Argument from Good Reasons

The crux of the argument can be outlined as follows:

  1. If I desire to do something (do something to get pleasure), then I can justify doing it by reference to the desire (pleasure).
  2. If I can justify doing something, then I have a good reason for doing it.
  3. If I have a good reason for doing something, then it is what I ought to do.
  4. If I desire to do something (do something to get pleasure), then it is what I ought to do.
This argument, like the previous argument, fails to justify ethical egoism, but for different reasons. First, it provides only a sufficient condition of obligation, whereas ethical egoism states a necessary condition as well. Admittedly, this is not a vital point, because we can design a parallel argument which would allow us to conclude that if I do not desire to do something, then it is not what I ought to do. The second reason is that ethical egoism refers to what maximizes my pleasure rather than to what merely gives me pleasure. We can fix this, however, by referring to what I most desire and what I do to maximize my pleasure. Let us assume, then, that from this argument, with the parallel negative argument, we can infer that ethical egoism is true. But is the argument sound?

Objection: Desiring to Do Something Does Not Justify Doing It

We have based premise (1) on the fact that we quite often can answer a "Why?" question once and for all by saying that we wanted to. Thus we often can answer a request for justification by such an answer. But we have inferred from this that we provide a justification for what we do whenever we answer in this way. There are two reasons why this is a mistaken inference. First, when I reply to a question such as, "Why did you go to the movies instead of studying?" by saying, "I wanted to," or "I felt like it," my answer does not provide a good reason for what I did, but rather is used to refuse to give a reason or to claim that no reason is necessary, ft functions more like "No reason," "I don't know," "Because." or even a shrug of the shoulders. Thus when someone answers this way he is not justifying his action, but rather is claiming that there is nothing to justify or is refusing to give a justification. Such an answer may stop the questioning but it does not justify the action. This can be illustrated by a different example. Suppose that you ask someone why he shot that old woman as she crossed the street and that he answers that he felt like it and will say no more. He might have just as well shrugged his shoulders, for in neither case would he justify what he did. In this example we are not satisfied with his answer because here, unlike the first example, a justification is called for and he has provided none. Thus although we can sometimes answer requests for justification by talking about our desires and pleasure, we have not provided a justification for our actions, because such an answer is adequate only when no justification is required. Secondly, moreover, even if we were to agree that sometimes we can justify our actions in this way, the shooting example shows that there are many times when we cannot. There are situations in which only I am affected, so that the only morally relevant factors are my own preferences. But where others are affected there are other morally relevant factors. Thus we can reject premise (1) and with it this argument for ethical egoism.

Before we move on we should also note that premise (3) is false, because this reemphasizes an important distinction. It is true that if I have a good reason for doing something then it is permissible for me to do it, that is, it is not wrong for me to do it. But it does not follow that it is always something I ought to do. Often when we justify an action we show that it is not prohibited rather than that it is obligatory. For example, it seems that many times people can provide good reasons for not helping someone who is being attacked by a gang of knife wielders in the caverns of some large city. But although such reasons show he has no obligation to help, and therefore is normally permitted not to help, they would not show that he had an obligation not to help. Once again we must distinguish between 'no obligation to' and 'obligation not to' to emphasize that, although the person who does not help in this case should not feel immoral, neither should he feel particularly moral.

Rejection of Egoistic Hedonism: Prescribes Morally Repugnant Acts

We have found no way to justify that species of ethical egoism we have called egoistic hedonism. Should we reject it? We have been operating on the principle that if an ethical standard prescribes certain actions a person feels certain are morally wrong, and these beliefs are not inconsistent with his other beliefs, then the person has some reason to reject the standard. In this case, in addition, there are no counterbalancing arguments in favor of egoistic hedonism. Moreover, the actions prescribed by this standard, we think, are ones which virtually anyone would feel are morally wrong. Thus, consider a sadist or someone who hates a whole race of people. The catalog of sadistic pleasures found in Justine by the Marquis de Sade is enough to show that many actions which give people pleasure are nonetheless morally repugnant to virtually everyone who considers the matter. The pleasure some Nazis found in torturing, maiming, and killing Jews, the attitudes of thrill murderers, and the boastful and brash killings of some black persons by certain white Southerners all testify that egoistic hedonism would not only permit, but also make obligatory some of the most horrendous crimes ever committed. We should, then, conclude not only that there is some reason to reject this standard, but also that egoistic hedonism ought to be rejected outright because by making each person's pleasure one's guide to what is right and wrong, it prescribes the kind of selfishness which ignores the happiness and welfare of any other person.


We have rejected egoistic hedonism which is one form of ethical egoism, but should we also reject ethical egoism in general? To answer this we must do the things we did in evaluating egoistic hedonism, that is, find out whether there are any arguments which justify it and find out whether it prescribes any actions that we are convinced are wrong. We have defined ethical egoism as the theory that makes the following claim:

Each person ought to act to maximize his own good or well-being.
Corresponding to the standard we used for egoistic hedonism we have the following:
A person ought to perform an action if and only if he does it to maximize his own well-being (that is, if and only if he does it to maximize his own best interest).
We have come to this standard by substituting 'well-being' for 'pleasure.' Consequently, we can arrive at the arguments for ethical egoism by the same substitution. What we find, then, is an argument based on the theory that we always act for our own well-being or self-interest, and an argument based on the claim that we can justify our actions by reference to self-interest. We shall leave it to the reader to substantiate that the arguments in this form are no better than those in the previous form. Consequently, if the general theory of ethical egoism prescribes actions that we are certain are morally wrong, we can reject it with egoistic hedonism.

Objection to Ethical Egoism: Prescribes Morally Repugnant Acts

The first thing to note in examining what ethical egoism prescribes is that it does not prescribe many of the specific actions prescribed by egoistic hedonism, because what provides me with maximum pleasure often is not what maximizes my well-being. This is especially noticeable if we identify our well-being and self-interest with mental and physical health and abilities. Some people have a choice between, on the one hand, a life of intense pleasures in which their health deteriorates and they do not develop their abilities, and, on the other, a restrained, often arduous and regimented life in which they maintain their health and develop their abilities. It is not unlikely that the first life, even if considerably shortened by an early death --especially if it came quickly and painlessly -- would contain more pleasure than the second. But the second would be more conducive to the well-being of the person. Thus these two different standards might often differ in what they prescribe, so that the examples we used against egoistic hedonism cannot be used in their present form against ethical egoism. Nevertheless examples can be devised which will provide grounds for rejecting ethical egoism. One example is the case where three people have a disease which is fatal unless they take certain pills. One of these people, unknown to anyone else, has the only three pills available, and he knows that if a person takes one he has 90 percent chance of surviving the disease, if he takes two he has a 94 percent chance, and if he takes all three he has a 99 percent chance. What ought he to do? It seems clear that he ought to give each person one pill. But ethical egoism prescribes that he maximize his own well-being; that in this case he ought to take all three pills himself and let the others die.

Plato on Morality and Self-interest

Only one reply seems open to an ethical egoist at this point and that is the claim that taking all three pills is not really in the person's own best interest, but it is not clear how this can be defended. Most modern philosophers have not tried to defend it and have in general rejected ethical egoism. However, this was not true of the ancient Greek philosopher Plato. He was interested in attempts to justify actions as moral or immoral and he thought that one way to do so, and perhaps the only way, is by establishing that moral actions benefit the doer and immoral actions injure the doer. That is, he tried to show that acting morally is in the best interest of the doer and acting immorally is against his best interest, in spite of the way it often seems. If he. had succeeded, then it might be claimed that we would have the best defense of morality possible, because if we could convince someone that acting morally benefits him, then he would be no less than foolish if he did not act morally.

We can see Plato's argument by reading an excerpt from the debate between Socrates, who expresses Plato's view, and Polus in the Platonic dialogue Gorgias. In this part of the dialogue Socrates is trying to show Polus that the worst thing that can happen to a person is that he do an unjust or immoral act and escape punishment for it. Thus the person who commits an unending procession of horrendous crimes against humanity and, escaping punishment, provides for himself a luxurious life of ease and pleasure which he seems to enjoy completely, is, according to Socrates, in a worse position than a person who committed the same crimes and is caught and punished. Further, both people are in a worse position than a person who always acts justly and as a consequence spends his life in unending pain and burdensome toil. Thus, according to Socrates, no matter how it might seem, it is always more in your own self-interest to act justly. We can see how Socrates argues for this implausible position by joining the discussion at the following point:

Socrates: Look at the matter in this way: In respect of a man's estate, do you see any greater evil than poverty?

Polus: There is no greater evil.

S: Again, in a man's bodily frame, you would say that the evil is weakness and disease and deformity, and the like. And do you not imagine that the soul likewise has some evil of her own? And this you would call injustice and ignorance and cowardice, and the like?

P: Certainly.

S: So then, in mind, body, and estate, which are three, you have pointed out three corresponding evils -- injustice, disease, poverty. And which of the evils is the most disgraceful? -- Is not the most disgraceful of them injustice, and in general the evil of the soul?

P: By far the most.

S: And if the most disgraceful, then also the worst?

P: What do you mean, Socrates?

S: I mean to say that what is most disgraceful has been already admitted to be so, without exception, because it is most painful, or hurtful, or both. And now injustice and evil in the soul has been admitted by us to be most disgraceful. And most disgraceful either because most painful and causing excessive pain, or most hurtful, or both?

P: Certainly. '

S: And therefore to be unjust and intemperate, and cowardly and ignorant, is more painful than to be poor and sick?

P: Nay, Socrates; the painfulness does not appear to me to follow from your premises.

S: Then since the evil of the soul is of all evils the most disgraceful, but (as you argue) is not so by reason of its painfulness, the cause must be some enormous harm and evil, of preternatural magnitude. And I take it that that which is greatest in harmfulness will be the greatest of evils?

P: Yes.

S: Then injustice and intemperance, and in general the depravity of the soul, are the greatest of evils?

P: That is evident.

S: Now, what art is there which delivers us from poverty? Does not the art of making money? And what art frees us from disease? Does not the art of medicine? And what from vice and injustice? If you are not able to answer at once, ask yourself whither we go with the sick, and to whom we take them.

P: To the physicians, Socrates.

S: And to whom do we go with persons acting unjustly or intemperately?

P: To the judges, you mean.

S: -- Who are to punish them? \

P: Yes.

S: And do not those who rightly punish others, punish them in accordance with a certain rule of justice?

P: Clearly.

S: Then the art of money-making frees a man from poverty; medicine from disease; and justice from intemperance and injustice?

P: That is evident.

S: Which, then, is the best of these three?

P: Justice, Socrates, far excells the two others.

S: And justice, if the best, gives the greatest pleasure or advantage or both. But is the being healed a pleasant thing, and are those who are being healed pleased?

P: I think not.

S: A useful thing, then?

P: Yes.

S: Yes, because the patient is delivered from a great evil; and it is worth his while to endure the pain, and get well?

P: Certainly.

S: And would he be the happier man in his bodily condition, who is healed, or who never was out of health?

P: Clearly he who was never out of health.

S: Yes; for happiness surely does not consist in being delivered from evils, but in never having had them. And suppose the case of two persons who have some evil in their bodies or in their souls, and that one of them is being treated and delivered from evil, and another is not being treated, but retains the evil -- which of them is the more miserable?

P: Clearly he who is not being treated.

S: And was not punishment said by us to be a deliverance from the greatest of evils, which is vice? For justice chastens us, and makes us more just, and is the medicine of our vice. He, then, has the first place in the scale of happiness who has no vice in his soul; for this has been shown to be the greatest of evils. And he has the second place, who is being delivered from vice. That is to say, he who is receiving admonition and rebuke and punishment. Then he lives worst, who, being unjust, is not being delivered from injustice?

P: Certainly.6

We shall leave it to the reader to discover the flaws in the steps by which Polus let himself be led to Socrates' position, but it should be noted here that even if Socrates had succeeded in justifying his position he would have done so only by stripping ethical egoism of any value as a moral standard. Socrates claims that being just or moral is what is most beneficial to the soul and thereby to human beings, in spite of appearances to the contrary. Thus we must find out what is in our own best interest by finding out what are the right things to do. That is, we must know what is right in order to find out what is best for ourselves, because no matter what may seem to be most beneficial, it is not if it is unjust. Thus instead of self-interest providing the criterion of what is right as required by ethical egoism, we would need some independent ethical standard to determine what is right in order to discover what is really in our own best interest. Plato, by making what is right the basis of deciding what is most beneficial, has made it impossible to use ethical egoism as a criterion for deciding what is right.

Because there seems to be no better way to justify the general theory of ethical egoism than there is to justify that brand of egoism we have called egoistic hedonism, we should conclude that ethical egoism ought to be rejected. This is because there is no sound argument to support it and because it prescribes certain morally repugnant actions. These are actions which, if we look back at the relevant examples, we find have one morally relevant feature in common. In each case the action ethical egoism prescribes is morally repugnant because the standard ignores the happiness and welfare of other people affected by the prescribed action. In short, the standard of ethical egoism seems to be misguided because it ignores one ingredient of morality which seems to be essential, namely, impartiality. In deciding what ought to be done it seems that each person who is to be affected by the action should be taken into account. No one should be ignored and no one should be given privileged status. Let us turn to a theory which explicitly incorporates impartiality into its standard.


Utilitarianism is not only the name of a particular ethical theory, but it is the name of the doctrine that called for social reforms to be achieved by bringing the actions of persons and also governments in line with the ethical principle of utility. This principle as an ethical and social weapon for reform received its first eloquent expression in the writing of Jeremy Bentham. who defined it as follows:

By the principle of utility is meant that principle which approves or disapproves of every action whatsoever, according to the tendency which it appears to have to augment or diminish the happiness of the party whose interest is in question; or, what is the same thing in other words, to promote or to oppose that happiness, I say of every action whatsoever; and therefore not only of every action of a private individual, but of every measure of government.7
As Bentham himself realized, it might be more perspicuous to call his ethical principle the greatest happiness principle rather than the principle of utility, because the principle is to be concerned with happiness, pleasure, and pain of the parties affected by actions. And as he said in a footnote added later,
The word 'utility' does not so clearly point to the ideas of pleasure and pain as the words 'happiness' and 'felicity' do: nor does it lead us to the consideration of the number, of the interests affected; to the number, as being the circumstance, which contributes, in the largest proportion, to the formation of the standard here in question; the standard of right and wrong, by which alone the propriety of human conduct, in every situation, can with propriety be tried.8
It is important, then, to note that utilitarianism, which is the theory proposing the principle of utility as the correct ethical standard, equates the utility of something with its tendency to produce happiness or pleasure. Thus, the word 'utility' as we shall be using it here is not synonymous with 'usefulness,' so that when we consider the utility of something we shall not be considering what use it has or how useful it is but its relationship to the production of happiness.

The Principle of Utility

Let us try to state the utilitarian standard clearly. A first approximation is the following:

An action ought to be done if and only if it maximizes the pleasure of those parties affected by the action.
This statement, however, can be made more precise by specifying what is to count as an affected party. It may seem evident to many that the reference is to persons, but Bentham realized that the state or community as a whole can be an interested and affected party. There are, for example, crimes against the state, and some heads of state have insisted that the interests of the state are distinct from those of its citizens. Thus leaders have called on citizens to sacrifice themselves for the fatherland that is, sacrifice their own happiness, even their lives, for the state. There have been people who claim that the state is not only a distinct individual, but that it is an individual of more worth than any or even all of its citizens. It is surely important for our purposes, then, whether or not we are to count the state or community as a separate affected individual in formulating this standard. Bentham realized this and clarified what he meant by 'party.'
The community is a fictitious body, composed of the individual persons who are considered as constituting as it were its members. The interest of the community then is, what? -- The sum of the interests of the several members who compose it.9
According to Bentham, then, we need not consider the state as a separate affected party, so that we can change the standard to read:
An action ought to be done if and only if it maximizes the pleasure of those people affected by the action.
This fomulation is still ambiguous, however, because it might be interpreted as stating that an action is right only if the pleasure of each person affected is maximized. Not only is this not what Bentham meant, it is a standard which could rarely be met. In most situations it is not possible to maximize the happiness or pleasure of each person involved. Usually someone is going to be less than completely satisfied with what happens. What Bentham means is that the action which ought to be done in a particular situation is that one which maximizes the sum total of pleasure produced. Thus although in many situations some of those affected will be made unhappy and made to feel pain, we might try characterizing right actions as those which minimize the number of people made to be unhappy and to feel pain. However, even this modification is not quite right because an action which causes several people to have a slight headache is better than an action in which, all else being equal, only one person suffers an almost unendurable pain. Thus we must consider not merely how many people receive pleasure or pain from the action, but also how intense each pleasure or pain is. Thus let us take the principle to consider the total amount of pleasure and pain produced where the amount is a function of intensity per person and number of people affected. We can now state the principle in the following form:
An action ought to be done if and only if it maximizes the total amount of pleasure of those persons affected by the action.
We are not yet finished amending it. Although it may seem obvious that the feature of morality which we saw was missing from ethical egoism -- that is, impartiality -- is included in the preceding formulation of the utilitarian principle, there is still room for partiality. How we arrive at the total amount can be affected by many factors, including whether or not we give equal weight to each person affected. In any society where some people are considered second-class citizens then someone using the last version of the principle might weight the percent of the total pleasure or pain contributed by each person according to his status as a full-fledged or secondary citizen. It seems clear that black pleasure and pain in many parts of this country are not added in on equal grounds with white pleasure and pain. There are also examples where the pleasures of kings have been taken to count for more than those of their subjects. Bentham, however, wants no such fractionalizing, so we must state the factor of impartiality explicitly. This will give us the final version of the principle of utility:
An action ought to be done if and only if it maximizes the total amount of pleasure of those persons affected by the action, counting each person as one and no person as more than one.
We shall use this statement in our critical evaluation of Bentham's standard, which we shall begin by examining the proofs that have been offered to support it.

Arguments for the Principle of Utility

As Bentham realized, two kinds of proofs can be offered in defense of an ethical principle -- one he called direct proof, the other we can call indirect proof. The first kind of proof is a deductive proof in which the conclusion is the principle itself. Thus this kind of proof argues directly for the principle. In an indirect proof the principle is supported indirectly by refuting objections to it and by showing that there are objections to the opposing alternatives. This is the way we argued in critically evaluating the various positions proposed as solutions to the mind-body problem and the way we have been approaching the various ethical theories. Obviously an indirect proof cannot provide an argument as rigorous or grounds as solid as can a direct proof. Let us then begin by seeing whether there are any direct proofs available for the principle of utility.

Direct Proofs for the Principle of Utility: Deriving 'Ought' from 'Is'

Bentham thought that there were no direct proofs of his principle, "for that which is used to prove everything else, cannot itself be proved: a chain of proofs must have their commencement somewhere."10 Bentham's point here is that he is proposing the principle of utility as the ultimate or basic ethical standard. Therefore although other ethical principles may be deducible from it, it is not deducible from any other ethical principle. Although we can deduce certain obligations from certain ethical standards and perhaps those standards from others, the process of deduction must begin with some one or more ethical principles that are not deducible from any others. These are the basic principles. For Bentham, as for most ethical theorists, there is only one basic principle and this cannot be deduced from a more basic ethical principle.

It may occur to someone that although the basic ethical principle cannot be deduced from other ethical principles, this does not show that the basic principle cannot be deduced from any premises at all. Indeed why can the basic ethical principle not be deduced from some factual premises about the way things are? This has been tried. People have tried to deduce moral obligations from the nature of human beings, from the facts of evolution, or from facts about societies, cultures, and economic classes. In each case they have tried to deduce a normative ought-statement from a factual is-statement.

Hume's Objection: No 'Ought' Is Deducible from 'Is'

One of the first to cast suspicion upon deducing 'ought' from 'is' was David Hume, who said,

In every system of morality, which I have hitherto met with, I have always remark'd, that the author proceeds for some time in the ordinary way of reasoning, and establishes the being of a God, or makes observation concerning human affairs; when of a sudden I am surpriz'd to find, that instead of the usual copulations of propositions, is, and is not, I meet with no proposition that is not connected with an ought, or an ought not. This change is imperceptible; but is, however, of the last consequence. For as this ought, or ought not, expresses some new relation or affirmation, 'tis necessary that it shou'd be observ'd and explain'd; and at the same time that a reason should be given, for what seems altogether inconceivable, how this new relation can be a deduction from others, which are entirely different from it.11
Hume is making the straightforward logical point here that no ought-statement, that is, one that makes only an ought-claim and therefore no factual claim, is logically deducible from a factual is-statement. that is, one that makes only a factual claim and therefore no ought-claim. To see Hume's point here, consider any two propositions, P and Q, and imagine that Q is logically deducible from P. If it is, then an explicit self-contradiction will be deducible from the conjunction of P and not Q. For example, let P = 'It is raining and it is cloudy,' and let Q = 'It is raining.' It should be obvious that P entails Q or, put another way, that Q is deducible from P. After all, P is a conjunction and Q is one of the conjuncts. Now consider the conjunction P and not Q. In words it would be,
It is raining and it is cloudy and it is not raining.
It should be clear that from this proposition we may deduce the explicit self-contradiction,
It is raining and it is not raining.
since, again, the conjunctive propositions includes the relevant conjunct. This illustrates the key idea that if Q is deducible from P, then a self-contradiction is deducible from the conjunction of P and not Q.

We can understand Hume's point by utilizing this key idea and reasoning backyards, so to speak. Suppose we have a purely factual claim with no 'ought,' such as the proposition A, which is 'Helping others is maximizing happiness,' and a normative ought-claim, not B, which is 'We ought not help others.' The first thing to notice is that the conjunction A and not B is not a self-contradiction. Moreover, no self-contradiction is deducible from the conjunction A and not B taken by itself. Hence, (here is where the reasoning backwards comes in) given what we called the 'key idea,' we may infer that B is not deducible from A. And, Hume would say, what holds for propositions A and B here holds generally, for any pair of propositions one of which is a purely factual claim and the other of which is a purely normative, ought-claim. No ought-claim is deducible from a purely factual claim. We thus cannot deduce that an action ought to be done from the premise that it maximizes the general happiness or the total amount of pleasure.

We have not yet reached a conclusion about the principle of utility or any other ethical standard, because such standards typically include both ought-claims and is-claims (factual claims). However, we can use Hume's conclusion to draw one about ethical standards. Let us assume that O is some ought-statement, F is some factual statement which consists of a conjunction of all true is-statements that make factual claims, and S is some ethical standard such that O or some other ought-statement is deducible from S, depending upon which factual statements are conjoined with S. Thus if S were the principle of utility and F includes the statement F1, 'A is an action that maximizes the total amount of pleasure,' then we could deduce O, 'A ought to be done.' Now we have seen that no ought-statement is deducible from any purely actual statement. Consequently, O is not deducible from F alone, but, we can assume, O is deducible from S and F. From this we can conclude that S, which stands for any ethical standard, is not deducible from F alone, the conjunction of all true factual statements. Thus, no ethical standard is deducible from any or even all true factual demises.12

The reasoning here is somewhat intricate. However, it can be simplified in the following way: Suppose, that S is some ethical standard such as the principle of utility, F is the conjunction of all true factual claims, and that O is some ought-claim. Assuming that F includes F1 as a conjunct, we have

1. O is deducible from the conjunction S and F.
Now we already know, from an argument given earlier on pp. 293-94 that no purely factual claim entails a purely ought-claim. Thus, we may say,
2. It is not the case that O is deducible from F. (Alternatively: It is not the case that F entails O.)
From (1) and (2), we want to conclude that,
3. It is not the case that S (an ethical standard) is deducible from F (set of all true factual claims).
To see how to get this conclusion from (1) and (2), we may use an indirect proof or argument. That is, we may start by assuming the very opposite of (3), namely,
4. S is deducible from F.
Now we may certainly allow that,
5. F is deducible from F.
But, if S is deducible from F, and F is likewise deducible from F, then so is their conjunction. That is, from (4) and (5), we get,
6. The conjunction S and F is deducible from F.
Notice, though, that when (6) and (1) are taken together, we get directly,
7. O is deducible from F.
The move from (1) and (6) is just a type of hypothetical syllogism: If the conjunction S and F is deducible from F [ = (6)], and O is deducible from the conjunction S and F [ = (1)], then O is deducible from F [ = (7)].

Statement (7), however, is the very opposite of (2). These two statements taken together amount to a contradiction, to the effect that O both is and is not deducible from F. We derived this contradiction by means of the introduction of (4) which, we may recall, is the very opposite of (3). So, we may conclude that (4) is, so to speak, the logical culprit here. From premises (1) and (2) and (4) we find that we may deduce a self-contradiction. Hence, using the very same reasoning we used earlier to discuss 'is' and 'ought,' we may conclude that (3) is deducible from the conjunction of (1) and (2).

A Further Objection: Naturalistic (Definist) Fallacy

We have seen that no ultimate ethical standard is deducible from any other ethical standard and that no ethical standard is deducible from purely factual premises. It would seem, then, that we could conclude that there is no direct proof possible for an ultimate ethical standard. However, such a conclusion would be premature. Although no ought-statements and no ethical standards are deducible from a set of premises all of which are factual is-statements, it might be true nevertheless that with the addition of only certain analytic premises we can deduce some ought-statement or some ethical standard. If this can be, done, then, because the additional premise is logically necessary, we can conclude that it is logically necessary that if the factual premises are true, then so also is the ought-conclusion. That is, the factual premise would entail the ought-conclusion, and, after all, 'ought' could be derived from 'is.'13

To see how this might be applied to 'ought' and 'is,' let us consider the following argument:

  1. A is an action that maximizes the total amount of happiness.
  2. A ought to be done.
This argument with a factual is-premise and an ought-conclusion is invalid. But if we add as a premise:
  1. Whatever maximizes the total amount of happiness is what ought to be done.
then the argument is valid. And if (3) is an analytic statement and therefore necessarily true, then we can conclude that (1) entails (2). Consequently, someone might offer the preceding argument to show how 'ought' can be derived from 'is.' This, of course, raises the question of whether (3) is analytic, that is, a statement whose truth can be established by appealing only to logic and the meaning of its terms. Some people seem to have thought that it is, and the view has even been ascribed to Bentham in spite of the fact that he thinks the principle of utility requires an indirect proof. We should, therefore, examine the claim that (3) is analytic, because if it is, then we need proceed no further in our search for a justifiable ethical standard.

Some might argue that (3) is analytic on the grounds that, first, it is analytic that what ought to be done is what maximizes what is good, and second, that the general happiness is, by definition, what is good. This is the way that G. E. Moore interprets Bentham when he claims that Bentham, like many others, commits what Moore calls the "naturalistic fallacy."14 This fallacy, according to Moore, is committed by anyone who defines an ethical term such as 'good,' or 'right,' or wrong' by terms that are purely factual or descriptive, and therefore have no evaluative force. Thus the naturalistic fallacy is committed when someone defines ethical terms such as 'good' using only such empirical terms as 'pleasure,' 'happiness,' 'desire,' or 'interest.' However, the fallacy is not restricted to those definitions that contain only naturalistic, that is, empirical terms. It has been pointed out that this fallacy might better be called the definist fallacy because, as Moore says himself, it is committed whenever someone defines an evaluative term such as 'good' by any nonevaluative term.15 Thus not only naturalistic or empirical definitions, but also metaphysical and religious definitions would involve the fallacy. Thus if Moore is right, to define 'good' as 'what God wills' is to commit the same fallacy as to define it as 'pleasure.'

Now we must ask why Moore thinks that any such definition is fallacious. His primary reason is that any such definition would make many open or debatable questions closed and trivial. For example, if someone were to define 'what is always good' as 'pleasure,' then the seemingly debatable question 'Is pleasure always good?' becomes no more than the trivial question 'Is pleasure pleasure?' It surely seems worth debating whether pleasure is always good, but no one would spend time debating whether pleasure is pleasure. Another way of seeing this is by realizing that many statements we use to commend or condemn someone for doing something would become mere trivially true analytic sentences and lose their evaluative force. For example, if I tell someone that he ought to promote the general happiness because promoting the general happiness is promoting what is good, I mean to support a certain kind of action by commending it. But if 'what is good' means 'the general happiness' then all I have said is that he ought to promote the general happiness because promoting the general happiness is promoting the general happiness. This latter claim is not only absurd, but it is clearly not a case of supporting something by commending it.16 I might just have said, "because killing is killing," or "promoting misery is promoting misery." But the original claim is not absurd. Therefore the latter is not an adequate translation of the original claim, and any other translation which leaves out the evaluative, and thereby the moral, element will also be inadequate.

The consequences of this fallacy are important. We can now state that no ethical claim is derivable from factual premises, because none is entailed by any factual statement. Any such entailment would involve the naturalistic or definist fallacy. Therefore we can also conclude that no ethical standard is entailed by any factual statement. In this way we have established what has been called the autonomy of ethics. That is, no ethical statements are derivable from any nonethical statements, so that no scientific findings entail any ethical principle, no metaphysical claims entail any ethical principle, and no (nonethical) religious claims entail any ethical principle. We cannot, then, hope to find a direct proof of the principle of utility or of any other ethical principle in which the principle is to be deduced from nonethical premises. And because we have seen that no ultimate ethical principle can be deduced from ethical premises, we can conclude that Bentham was right: There is no direct proof of the principle of utility or of any other ultimate ethical principle.

Bentham's Indirect Proof of the Principle of Utility

Bentham uses just the kind of indirect proof we are using throughout the chapter. First, he claims,

By the natural constitution of the human frame, on most occasions of their lives men in general embrace this principle, without thinking of it: if not for the ordering of their own actions, yet for the trying of their own actions, as well as those of other men.17
That is, according to Bentham, the principle of utility prescribes actions which in their uncritical way human beings believe to be right. Second, all principles which differ in what they prescribe from the principle of utility are confronted with objections sufficient for rejecting them. From these two premises, Bentham concludes that we are surely justified in accepting the principle of utility as the correct ethical standard.

Although by and large Bentham leaves it to the reader to investigate whether his principle is in line with our ordinary ethical beliefs, he does provide reasons for rejecting all opposing principles. He says that any principle different from his is either completely opposed to it or only sometimes opposed to it. The first opposing principle he calls the principle of asceticism, which, he says, "like the principle of utility, approves or disapproves of any action, according to the tendency which it appears to have to augment or diminish the happiness of the party whose interest is in question; but in an inverse manner: approving of actions in as far as they tend to diminish his happiness; disapproving of them as far as they tend to augment it."18 As Bentham points out, if such a principle were consistently followed, the earth would be turned into a living hell in a very short time. But Bentham's main attack is to point out that humans are incapable of consistently pursuing this principle. Consequently, because, as we have already seen, 'ought' implies 'can,' we can conclude that it is false that anyone ought to use such a principle. We can agree with Bentham that we should reject this principle.

All principles of the second kind opposed to the principle of utility, the kind which only in some situations opposes what the principle of utility prescribes, Bentham lumps together as various versions of what he calls the principle of sympathy and antipathy. By this he means

that principle which approves or disapproves of certain actions, not on account of their tending to augment the happiness, nor yet on account of their tending to diminish the happiness of the party whose interest is in question, but merely because a man finds himself disposed to approve or disapprove of them: holding up that approbation or disapprobation as a sufficient reason for itself, and disclaiming the necessity of looking out for any extrinsic ground.19
Such principles, as Bentham points out, do not appeal to any standard independent of the feelings and sentiment of those who propose the principles. The appeal is in every case to what someone or other happens to approve or disapprove. Surely no justifiable, ethical standard can be derived in this way. If Bentham is correct here, we should reject not only the principle completely opposed to his own, but also all those sometimes opposed. Only Bentham's principle would remain.

Objection to Bentham's Proof: Does Not Disprove All Opposing Views

There are two points at which we can attack Bentham's proof: his reason for rejecting all principles that are in some situations opposed to the principle of utility, and, second, his claim that no actions prescribed by his principle are morally repugnant. First, consider how Bentham characterizes all versions of the principle of sympathy and antipathy. No such principles, he claims, are standards independent of the feelings of people. This, of course, is not enough to distinguish these principles from his own, which is concerned with the happiness of people. However, he goes on to characterize these rival theories as replacing an objective standard by a mere reliance on feelings of approval and disapproval. Thus, according to Bentham, all these theories reduce to claims that we should make moral judgments merely on the basis of how we feel at the time. We can agree with Bentham that all versions of the principle of sympathy and antipathy should be rejected, but what seems clearly false is his claim that all principles sometimes opposed to his own are versions of the principle of sympathy and antipathy. Consider, for example, ethical egoism, which sometimes prescribes actions opposed to Bentham's principle of utility. It is clearly an objective standard applicable to all people at all times, and does not prescribe actions on the basis of what someone happens to find right or wrong at the moment. It is sometimes opposed to Bentham's principle, though not a version of the principle of sympathy and antipathy. Bentham's defense of his own principle fails, consequently, because he has not considered all rival principles.

Bentham might reply that principles sometimes opposed to his own also fail because they do not consider all people involved. But although this is true of egoism, it need not be true of any other rival to Bentham's principle, for they could be genuine rivals and consider all people involved so long as they did not consider only the happiness of all involved. It will not do, of course, for Bentham to reject the opposing principles on the grounds that they do not consider just the happiness of all. If he did, he would only show that they differ from his own principle, but this is not a sufficient reason for rejecting them. Therefore, this part of Bentham's indirect proof fails because he has failed to show that only the principle of utility and the previously rejected principle of asceticism are universally applicable principles which can be applied in an objective way.

The Hedonic Calculus

Although, as we have seen, Bentham has not shown that all principles different from his own can be rejected, this failure is not vital if, as Bentham thinks, his principle and his principle alone prescribes no actions that are morally repugnant to human beings. However, if we find situations in which what his principle prescribes is morally repugnant, then he is in serious difficulty. Let us, then, try to think of such a situation. To do so we must get some idea of how we are to arrive at a conclusion about what maximizes the total amount of pleasure in any one situation. The method proposed by Bentham is what has been called the hedonic calculus, because it proposes a way to calculate the total amount of pleasure by bringing in all the relevant factors. According to Bentham there are seven different relevant factors which we can break down into three different basic categories. The first kind of factor is that which includes the relevant characteristics of each pleasure and pain produced by the action being considered; the second kind includes the tendency of a particular pleasure or pain to be followed by more pleasure and pain; and the third kind consists in the method of including in the calculations all the pleasures and pains which result from the action being considered. Let us list these factors as follows:

    Intrinsic Characteristics of Pleasure and Pain
  1. Intensity of each pleasure or pain.
  2. Duration or length of time of each pleasure or pain.
  3. Probability that the pleasure or pain will occur after the act.
    This is affected by:
  4. Propinquity or nearness in time of the pleasure or pain to the act.

    Consequential Characteristics of Pleasure and Pain

  5. Fecundity or probability that the sensation will be followed by other sensations of the same kind.
  6. Impurity or probability that the sensation will be followed by other sensations of the opposite kind.

    Summation of All Pleasures and Pains Resulting from Act

  7. Extent of pleasures and pains.20

We can illustrate by a simple example how these factors might affect the sum total of pleasure and pain resulting from an act. Let us say that you, a person with barely enough money to eat, find a wallet containing $1,000 and cards identifying the owner as a multimillionaire. You plan to send back the wallet, but you are debating whether or not to send back the money. What should you do? To decide, you turn to the hedonic calculus. You calculate that because neither you nor the millionaire have any dependents no one must be considered but the two of you. You only have to weigh your pleasure and his pain if you keep the money against your pain and his pleasure if you send it back. We can surely assume that the intensity of pleasure you can obtain by using the money to buy food, drink, and entertainment far outweighs the intensity of the millionaire's irritation at not having the money returned. Furthermore, the duration of your pleasure will probably far exceed his irritation. We can assume that it is quite probable that you will get the pleasure and he will become irritated, so that factors (3) and (4) will not have much effect. We can also discount the effect of (5) and (6) in the case of the millionaire, because once his irritation is gone he will have too many more important things to think about. But if we assume that you probably will drink too much as a result of keeping the money, we can say that the pleasure is somewhat impure because of the hangover which follows it. Thus we must subtract some part of the total of your pleasure. And because such pleasures usually are not followed by additional pleasures as well as the pain of hangovers, we can conclude that your pleasure is not at all fecund. Nevertheless, it seems clear that if you keep the money your pleasure far exceeds the millionaire's displeasure, so that there is a considerable overall increase in the total amount of pleasure. But if you return the money, the little pleasure the millionaire receives hardly outweighs the unhappiness you feel when you think of the good times you are missing. Given all this, the decision is easy. You should, if you apply the principle of utility, keep the money.

We have seen a simple example of how the principle of utility is to be applied in a specific situation. The question before us is whether there are certain situations in which the principle would prescribe morally repugnant actions. Someone might claim that we already have found such a situation because we always ought to return lost items to their owner. However, there are exceptions to this rule, such as the one cited by Plato where we should not return a lethal weapon to its rightful owner who has become a homicidal maniac. Furthermore, although the example we have used may seem to some to be a case where what is prescribed by the principle of utility is wrong, what it prescribes is not a clear-cut example of a morally repugnant act. It is by no means a clear counterexample to Bentham's claim that his principle generally prescribes actions in line with what we think is right. We need a stronger case to refute Bentham's claim.

An Objection to Bentham's Principle: Sadistic Pleasures

We are to use the hedonic calculus to find out what we ought to do, so that if the calculus prescribes an obviously immoral action we can reject Bentham's principle. Let us take an example from the Marquis de Sade. There is a roomful of men who get extreme pleasure from the sadistic mutilation of the girl Justine.21 She suffers great pain, but all the men enjoy great pleasure so that the sum total of pleasure in this case is greater than if the men forego their pleasure by allowing Justine to go her way unharmed. If we apply Bentham's principle, once again it is clear what ought to be done. The men should take their sadistic pleasures and Justine should suffer. But this is surely morally repugnant. Something has gone horribly wrong with a principle which prescribes such sadistic acts. It might be objected, however, that because the mutilation Justine suffers results in prolonged pain, whereas the pleasures of the sadists are short-lived, the total amount of pain outweighs the total amount of pleasure. This objection can easily be avoided by changing the situation to one in which this particular group always kills the object of their sadism at the end of the festivities by skillfully administering a drug that kills quickly and painlessly. Here we have an example in which murder, by cutting short sadistically inflicted pain, would, on Bentham's principle, remove an objection to willful injury.

Consider another example which illustrates again how emphasis on pleasure as the summum bonum can justify murder. Let us replace the sadists by a cult of people who hate pain but who get immense pleasure from mutilating a warm human body. This group carefully picks a victim who has no close family or friends and whose life is not particularly pleasurable. If they can, they try to choose someone who suffers from an illness so that they can eliminate pain. They kill such a person as skillfully and painlessly as the sadists; then they have their joyous rites. Such murders seem justified by Bentham's principle, but clearly they are wrong. Somehow, although the principle is, as we have seen, impartial, it still has omitted something essential to morality. We should, then, reject Bentham's principle of utility as we have rejected ethical egoism before it, because we have found no reason to accept the principle, but we have found reason to reject it.

This does not mean, however, that we have found reasons sufficient for rejecting utilitarianism, because Bentham's version is only one particular version. Another version, that proposed by John Stuart Mill, who followed Bentham in his ideas about social reform, is an explicit attempt to meet the kind of objection we have just raised. Let us turn, therefore, to consider Mill's ethical theory.


John Stuart Mill, whose father James Mill was a contemporary follower of Bentham, had ample opportunity to become acquainted with all the various objections raised against Bentham's theory. Consequently, in his book Utilitarianism, Mill set out to state and justify a version of the utilitarian principle. Like Bentham he attempted to refute objections to the principle and raise objections to opposing principles. Unlike Bentham he tried to construct a less indirect proof of his principle, but his proof was an obvious failure. We are primarily interested here in his defense of utilitarianism, in particular his refutation of the objection that if we treat all pleasures equally, as we must in applying the hedonic calculus, then sadistic pleasures as well as the merely bodily pleasures are to count on a par with the pleasures of contemplation, creation, discovery, and other so-called mental pleasures. That is, it is better to be a pig satisfied than a human dissatisfied; better in some situations that sadists are satisfied than that they are not satisfied. Mill answers this objection as follows:

It is quite compatible with the principle of utility to recognize the fact that some kinds of pleasure are more desirable and more valuable than others. It would be absurd that, while, in estimating all other things, quality is considered as well as quantity, the estimation of pleasures should be supposed to depend on quantity alone.22

Quality versus Quantity of Pleasure

It is obvious that, contrary to what Mill says, qualitative distinctions between pleasures are inconsistent with at least one form of utilitarianism, namely, Bentham's. The only characteristics of pleasure and pain we are to consider in the hedonic calculus are their intensity and duration. There is no factor available for distinguishing among different kinds of pleasures and different kinds of pains. Thus Mill has taken a radical departure from Bentham's theory. Just how radical his departure is can be seen by examining the criterion Mill proposes to distinguish between qualitative levels of pleasures. He says,

Of two pleasures, if there be one to which all or almost all who have experience of both give a decided preference, irrespective of a feeling of moral obligation to prefer it, that is the more desirable pleasure. If one of the two is, by those who are competently acquainted with both, placed so far above the other that they prefer it, even though knowing it to be attended with a greater amount of discontent, and would not resign it for any quantity of the other pleasure which their nature is capable of, we are justified in ascribing to the preferred enjoyment a superiority in quality, so far outweighing quantity as to render it, in comparison, of small account.23
Mill's criterion tells us to decide which pleasures are qualitatively superior by what amounts to a poll of those who have experienced the pleasures in question. This seems to be an eminently democratic way to decide the issue, but we shall see that it is not. It is possible that the results of such a poll will show merely a wide range of disagreement or even a preference for "pig" pleasures. Mill, however, seems to dismiss this possibility immediately, for he assumes that the "verdict of the only competent judges" will be that "the pleasures derived from the higher faculties [are] preferable in kind, apart from the question of intensity, to those of which the animal nature, disjoined from the higher faculties, is susceptible. . . ."24 It seems clear to Mill that the nobler pleasures, those associated with a person's intellect, will win the poll over the lower bodily or "pig" pleasures. Thus, for Mill, utilitarianism can avoid the objection that it is a pig philosophy. To understand why Mill is so certain of the outcome of such a poll we must concentrate on the key phrase, 'only competent judges.' In using this phrase Mill implies that the person who has savored the nobler pleasures but who prefers bodily pleasures is a backslider, a person of weak will who is not competent to judge. His vote, therefore, is not to be counted.

It may be that we can find some way to justify revoking the voting rights of the skid-row inhabitants who have fallen from some previous higher estate, but it is not clear what grounds there would be. It is by no means clear, however, how we are to handle sadists, masochists, arsonists, voyeurs, and others who might prefer exotic pleasures to the "noble" ones. Perhaps we should call these people perverts and allow only normal people to decide. But even if we could decide who is normal in some way that is not question-begging, we would still find many men like P. H. Lawrence who, if they were asked to choose between intellectual pleasures and sexual pleasures, would without hesitation claim the latter should be chosen. It would be very hard to show that these people are backsliders or perverts. Furthermore, they often try to justify their choice on the grounds that, for example, without sexual pleasures people become isolated, lonely, hollow shells with no capacity to communicate with their fellow human beings. These people often argue that in this age of alienation and automation the one way to avoid dehumanization is through a passionate clinging together built upon the emotional base of the joys and pleasures of shared sexual acts. There are of course many others, including many philosophers, who agree with Mill, but taking a poll is hardly the way to show they are right. And what happens to the great number of people who all their lives long have little chance to experience the noble pleasures through no fault of their own? On this issue these people do not count as one and, as a result, once a hierarchy of pleasures is decided, they might not count as one when applying the utilitarian principle.

A poll does not seem the way to decide the issue, but how else can it be decided? Very often when the issue is debated, the argument proceeds by referring to what the various pleasures are associated with or lead to. The qualitatively superior pleasures turn out to be those associated with what is better, for example, a person's intellect or a person's love of his fellow human being. But once this line is taken utilitarianism has been abandoned, because the basic ethical principle is the one used to distinguish the hierarchy of things which are good and for this there is no need of a reference to pleasure. A utilitarian cannot take this line. If we are going to be utilitarians we must agree either that all pleasures or else that only certain pleasures are the only things intrinsically good. If we take the first alternative, then the objection arises that utilitarianism implies that it is better to be a pig satisfied than Socrates dissatisfied. If we try the second, then we can merely list the qualitative hierarchy of pleasures without justifying the list by reference to anything else intrinsically good. Consequently, there will be no way to decide among alternative lists and thus no basis for deciding what ought to be done in particular situations. Because neither alternative is attractive, perhaps we should abandon utilitarianism and with it the claim, which we have been considering since our examination of egoistic hedonism, that pleasure is the one thing intrinsically good.


There are additional problems facing utilitarianism. Both Mill's and Bentham's versions face two additional serious objections. The first is based on the inability of utilitarianism to account for special duties. There surely seem to be duties or obligations that some people have because of their particular and special status, but that other people do not have. Consequently, these duties are different from the obligations we all have, for example, to maximize happiness. People who are parents, teachers, or judges, for example, have special obligations to their children, students, and defendants, respectively, obligations which others do not have to these same people. Utilitarianism seems unable to account for these duties. When a teacher grades a paper or an examination, she does not decide the grade on the basis of what would maximize the overall happiness in this particular case. She tries to grade solely on the basis of the quality of the work done even if the resulting grade produces more pain than pleasure. If she produces more pain, is she being immoral? Many students seem to think so, but it can hardly be right to adjust a grade according to how it affects the happiness of all those concerned. We can imagine a lonely student unjustly despised by his fellow students who would be handicapped unfairly relative to more popular students. Thus teachers seem to have, because of their unique position, an obligation completely independent of the principle of utility. Special duties, therefore, present another problem for utilitarianism and for its central claim that only pleasures are intrinsically good.

We have just seen an example in which applying the principle of utility results in someone being treated unfairly. A similar problem could arise if a judge or jury were instructed to decide guilt on the basis of what maximizes the overall happiness. This would be unfair, unjust in many cases This points to perhaps the most serious problem facing utilitarianism, the problem of fairness or justice. This seems to be a problem independent of the problem of special duties because not only judges and juries ought to be just; it is an obligation each of us seems to have regarding our fellow human beings. It may seem strange that fairness should be a problem for utilitarianism, because we moved to utilitarianism from ethical egoism in search of an impartial standard. It is true that utilitarianism is impartial in counting each person as one and no one more than one regarding at least quantities of pleasure and pain, but this is not the only kind of morally relevant impartiality, and surely not the only kind that can be relevant to justice.


The problem justice raises for utilitarianism is demonstrated by a scapegoat example. Imagine a town in which the young daughter of a prominent family had recently been kidnapped in broad daylight, then raped, and brutally murdered. The police are completely baffled and among the citizenry, aroused by the local newspaper, there is growing contempt for the police. It is becoming harder for the police to control the youth of the town, crime is on the increase, and fear is widespread. It seems that something should be done to restore confidence in law and the police. At this point the chief of police decides to find someone who can be accused of the crimes and brought to a speedy and decisive trial. The first hobo off the next through train is apprehended, and with planted witnesses and carefully chosen jurists quickly condemned to death. The town breathes easier, the police are praised, happiness and tranquility are restored except for one police patrolman who knew that the executed man was not guilty. But he is quickly reassured by the police chief who, not previously known for his morality, gives him a quick course in utilitarianism and then shows how the overall happiness has been maximized.

This is an example of an obvious miscarriage of justice. Such a case might occur, as can be readily shown if we describe the city as in the southern United States and both the rapist and hobo as black. Nevertheless no such cases should occur and certainly any ethical principle which prescribes them is clearly wrong. Thus utilitarianism, because it sacrifices justice to the overall happiness and thereby omits an essential ingredient of impartiality, should be rejected in favor of an ethical theory that takes justice as an essential part of morality. To find such a theory we shall turn to a standard that differs radically from any we have examined so far in that it does not consider the consequences of an act as relevant to deciding the rightness of the act. Such an ethical theory has been called "deontological" because it stresses that morality is essentially based upon the relationship an act has to moral laws or principles rather than its relationship to its consequences.


All the ethical theories we have examined so far have had two things in common. They propose something as the summum bonum or highest good, and they prescribe that what ought to be done is to maximize whatever is the highest good. For example, both egoistic hedonism and Bentham's utilitarianism agree that because pleasure or happiness is the one thing good in itself, it is the summum bonum and we ought to bring it about wherever possible. Where they differ is in their claims about whose pleasure each person ought to maximize. For such theories what is morally important is whether or not our actions have consequences that bring about what is the highest good. Theories that emphasize the consequences of actions have been called "teleological" ethical theories.

The Highest Good: A Good Will

The great German philosopher Immanuel Kant propounded an ethical theory that is very difficult to interpret, but it has generally been construed as a prime example of a deontological theory. We shall follow this interpretation. Kant began his search for a basic ethical principle in the same way as Bentham and Mill. He also began by attempting to find the highest good. What he concluded, however, was so different from the conclusions reached by the others that on his theory what is the highest good can be realized regardless of the consequences of an act. To see how he arrived at this conclusion we must understand the conditions he required of anything for it to be the highest good. According to Kant, the highest good must not only be good in itself, it must also be good without qualification.25 This means that there are no situations in which the addition of what is the highest good makes the situation morally worse. Using this as his criterion Kant can eliminate all the leading candidates for the highest good for the reason that each one when added to certain situations makes them worse. He eliminates the "higher" faculties such as intelligence and judgment because if a person with evil designs also has a high degree of intelligence, the results are worse. He eliminates traits of persons such as courage, resoluteness, and perseverance because "they can become extremely bad and harmful if the will, which is to make use of these gifts of nature and which in its special constitution is called character, is not good."26 He rejects what he calls gifts of fortune, including power, riches, honor, and the utilitarian candidate, pleasure, because they too can make certain situations worse than if they were lacking. If, for example, we were to hear that the executioners at Auschwitz got pleasure from their horrible deeds and even if we agree that pleasure is good in itself, we would not think that this pleasure made the situation better. Instead we would think it made their deeds all that much worse. After rejecting these candidates Kant proposes the one thing he can find that meets his criterion. He says, in a famous passage:

Nothing in the world -- indeed nothing even beyond the world -- can possibly be conceived which could be called good without qualification except a good will.21
Kant claims that the one thing good without qualification is a good will, but explaining what he means by 'good will' is far from easy. For our purposes it will be enough to begin by noting that according to Kant "the good will is not good because of what it effects or accomplishes or because of its adequacy to achieve some proposed end."28 This is because what we do as a result of willing may, through chance, bungling, or interference of others, be quite the opposite of what we decided. We know of the good-intentioned bungler and the villain who in spite of all he plans actually aids the hero. Kant says that the will "is good only because of its willing, that is, it is good of itself."29 This means that whether a will is good does not depend on the consequences of willing but upon the manner of willing. This is brought out by the following definition we can use to express Kant's point:
S has a good will = df. S acts out of respect for moral laws.
This is still only a beginning because we have introduced two new terms Kant uses, both of which require explanation, 'act out of respect for' and 'moral law.' The first can be explained by distinguishing it from 'act in accordance with' in the following way:
S acts in accordance with principle P = df.S does something that is consistent with what P prescribes.

S acts out of respect for principle P = df.S does something solely for the reason that what he is doing is consistent with what P prescribes.

We can often act in accordance with a principle without even being aware of it or even when trying to violate it. Most of us when we drive a car, act in accordance with laws regarding speed limits, sometimes because we want to, other times without any thoughts or desires about it, and sometimes even when we try to break the law if, for example, we mistakenly think the limit is lower than it is. In none of these cases do we act out of respect for the laws. We act out of respect for a law only when our decision to do something is based on, and only on, the reason that what we do is consistent with what the law prescribes. Thus to act out of respect for a law we must decide on the basis of reason alone, that is, without reliance on our inclinations and desires, to do what is consistent with what the law prescribes. If we then act on the basis of our decision, we can be said to act out of respect for the law.

The Moral Law and the Categorical Imperative

The expression 'moral law' is more difficult to explain. We know three things:

  1. A moral law prescribes what ought to be done.
  2. What ought to be done is to bring about whatever is the highest good.
  3. A will that acts out of respect for moral laws is the highest good.
From this we can conclude that a moral law prescribes just one thing, namely, that we act out of respect for moral laws. There are two important consequences of this. First, there is only one moral law, for there is only one thing prescribed. Second, because the moral law merely requires that we act out of respect for itself, it is unlike any previous basic ethical principle we have examined. They all prescribe what acts we should do, but this one prescribes how we should do any act. Therefore it is not the particular actions a law prescribes that make it a moral law, that is, it is not the particular content of any law that makes it moral. And because any particular law consists only of some particular content embedded in a lawlike form, it must be this lawlikeness of a law that makes it moral. Thus if we can find a law that expresses merely this lawlike form of laws, then we will have found the one and only moral law.

What is the form of all prescriptive laws? They can be distinguished from explanatory laws, such as scientific laws, in that they can be expressed as imperatives about people's actions. Thus legal prescriptive laws are often stated in the imperative mood such as "Do not speed!" and "No smoking!" And because there is no restriction upon the moral imperative except that it expresses the form of lawfulness, there are no conditions that must be met for it to be applicable. It is then an unconditional or categorical imperative and an imperative with universal application. Thus the moral law, by requiring that we act out of respect for itself, requires that we act out of respect for universal and unconditional lawfulness. The moral law requires that whenever we decide to do something we should decide to do it solely for the reason that doing it is consistent with what universal and unconditional lawfulness requires. And, as we are interpreting Kant, universal and unconditional lawfulness requires that the principles we actually base our decision on, what Kant calls "maxims," should have the form of universal and unconditional laws. The moral imperative, therefore, requires that we are morally permitted to act on a maxim only if our decision to act on it is consistent with our willing to make the maxim a universal and unconditional law governing the actions of everyone, including ourselves. Kant formulates the moral imperative as

Act only according to that maxim by which you can at the same time will that it should become a universal law.30


The preceding formulation of the categorical imperative is not the only formulation given by Kant, but it is the one he derives first. We shall examine his second formulation later on p. 311. One thing both formulations have in common is that they prescribe principles, and therefore actions based on the principles, independent of the consequences of the actions. An ethical theory that takes this as its basic ethical principle is a deontological theory. This, like other ethical theories, faces objections, but before we assess the objections we must decide whether we should interpret Kant's principle as expressing both a necessary and sufficient condition of moral permission, or merely a necessary condition because of the word 'only.' That is, it seems equivalent to

You are permitted to act on a principle P only if you can will P to be a universal law.
Furthermore, if we try to interpret it as a sufficient condition as well, then immediate objections arise. If the possibility of someone willing that a principle be a universal law is a sufficient condition of the principle being one he ought to act on, then we get morally repugnant results. For example, a masochistic sadist might have no trouble willing that the principle "Give Smith five lashes a day," be universalized to "Give everyone five lashes a day." But it should not be concluded from this that he is permitted to act on the principle of giving Smith five lashes a day. Therefore we should restrict the principle to stating merely a necessary condition.

Once we restrict the categorical imperative in this way, however, another objection rises. The restricted imperative is not helpful in cases where we can will a principle universalized, but are not sure of whether we ought to act on the principle. The most the imperative can tell us is that if we cannot will a principle to be universalized then we are not permitted to, that is, we ought not act on it. Consequently, Kant's imperative, although it may be an essential element of a basic ethical principle, cannot be the basic principle, because it is not applicable in many situations. Indeed, it might also be objected, it is not clear how it is applicable in any situation because it is not clear how we can derive particular obligations from such an abstract principle. Kant tries to rebut this second objection by showing how to derive particular duties from his imperative. What he attempts to do is show that someone who does a particular act on the basis of a particular immoral maxim would become involved in some kind of inconsistency if he also willed that the maxim become a universal law. Thus what Kant means by "You cannot will the maxim you act on to be a universal law" is that if you do, then you will be in some way inconsistent and thus your decision will be irrational. But making an irrational decision is contrary to acting out of respect for the moral law, because, as we have seen, we act out of respect for the moral law only if we decide on the basis of reason alone to act in accordance with it.

Let us examine two of Kant's examples to illustrate his method. One duty he derives from the first formulation is the duty not to make a deceitful promise, in order to borrow money, for example. In this case, according to Kant, the maxim would be

When I believe myself to be in need of money, I will borrow money and promise to repay it, although I know I shall never do so.31
If this maxim is universalized, we would have a law that whenever anyone needs money he makes a lying promise in order to get it. If this were a law governing everyone's actions, then, according to Kant, no one would believe a promise made under such circumstances and no one would be fooled into believing the false promise. The result is an inconsistency between the intention of the liar to deceive others and his willing a universal law that eliminates the deception. We can conclude then that we ought not make a lying promise. Here is an ethical principle prohibiting specific acts so that, if Kant's derivation is sound, he has shown us how to apply his abstract principle to specific acts. No act of lying is right.

Another of Kant's examples concerns the person who decides not to help someone who needs help. In this example Kant understands the maxim to be the following:

I shall not help another person even when he needs help.
If we were to make this into a universal law it would be a law that no one is to help any other person who needs help. But, claims Kant, all of us desire that someone help us when we are in trouble so that our desire for help would conflict with our willing this to be a universal law governing all human actions. Thus we have an obligation to help others in a specific situation when they need help.

Objection to First Formulation: Which Maxims to Universalize?

We have seen from the preceding two examples that Kant's method of deriving specific duties from the first formulation of the categorical imperative depends upon deriving an inconsistency when certain maxims are universalized. There are two basic problems for this derivation. The first is the problem of applying the first formulation to maxims. To which is it to be applied and to which should it not be applied? The second is the problem of whether or not Kant can, as he claims, derive a clear inconsistency in applying the first formulation. To see the first problem consider a wretched, starving person who makes a promise which he knows he cannot keep to an extremely wealthy person in order to get money for much-needed food and medicine. To what maxim are we to apply the imperative? Is it to Kant's quite general maxim or a more restricted one, such as:

Whenever I am starving and need food and medicine and the only way to get it is to make a deceitful promise, I will deceitfully promise a wealthy person who can spare the money.
It is by no means clear that this is an immoral maxim even if the person's intention in acting on it is in some way inconsistent with his willing to universalize it.

Consider also a very sly universalizer who whenever he makes a deceitful promise claims that his maxim is something like the following:

Whenever anyone is six feet tall, has one blue and one brown eye, a three-inch scar on his left cheek, a bullet wound in his right palm, a gold ring in his left ear, and needs money, he is to borrow money and make a deceitful promise to repay it.
What makes this universalizer sly is that the only person who fits this description is himself. Furthermore, he claims that this maxim is universal as it stands, for it is of the form:
Whenever anyone is X, he is to do Y.
which is the form of Kant's universalized maxim. Indeed, the maxim applies to everyone who is X. It is only a contingent fact that only he is X. Consequently, our universalizer, who uses this form with the preceding description for all his maxims, finds that nothing is forbidden and nothing is obligatory, because all his maxims are universal. Thus he can act on them and will them to be universal laws without inconsistency. The obvious reply to this is to say that some restriction must be placed on what we are allowed to substitute for 'X,' but it is not clear how to allow a phrase such as 'desperately in need of food,' but rule out the longer phrase invented by the sly universalizer.

Another Objection: Cannot Derive Specific Duties

The first is not the more serious problem, however, because it may be possible to place a satisfactory restriction upon the application of the imperative but it is not clear how to avoid the second problem. It is essential for Kant to derive some kind of inconsistency. The most plausible example he gives is the case of deceitful promising, but even here his derivation fails. There is an inconsistency only if someone decides to deceive someone and also decides to do something that stops him from deceiving the person. But the deception would not be stopped if the only thing that were to happen is that the actions of everyone who needed money became governed by a law requiring them to make deceitful promises. If the person a liar was trying to deceive did not know that there was such a law or did not realize that this was a situation covered by the law about needing money, then there is a very good chance that he would be deceived, especially if the deceiver were clever. Even if this practice had been occurring universally for centuries, there is, as the saying goes, "a sucker born every minute." There is, unfortunately, very often little resemblance between what people are willing to believe and the truth.

The problem is more evident in the second example, for in order to arrive at the inconsistency Kant must claim that all of us desire someone to help us when we are in trouble. If someone did not have this desire, then his universalizing the maxim not to help another would not be inconsistent with any of his desires. He would consequently, not be obligated to help others. There are, we are sure, some people who do not have this desire -- people, for example, who claim they belong to that almost mythical breed of people known as rugged individualists. Kant can at most claim that we nonrugged people would become involved in an inconsistency, but even here problems arise. First there is, as before, the problem of restricting the application of the imperative. Even if we specify in the maxim merely the way in which help is needed, such as 'needs help crossing the street,' some of us are at least rugged enough not to desire this kind of help. Second, in absolving the rugged individualist from responsibility for helping others, Kant seems to condone what we might call the rugged individualist fallacy: Because I need no help and everyone should be like me, I have no obligation to help anyone. Unfortunately, whatever we should all be, most of us are not rugged individualists. We sometimes need help and therefore there are times others should help us whether or not they need help themselves.

There, are, then, serious difficulties facing Kant's first formulation of the categorical imperative, difficulties which eliminate it as being sufficient to stand alone as the basic moral imperative. However, we should not reject it entirely, because it may be an important element in a satisfactory formulation of such an imperative.


Having rejected Kant's first formulation someone might wonder why we are going to examine his second formulation and, indeed, why we have considered Kant at all when our purpose has been to find an ethical theory that can accommodate justice. Although it is not obvious that the first formulation is related to justice, Kant, by requiring that the maxims we act on be universalized so as to be equally applicable to all, has incorporated into his imperative something essential to justice. When we get to the second formulation, however, we shall see clearly how Kant's theory overcomes the kind of difficulty justice poses for utilitarianism.

Kant, in stating his second formulation, gave expression to one of the greatest humanistic doctrines. It sums up in one short imperative the doctrine of the dignity and worth of the individual person:

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.32
There are two important prescriptions in this imperative. We are to treat people as ends, that is, we are to treat them as beings having intrinsic value in themselves regardless of whatever value they may have or lack as a means to some end. We are also never to treat people as things that are mere means. That is, although we can, do, and often must treat people as means, in so doing, we must also treat them as ends. Thus the farmer treats his plow and hired hands as means, a manufacturer treats his machines and laborers as means, and a student treats his books and teachers as means. But although it is all right to treat the plow, the machinery, and the books merely as means, the hired hands, laborers, and teachers must also be treated as ends. This implies that no person should be a slave or racially discriminated against or used as a scapegoat. Each person is an end in himself and must be treated that way. Here, surely, is the very essence of justice.

Kant claims that this formulation is just another way to express the very same moral law expressed by the first formulation. Although it is by no means clear why Kant thought this, perhaps the following will help explain what he had in mind. We have already seen that Kant holds that the good will is the highest good, so that a good will is an end in itself and should be treated as such. But we can treat a faculty of some being as an end in itself only by treating the being itself as an end. And because only a being who can act out of respect for law, that is, only a rational being, can possess a good will, it follows that we should treat rational beings with good wills as ends. Furthermore, because we cannot know from its effects whether a will is good, we cannot be sure whether or not any particular will is good. Consequently, in order not to omit any beings with good wills, we must treat all rational beings and therefore all human beings as ends in themselves. In this way, starting from the same premises as were used to arrive at the first formulation, we arrive by a slightly different route at the second formulation. In such a way Kant might have arrived at the second formulation and at the conclusion that it was equivalent to the first.

Another reason Kant might have had for thinking that the two formulations are equivalent, and thus formulations of the same law, is that he thought the same duties could be derived from each. He illustrated this by deriving the same duties from each formulation. Let us examine how he derives these two duties previously discussed. We shall see that the derivation is easier and more plausible in this case. Kant derives the duty not to make deceitful promises from the prescription not to treat people as means only by arguing that anyone who intends to make such a promise "sees immediately that he intends to use another man merely as a means."33 In this Kant surely is right, for deceiving someone to get something for ourselves is using the other merely as a means to our own gain. The obligation we have to help others is derived from the other prescription in the second formulation, that we treat people as ends. This means that we ought to further the well-being of people because this is how we should treat an end in itself. Consequently, it is not enough that we avoid treating people merely as means; we ought to do more.

Humanity, might indeed exist if no one contributed to the happiness of others, provided that he did not intentionally detract from it; but this harmony with humanity as an end in itself is only negative rather than positive if everyone does not also endeavor, so far as he can, to further the ends of others.34
Although Kant is far from being a utilitarian he arrives at an obligation which sounds quite utilitarian, that we ought to promote the happiness of others by treating each person as an end. Thus Kant's second formulation may provide a way to accommodate the greatest happiness principle and justice. However, in order to evaluate it we must see whether it faces other problems.

An Objection to Kant's Theory: Not Applicable in All Situations

There are three objections that cast doubt on Kant's second formulation of the categorical imperative, but they can all be adapted to apply to the first formulation also; thus they are really objections to the whole of Kant's ethical theory. We have already seen the first objection as it applies to the first formulation. Neither formulation is applicable in all situations. This problem arises for the second formulation in two kinds of situations. The first situation is one in which all the possible alternatives require treating someone as a means only, such as in the example of an overcrowded lifeboat where someone must be sacrificed as a means of preserving the others. Kant's imperative provides us with no way to decide. The second situation is one in which all the alternatives allow us to treat someone as an end, but each alternative involves different people. This is the problem confronting someone who is in charge of distributing welfare funds so limited that not everyone needing help can be aided. Kant's imperative does not provide a way to decide between, for example, a family with a talented child, another with a child needing medical attention and another with a mentally disturbed child who terrorizes the neighborhood. Although in each of these kinds of cases the decisions are most difficult, a principle which would provide a way to distinguish among the alternatives would be superior in at least one respect to Kant's imperative.

A Second Objection: Absolute Versus Prima Facie Duties

The second objection is a more serious one. According to Kant the duties we derive from the categorical imperative are absolute duties. He is therefore committed to what we previously called action absolutism, that is, that certain acts are always right or always wrong. Thus for Kant the obligation not to lie is an absolute duty, so that we ought not to lie under any conditions. This, however, leads to some results that are surely morally repugnant. Assume that you are trusted by the local Nazi commander in occupied Holland and that you are harboring an important Jewish refugee for whom the commander is searching. The commander comes to your door and, trusting you, asks if you are hiding the refugee. You know that he will go away without searching if you say you are hiding no one, and that saying nothing would amount to telling him the truth. It seems clear that you should lie in this situation, but a Kantian who remembers the absolute duty to tell the truth would say that you should admit you are hiding the refugee. Clearly such a Kantian is wrong.

It seems, therefore, that the duties derived from the categorical imperative should not be construed as absolute duties but rather as what have been called prima facie duties.35 That is, Kantian duties are duties we are required to perform unless they are overruled or overridden by something else required of us. Thus the duty not to lie is not an absolute duty, but merely a prima facie duty, because in some situations it is overridden by some other prima facie duty, such as helping a deserving friend in distress. And, in any particular situation, that prima facie duty which overrides all others is our duty proper, that is, what we ought to do in the situation.

We can state the distinction between an absolute duty and a prima facie duty by defining the key terms.

A has a prima facie duty to do P = df. There is something C1 that requires A to do P.
However, because a prima facie duty can be overridden and therefore might not be what we ought to do, we should also relate 'prima facie duty' to 'override' and to 'ought.'
A's prima facie duty to do P is overridden in situation S = df. There is something C1 that requires A to do P, but there is something else C2 such that C1 and C2 together do not require A to do P in situation S.
This expresses what happens to the prima facie obligation to tell the truth in the refugee example. In this situation there is something, for example, a Kantian rule, that taken by itself requires you to tell the truth, but that, when taken with the conflicting requirement to help the refugee, does not require that you tell the truth. Furthermore, in this situation you surely seem to be required to do something instead of telling the truth. What you are required to do, what you ought to do, what we have called a duty proper, is to lie. Thus

A ought (has a duty proper) in situation S to do P = df. There is something C1 that requires A to do P in situation S, and there is nothing else C2 such that C1 and C2 together do not require A to do P in situation S.36
Thus any prima facie obligation that is not overridden in some situation is a duty proper and we ought to do it in that situation.

We can now define 'absolute duty' as a duty for which there is no situation in which someone is required to do something instead of it. It is, consequently, a duty that is never overridden.

A has an absolute duty to do P = df. There is something C1 that requires A to do P, and there is no situation in which A is required to do something instead of P.
Once we see what is necessary for a duty to be an absolute duty we can also see that there are very few, if any, duties that are absolute. There generally is some situation in which a duty is overridden, indeed, in which we are required to do something else. Most of our duties, therefore, are more properly called prima facie duties.

It seems that Kantian duties are more accurately described as prima facie duties than as absolute duties for, as we have seen, there are situations in which they are overridden. We can handle the refugee example in this way, but if we do we must conclude that Kant's theory is not adequate because it cannot accommodate the concept of overriding The problem is further emphasized by the third and most serious objection to Kant's theory.

A Third Objection: Cannot Resolve Conflicts of Duty

The refugee example not only shows that there is a problem for Kant's theory because it seems to prescribe morally repugnant actions in certain situations, but it can also illustrate the problem that conflicts of duty produce for his theory. In the refugee example the person is faced with what is clearly a conflict of duties, for he has a duty to help the refugee and a duty to tell the truth. In this case it should be easy to resolve the conflict, but Kant's theory cannot handle it. If, as Kant thinks, his theory prescribes absolute duties, then in this example the person ought to do two things he cannot possibly do together. Thus not only is he obliged to do something he cannot do, but he is also unable to justify choosing one rather than the other. If we interpret the theory to prescribe prima facie duties, then, although the person is not obliged to do two conflicting things, he still has not way of deciding what to do. Consequently, Kant's theory cannot handle conflicts of duty. It seems that he has so divorced morality from the consequences of our acts that in a case like the refugee example where the consequences seem most relevant, Kant's theory is unable to help us.

We have found three objections to Kant's deontological ethical theory which, taken together, constitute a sufficient reason for rejecting the theory as providing the basic ethical principle. Thus, we must continue our search. We should not, however, reject Kant's theory outright because it embodies something that seems essential to any satisfactory basic principle, the requirement to treat all people as ends, and therefore justly. It seems, consequently, that Kant's imperative should be included in any satisfactory basic standard. The question is how to include it. A recent and much-discussed answer is the theory called rule utilitarianism.


Two of the main problems facing Kant's theory are that it cannot handle conflicts of duty and that there are situations in which it is not applicable. Two of the main problems for utilitarianism are the problem of the lower pleasures and the problem of justice. Because Kant's theory can accommodate what causes the utilitarian trouble and utilitarianism can avoid what causes Kant trouble, it seems that if the two could be encapsulated in one theory which eliminated the weaknesses of both while maintaining the strong points of each, then we might well have a satisfactory theory. Kant's theory stresses the importance of moral laws that prescribe duties. This enables it to account for justice. Utilitarianism, on the other hand, proposes a standard that can be applied to every situation. It can also be applied to laws; that is, we can evaluate a law by deciding whether or not its enforcement tends to maximize the overall happiness. Indeed it has been claimed that the way to evaluate any juridical law, newly proposed or in force, is to apply the principle of utility to it, because the purpose of government and thus of juridical laws is to maximize the general welfare.

Acts, Laws, Judges, and Legislators

The applicability of the utilitarian principle to juridical laws has led some philosophers to propose that the correct ethical standard should be constructed on an analogy with the way the utilitarian principle applies to juridical laws. To do this we must understand and distinguish between the relationship of a judge to a law and the relationship of a legislator to a law. P. H. Nowell-Smith claims,

The duty of the judge is to pronounce verdict and sentence in accordance with the law; and the question "What verdict and sentence ought he pronounce?" turns solely on the question "What verdict and sentence are laid down in the law for this crime?" As judge, he is not concerned with the consequences, beneficial or harmful, of what he pronounces. Similarly, the question "Was that a just sentence?" is one that cannot be settled by reference to its consequences, but solely by reference to the law.37
The judge, as Nowell-Smith points out, is concerned with deciding individual cases and in doing so he can rely only on the laws in effect. He cannot use the consequences, whether for good or ill, to help justify his decision. The judge, therefore, because his decisions are bound by a network of laws, functions in a somewhat deontological way. As Nowell-Smith says, however,
The duty of the legislator is quite different. It is not to decide whether a particular application of the law is just or not, but to decide what laws ought to be adopted and what penalties are to be laid down for the breach of each law. And these questions cannot be decided in the way that the judge decides what verdict and sentence to pronounce.38
The legislator should evaluate laws by their consequences rather than by some other set of laws, although, of course, what the consequences of a particular law are depends in part upon what other laws are already in effect. The legislator, then, because he evaluates not by means of laws but by consequences, functions in a somewhat utilitarian way. Using this analogy with legality, Nowell-Smith concludes,
The obligation to obey a rule does not, in the opinion of most ordinary men, rest on the beneficial consequences of obeying it in a particular case in either the short or the long run, as utilitarians have almost always supposed. But the reasons for adopting a rule may well be of the kind that utilitarians suggest.39
In other words, Nowell-Smith is proposing an ethical theory that restricts the application of the utilitarian principle to rules of conduct rather than to particular actions. The particular actions we do and contemplate doing are to be evaluated by moral rules which are in turn justified by the utilitarian principle. Let us call moral rules that are justified by the utilitarian principle "utilitarian rules." We have, then, what has been called both a restricted utilitarianism, because it restricts the application of the utilitarian principle, and a rule utilitarianism, because it restricts the principle to rules. This theory, therefore, differs from a theory that applies the utilitarian principle to acts. Thus it differs from Bentham's and Mill's theories, which we have interpreted as versions of act utilitarianism. We can better understand this difference by stating a two-part principle containing the central doctrine of rule utilitarianism.
  1. Someone has a prima facie duty to obey a rule of conduct if and only if having the rule in effect tends to maximize the overall happiness of those to whom it applies (that is, the rule is a utilitarian rule).
  2. Someone has a prima facie duty to do an act if and only if the act is prescribed by a utilitarian rule.


We have arrived at the rule utilitarian principle with the hope that it will incorporate the strengths of Kant's and Bentham's theories while eliminating their weaknesses. Let us, therefore, see how rule utilitarianism fares, but let us do so by bringing to bear upon it all the problems and objections we used to reject all the preceding theories, and what, as a result, we have found to be required for a satisfactory ethical theory. Any completely satisfactory ethical theory must provide a basic ethical principle which:

  1. Is applicable in any situation requiring a moral choice. (Kant's theory and Mill's utilitarianism, which provides no justifiable way to evaluate pleasures qualitatively, fail to meet this condition.)
  2. Accommodates special duties. (Act utilitarianism and ethical egoism fail here.)
  3. Resolves conflicts of duty. (Kant's theory fails here.)
  4. Guarantees the treatment of people as ends and thereby guarantees justice and impartiality. (Act utilitarianism and ethical egoism fail here.)
  5. Provides for consideration of the consequences of actions for human happiness. (Kant's theory seems to fail here.)
  6. Prescribes no acts we feel certain are wrong. (Ethical egoism and Bentham's utilitarianism fail here.)
It is clear that rule utilitarianism meets condition (5) and there is no reason to think it cannot meet conditions (1) and (6), although it is hard to evaluate with regard to (1) and (6) because little work has been done concerning specific recommendations for utilitarian rules. Nevertheless there seems to be no reason why there cannot be a utilitarian moral rule covering every situation and why any acts prescribed by these rules should be morally repugnant. At any rate, for the present let us assume that rule utilitarianism meets conditions (1), (5), and (6). It may seem that it cannot meet (2), (3), and (4), but rule utilitarians have concentrated on showing how their theory meets these conditions. They claim that the special duties of the parent, judge, and teacher can be handled because there are utilitarian rules imposing these duties, that is, rules that can be justified as tending to maximize the overall happiness of those affected. Although this has not been established it is at least plausible to think that the practices prescribed by such "special" laws have beneficial consequences for people. Therefore we can also grant that rule utilitarianism seems to meet condition (2).

Condition (3) seems to raise a serious problem, however, because, as we have seen, when there are several moral rules there will be conflicts of duty. Rule utilitarianism is faced with conflicts of duties, and as we have stated its principle, there is no way it can resolve such conflicts. The rule utilitarian's reply to this is that the principle is incomplete as stated. A provision must be added to the effect that, when there is conflict between the prima facie duties prescribed by utilitarian rules, then the overriding duty, the one that ought to be done, is to be decided by direct application of the utilitarian principle to the action. Thus when and only when someone faces a situation in which two or more conflicting prima facie duties are prescribed by utilitarian rules, he is to decide which action to do by the utilitarian principle. In all other situations the principle is to be applied only to rules. Thus condition (3) is met (and, incidentally, met in a way that justifies lying in the refugee example).

This brings us to the problem of justice, which causes the act utilitarian so much trouble. Is the rule utilitarian able to avoid the pitfalls of the scapegoat example? He tries to avoid it by handling justice in the same way he handles special duties. He claims that the obligation to be just follows from a rule which can be justified by application of the utilitarian principle. Thus the practice of treating people justly is prescribed by a utilitarian rule, because it is a practice having beneficial consequences for those affected. It seems, then, quite plausible to conclude that rule utilitarianism is a satisfactory ethical theory because the basic ethical principle it proposes seems to meet all the conditions we have found required of any satisfactory ethical theory.

Objection to Rule Utilitarianism: No Guarantee of Justice

However, before we conclude we have found the theory for which we have been looking, we should consider in more detail how the rule utilitarian accommodates justice. On this theory justice is assured only as long as the general practice of being just tends to maximize the overall happiness. It is possible, therefore, that in some societies a rule requiring justice would not maximize happiness. It is possible that a law which forces people into slave labor might in certain circumstances tend to maximize the overall happiness, even counting the unhappiness of the slaves. In such a situation the guarantee of justice disappears. Thus although rule utilitarianism can accommodate justice, whereas act utilitarianism cannot, there is no guarantee that it will. The kind of justification of rules required by rule utilitarianism depends so much on the particular circumstances that we cannot be sure any particular rule will be justified.

This is not the only way justice can be thwarted on the basis of the rule utilitarian theory. By this theory the rule of justice is merely one of many rules justified by the utilitarian principle. We have seen that where there is more than one of these rules it is likely that they will sometimes conflict, and when they do we are to apply the utilitarian principle directly to the action to determine what we ought to do. It is, therefore, likely that there will be occasions when the prima facie obligation to be just will be overridden so that on those occasions we ought to be unjust. Consider, for example, a society in which one rule justified by the utilitarian principle is that respect for law enforcement ought to be maintained. What such a rule prescribes could quite easily conflict with the obligation to be just. In such a situation the principle of utility would be applied directly to the action and the scapegoat problem would arise again. Accordingly, even if a rule prescribing justice is justified, it may well be that in particular instances unjust treatment would be obligatory. Therefore, although rule utilitarianism seems preferable to the other theories we have examined, it still has a flaw. We should continue the search.

It seems that if we want justice guaranteed we must incorporate it in the basic ethical principle, rather than justifying it in some derivative way. The only theory we have found which does this is Kant's deontological theory. If we can somehow make Kant's principle basic and also retain the features of rule utilitarianism, we shall have found a satisfactory theory.

We have seen that Kant's second formulation has two parts, one which prescribes that we treat no person as a means only and the other that we treat all people as ends in themselves. If we can in some way give content to what it is to treat people as ends, we may be able to find the theory we want. And although we only noted it in passing, we have already seen the hint we want in Kant himself. We know that in treating someone as an end, the minimum requirement is that we bring about and maintain the conditions necessary for his continued existence. But as with anything that is an end we should also promote his well-being. According to Kant (see p. 312), we should also treat him in a utilitarian way; we ought to promote his happiness. This suggests that we should take Kant's principle, and thus justice, as basic and use the utilitarian principle to help derive certain duties consistent with Kant's principle. What is wrong with rule utilitarianism is that it makes justice derivative instead of basic. Indeed it is more reasonable to justify maximizing human happiness by some reference to the principle that humans should be treated as ends, that is, in a Kantian way, than to justify treating people justly by reference to the principle that we ought to maximize human happiness. Human happiness is morally important because human beings are important. It is not that humans are morally important because human happiness is.


Our job is to find some way to graft the utilitarian principle to Kant's second formulation. We know two things: First, the basic prescription is, if possible, to treat no one merely as a means, but if this is not possible in a particular situation, then we should treat as few people as possible as mere, means. The overcrowded lifeboat example illustrates a situation in which someone must be sacrificed, treated as a mere means, in order to save the others. In such a situation it is obvious that as few as possible should be sacrificed. Second, we should treat as many people as ends as possible. We have interpreted this to imply that we should actively promote the well-being of those affected by the action in question. However, because promoting the well-being of as many people as possible could conflict with treating as few people as mere means as possible, and because the most basic imperative is not to treat people as mere means, the second imperative must be restricted so that it is consistent with the first.

At this point an objection can be raised. We can avoid treating a person as mere means by doing nothing at all. Consequently, in any situation we can avoid treating anyone as a mere means. If we accept the preceding imperative as basic, we should sacrifice no one in the lifeboat example, because that would be to treat as mere means as few people as possible. But that would result in a needless loss of life. We must, then, find a different basic principle.

We can avoid the objection by construing the treatment of someone as mere means to include doing nothing to help him when he truly needs help, especially when his life is imperiled. Not to do anything to help someone in such a situation is to respond to him as something with no intrinsic worth. This amounts to treating him as a mere means. We can, then, take the basic imperative to be

In any situation, (a) treat as mere means as few people as possible, and (b) treat as ends as many people as is consistent with (a).
We have claimed that promoting someone's happiness is important for treating him as an end. We should, then, incorporate into our imperative a prescription to promote the happiness of those affected by an action. However, because promoting as much happiness as possible often conflicts with the previously stated basic imperative, any prescription to promote happiness must be restricted so that following it is consistent with what our basic Kantian imperative prescribes.

Although this gives the essential skeleton of the principle, there is still the question of how we are to relate the treatment of as many people as ends as possible to the promotion of happiness. The problem is that there are several conflicting ways we could do this. We treat one person as an end by promoting his happiness. We could, then, require the action that promotes to some degree the happiness of the greatest number of people, or we could be utilitarian at this point and require that it maximize the total amount of happiness, counting, of course, each one as one and no one more than one. Let us initially choose an act utilitarian interpretation that gives us the following principle:

An action ought to be done in a situation if and only if
  1. doing the action, (a) treats as mere means as few people as possible in the situation, and (b) treats as ends as many people as is consistent with (a), and
  2. doing the action in the situation brings about as much overall happiness as is consistent with (1).
As the reader can discover for himself, this principle seems to meet all of the first five conditions that any satisfactory ethical theory must meet, except for (3), which concerns special duties. By applying the act utilitarianism principle to treatment of people as ends we have allowed the problem of the special duties of teachers to arise again. However, because this problem can be handled by rule utilitarianism, we can accommodate special duties by applying the rule utilitarian principle. Here again we have a choice to make. We can assume, as a rule utilitarian does, that there are utilitarian rules covering every situation involving a moral choice. Or we can make provisions for the existence of some situations not covered by these rules by requiring that the act utilitarian principle apply in these situations. Let us here, however, accept the rule utilitarian's assumption. What we can call the utilitarian Kantian principle will be
An action ought to be done in a situation if and only if
  1. doing the action, (a) treats as mere means as few people as possible in the situation, and (b) treats as ends as many people as is consistent with (a), and
  2. doing the action is prescribed by any utilitarian rule that (a) does not violate condition (1) in the situation, and (b) is not overridden by another utilitarian rule that does not violate condition (1) in the situation.
To help understand this principle, let us see what it would prescribe in one particular lifeboat example. Let us assume that you are the captain of a ship that has just sunk, and you are in charge of the one remaining lifeboat, which has too many people crammed into it and three others, who are taking their turns in the water, hanging onto the sides of the boat. Suppose further that it is clear that a dangerous storm is quickly approaching and that the boat will capsize unless five people at minimum are cast adrift. You must decide what ought to be done. The utilitarian Kantian principle requires you to sacrifice some people, but as few as possible, in the situation in order to save the rest. In this way you would treat as few as possible as mere means, and as many as possible as ends in this situation.

Once this decision is made you are faced with the problem of finding a procedure for deciding who is to be sacrificed. One decision procedure which clearly treats no one as mere means is to draw straws, but another one is to ask for volunteers. The basic Kantian requirement expressed in condition (1) provides no way to choose between the two procedures. Thus you must consider any relevant utilitarian rules. To see which rules apply, let us further assume that five people in the boat have publicly volunteered to be sacrificed. Consider now the following rule:

Whenever it is required that some people be sacrificed to save others, and some people have publicly volunteered to be sacrificed, then there is a prima facie obligation to sacrifice the volunteers.
This rule clearly applies in this situation and it does not violate what the basic Kantian condition requires. Furthermore, it is reasonable to think it is a utilitarian rule because its being in effect tends to maximize the overall happiness of those to whom it applies. Indeed, it is quite likely that if this rule were not followed when it applies, there would be great unhappiness, and strong resistance, or even mutiny, when those who did not volunteer, but know others did volunteer, are asked to take a chance on being sacrificed. And, given the additional plausible assumption that this rule is not overridden in this situation, your obligation is to ask for volunteers, rather than have the passengers draw straws.

The principle we have finally reached is complex. As can be seen from the preceding example, it requires of anyone that he consider and relate many factors in order to decide what he ought to do in any particular situation. For many people in many situations it is practically impossible to complete such a complex task. Each of us should, of course, do the best he can, and where anyone has done a reasonably good job but failed to decide correctly, no blame or guilt should be attached to him. As brought out in the beginning of this chapter, the standards appropriate for morally evaluating actions are different from those appropriate for morally evaluating persons. Although we have not considered the latter kind of standard here, one thing is clear: Many actions that are quite clearly wrong do not reflect blame or guilt upon the doer.


We have not considered whether the utilitarian Kantian principle meets the sixth condition -- that it prescribes no acts we feel certain are wrong. This task is left to the reader. If he finds that it meets condition (6), then he has good reason to think that it is a satisfactory ethical theory. There may be, however, a better theory that we have not considered. There are, for example, ways we have not examined for applying the utilitarian principle to the basic Kantian standard. Consequently, although we have, in this book, stopped examining ethical standards, the examination itself is not over. But it has come a long way. Early in the chapter we found reason to reject several views: theological ethics, ethical relativism, nihilism, and egoism, but we left open the question of whether we should accept ethical skepticism. We decided that if by the conclusion of our examination of ethical standards, we found none satisfactory, we would have good inductive grounds for the claim of the ethical skeptic that there are no justifiable ethical standards. Although the examination is not completed, we have, even at this point, some reason to think we have found a standard that at least approximates a justifiable one. We may be wrong, but even if we are, there are definite signs that progress has been and still is being made. We have, then, grounds for concluding that ethical skepticism is wrong. Although the task of justifying an ethical standard is most difficult and often discouraging, we have found reason to keep on trying with hope of achieving success.


1 See pp. 215-218.

2 For a clear argument against theological ethics, see Jeremy Bentham's discussion of the theological principle in An Introduction to the Principles of Morals and Legislation in The Utilitarians (Garden City, N.Y.: Doubleday, 1961). John Stuart Mill also discusses theological ethics in Utilitarianism, also in The Utilitarians, p. 423.

3 This kind of example is discussed by Solomon E. Asch in Social Psychology (Englewood Cliffs, NJ.: Prentice-Hall, 1952), Chapter 13, especially p. 377, where he is trying to show how widely different ethical practices can result from divergent factual beliefs instead of from different ethical standards.

4 This statement of psychological egoism should not be read as asserting that a person does everything he is psychologically able to do. It has the force of "A human being is psychologically able to perform an action if and only if he does it (if he does it at all) in order to maximize his own pleasure or happiness." We make use of the simpler version just stated for convenience.

5 See Joseph Butler, Five Sermons (New York: Liberal Arts Press, 1950), pp. 12-17 and pp. 49-65.

6 Adapted with changes from The Dialogues of Plato, edited by B. Jowett (New York: Random

7 Bentham, The Utilitarians, pp. 17-18.

8 Ibid., p. 291.

9 Ibid., p. 18.

10 Ibid., p. 19.

11 D. Hume, A Treatise of Human Nature, edited by L. A. Selby-Bigge (New York: Oxford University Press, 1960), p. 469.

12 See Chapter 1, p. 23, for a discusssion of deducibility.

13 See Chapter 1, pp. 15-23, for a discussion of analyticity and entailment.

14 See G. E. Moore, Principia Ethica (New York: Cambridge University Press, 1960), pp. 5-21.

15 See W. Frankena, "The Naturalistic Fallacy," in W. Sellars and J. Hospers, eds., Readings in Ethical Theory, 2d ed. (New York: Appleton-Century-Crofts, 1970), pp. 54-62.

16 See R. M. Hare, The Language of Morals (New York: Oxford University Press, 1952), Chapter 5, for a more detailed discussion of how the naturalistic fallacy results in the word 'good' losing its function of commending.

17 Bentham, The Utilitarians, pp. 19-20.

18 Ibid., p. 21.

19 Ibid., p. 28.

20 See ibid., pp. 37-40, for Bentham's statement of the hedonic calculus.

21 Marques de Sade, Justine, translated by Richard Seaver and Austryn Wainhouse (New York: Grove Press, 1965).

22 Justine, p. 408.

23 Ibid., p. 409.

24 Ibid., pp. 411-12.

25 Kant discusses what is good without qualification in the first section of The Foundations of the Metaphysics of Morals (New York: Liberal Arts Press, 1959), pp. 9-10.

26 Ibid., p. 9

27 Ibid.

28 Ibid., p. 10.

29 Ibid.

30 Ibid., p. 39.

31 Ibid., p. 40.

32 Ibid., p. 47.

33 Ibid., p. 48.

34 Ibid., pp. 48-49.

35 The concept of prima facie duty derives from P. Ross, The Right ani the Good (New York: Oxford University Press, 1955), pp. 18-20.

36 This way of defining these distinctions is derived from R. M. Chisholm, "The Ethics of Requirement," American Philosophical Quarterly (1964), 147-53.

37 P. H. Nowell-Smith Ethics (Baltimore: Penguin Books, 1954), p. 236.

38 Ibid., p. 237.

39 Ibid., p. 239.


1. Which of the following do you think are moral judgments? Which are not? Explain your answers.

  • People have an inalienable right to life, liberty, and the pursuit of happiness.
  • Godpunishes those who break his laws.
  • Thou shalt not kill.
  • Stealing is against the law.
  • We should always obey the law.
  • The use of narcotics is harmful to society.
  • 'Ought' implies 'can.'
  • Nothing is right or wrong but thinking makes it so.

2. State whether or not each of the following are clear examples of ethical relativism as it is defined in the text. Explain your answers.

  • Polygamy was morally permissible in Abraham's time but it is immoral today.
  • What you claim is right is only your opinion and therefore it is no better than anyone else's.
  • What is wrong is that which harms society, so what is wrong in one society is often right in another.
  • The principle of utilitarianism may be all right for Western cultures, but it surely is wrong for Oriental cultures.
  • There are no ethical standards, each of us just "sees" what is right and what is wrong.

3. Evaluate the following argument:

Causal determinism is true, from which it follows that no one has free will and, consequently, no one is morally responsible for what he does. But if no one is responsible for what he does, then he cannot be obligated to do anything. That is, there is nothing he ought to do; nothing is right and nothing is wrong. Therefore, ethical nihilism is correct.

4. Evaluate the following argument:

Ethical skepticism is correct because, in the last analysis, no person can do any more than follow his own conscience in ethical matters, and nothing can be justified in that way.

5. Suppose that someone named Jones had risked his life to save a drowning child. Discuss the following explanation of his actions:

Jones did not really act unselfishly, that is, he did not risk his own life for the sake of the drowning child. Rather, Jones is the sort of person who gets satisfaction and pleasure as a result of helping others, and his desire for this kind of pleasure was so great it even outweighed his desire to protect his own life. Jones, therefore, was really acting to get the future pleasure, not to save the child.

6. It has been claimed that some pleasures (for example, aesthetic pleasures) are higher than others (for example, sexual pleasures). Explain what someone might mean by such a claim. Would a psychological hedonist be consistent if he maintained that some pleasures are higher than others? Would an ethical hedonist be consistent? Explain.

7. Egoistic hedonism was rejected on the grounds that there were no sound arguments in its favor and it prescribes morally repugnant acts. Suppose that an egoistic hedonist were to reply as follows:

My theory does not prescribe morally repugnant acts because such acts are clearly those that harm others and to harm another is to invite retaliation or punishment. Because it is obviously not in one's interest to bring injury to oneself, then on the basis of egoistic hedonism the correct conclusion is that we should not harm others. My theory, therefore, does not prescribe any morally repugnant acts.
Evaluate his reply.

8. Consider Socrates' argument on (pp. 287-289). Is he justified in his claim that when a person is unjust, the injustice is a disease of his soul and, as for all diseases, the person would be better off if he had never been unjust at all? Consider a person who lives "happily ever after" on the spoils of his crimes. Is there some way he is really more miserable than if he had remained in just but grinding poverty?

9. One objection to Bentham's version of utilitarianism is that it assumes a "hedonic calculus," that is, a system by which various pains and pleasures can be added and subtracted as if they were apples or oranges. Thus we are supposed to be able to add the pleasure you get from making an important discovery to the pleasure I get from drinking fine wine, and get a result that is, let us say, three times the amount that Jones gets from calling a bluff in poker minus the headache Smith has. Surely, the objection goes, the very idea of such a calculus is absurd. Consequently, Bentham's principle of utility is absurd. Evaluate this objection.

10. The following seems to be a case of deriving an 'ought' conclusion from an 'is' premise. Does it, therefore, refute Hume's claim that 'ought' cannot be derived from 'is'? Explain your answer.

  1. Old Ms. Smith is in need of help.
  2. If I ought to help old women in need of help, then I ought to help old Ms. Smith.

11. Is the following derivation of 'ought' from a factual claim valid? Explain your answer.

  1. Mike says, "I hereby promise to help Smitty escape from jail."
  2. If someone says that he promises something, then he is obliged to do what he promises.
  3. If someone is obliged to do what he promises, then he ought to do what he promises.
  4. Mike ought to help Smitty escape from jail.

12. Which of the following clearly commit the definist fallacy and which do not. Justify your answers.

  • What is good is what is commanded by God.
  • The word 'good' means 'what ought to be maximized.'
  • Pleasure is, by virtue of the meaning of words, the one and only thing good in itself.
  • The good is, clearly, that at which all things aim. To say that something is good is to say that it is an object of human interest.

13. Evaluate the following objection to utilitarianism:

According to the principle of utility we are to treat each person as one and no one as more than one. Thus, each person is to get equal consideration. But if we accept the principle of utility we cannot do this, for some people are much more sensitive to pleasures and pains than others, and so when we "add" pleasures or pains an extremely sensitive person would weigh more heavily in our calculations than normally sensitive people. Thus, if we use the principle of utility we cannot give all people equal consideration as the principle requires. The principle, therefore, is inconsistent and should not be followed.

14. Smith, a Kantian, decides to make it his rule to eat dinner at 6:00 P.M. Accordingly, he universalizes this rule, that is, conceives of it as universal law, and is horrified at the result. If everyone ate dinner at 6:00 P.M., essential services would go unmanned, patients would be left on the operating table, airliners would crash, ships would run aground, and so on. Because he could not will this state of affairs, he concludes that he must reject his rule. Yet the same objection applies to a rule prescribing dinner at any other time, and he is faced with starvation. Has Smith made a mistake in employing the first formulation of the categorical imperative? If so, spell it out carefully.

15. It has been objected that Kant's ethical theory requires people to be bloodless and inhuman. According to this objection, Kant says that if I come across an injured person lying by the roadside, it would be wrong to help him because I feel sorry for him. That would be to act out of inclination or desire; rather, I should act out of respect for the moral law, on the basis of reason alone. If Kant is right, then we should be cold, unemotional, and unsympathetic. We should be highly moral machines. Because this is surely a repugnant view of how people should act, Kant's theory should be rejected. Evaluate this objection.

16. Suppose you want to decide whether or not to cheat on an exam. Find the maxim by which you would be acting if you did so and test it by means of each version of the categorical imperative.

17. Consider the following objection to utilitarian ethical theories, and compare how an act utilitarian could reply to it with how a rule utilitarian could reply to it:

Jones is dying in great pain and has possibly a month to live with no hope of recovery. Jones has no relatives and no one who stands to gain or lose by his death. His doctor, a good utilitarian, decides upon a mercy killing. Jones, however, objects and says that he wants to live as long as possible, however great his pain might be. Nevertheless, his doctor goes ahead and ends Jones' life. This would seem to be the action that utilitarian ethical theories prescribe, but it is clearly wrong.

18. One moral issue not discussed in the text is the problem of punishment. Compare the ethical theories discussed in the text concerning the justification of capital punishment not only for crimes such as kidnapping, but also for cases of first-degree murder by a person already sentenced to imprisonment for the duration of his life for another crime. Explain which of the ethical theories you think provides the most reasonable position on capital punishment.

19. Morally evaluate acts of the following kinds by the standard stated on p. 321. If you think that the situation in some cases must be given in greater detail, you may make reasonable stipulations. Does the standard prescribe anything morally repugnant in these cases? If so, should we reject the standard? Explain.

  • Premarital sexual intercourse.
  • Refusing to serve in a war which you consider unjust.
  • Failing to report a friend whom you have seen cheating.
  • Retaliating in kind to someone who has harmed you.
  • Stealing from a large, impersonal institution, such as a university or bookstore.

20. In light of your answer to question 19 and the many other factors discussed in the text, do you think the ethical standard stated on p. 321 is justified? Explain your answer. If you think not, do you think some other standard is justified or do you conclude that we should accept ethical skepticism? Justify your conclusions.

21. Ethical standards are used not only for moral evaluation of the actions of individuals, but also for moral evaluation of the laws of governments and societies. Is there some way to adapt the utilitarian Kantian principle to apply to laws? For example, consider and critically evaluate the following which is based on the thesis that no law should prescribe that any person or group of persons is to be treated as mere means, and so the basic prescription is that laws, and indeed systems of laws, should be fair:

A law (system of laws) is morally just if and only if (a) what is prescribes is fair to all persons and groups to whom it applies, and (b) it is a utilitarian rule (system of rules) which is consistent with condition (a).

22. Consider this objection to the utilitarian Kantian standard:

An ethical standard that cannot be effectively used is worthless and ought to be rejected. The utilitarian Kantian standard is easily seen to be worthless, because it is so complex that in the attempt to actually use it to decide what to do, one would be caught in endless calculation. Hence, we ought to reject this standard.
Is this a reasonable criticism of the utilitarian Kantian standard? Defend your answer.