You are hereUtilitarianist Ethics
This paper has benefited from the complaints, exclamations, objections, and obscenities of posters to rec.arts.books and alt.postmodern.
Working Toward Good Results
I suspect that most Westerners at the end of the twentieth century would subscribe to the idea that when you do something, it's a good idea to consider what the results of your action will be. That is, if without mentioning "ethics" or "morality" you ask someone "Do you think it's a good idea, when you do something, to consider what the results of your action will be?" you will probably get an affirmative response. The question allows for our famous postmodern "apathy" in not specifying that you need ever be doing anything. The question is only about how you should behave if you do happen to do something. Of course, in ethics, doing nothing is as significant an action as any other (and "non-actions" is a concept I will criticize below), but here we are not talking about ethics. We are being intentionally vague. Some people may reply in the negative, saying that they prefer to be surprised by results, for example. And, given our vagueness, who could blame them? But I suspect that negative responses will be a distinct minority, and that only a very small minority of these will take the form of denying the relevance of results to the goodness of an action on the grounds that something else entirely determines that goodness.
I'm confident that many would go on to agree with the idea that it's best to do those things which you think will have the best results. Some will probably even find this to be almost a truism, almost not worth saying. What possible alternative to it could be proposed? Surely acting so as to produce bad or mediocre results, far from being ideal, would be bad or mediocre. Who would recommend THAT? And why in the world would one advocate acting without any thought for the results of actions? I expect that many people will answer these questions without thinking about "goodness" in an ethical sense. Instead they will think about other measures of success: completing projects, creating art works, winning games, earning profits. With these things in mind, they will assert that it is OF COURSE best to act so as to produce the best results, just as it is worst to act so as to produce the worst results, fastest to act so as to produce the fastest results, meanest to act so as to produce the meanest results, most beautiful to act so as to produce the most beautiful results, and so on.
Consequentialism and Utilitarianism
But something changes as soon as we specify that we are talking about the "field" of ethics, that is to say: about all actions, not just actions in a specific business or sport or research project. Certainly the set of all actions includes these actions. In fact, it may consist of nothing but the composite collection of all such actions if we extend the list of specific types of actions indefinitely to include personal relationships, dinner recipes, philanthropic missions, animal training sessions, computer programming tasks, and so on. If it is best to work for the best results in each of these, surely it is best to work for the best results in all of them. This is not a common view, not if the word "ethics" is used. Reviewing discussions that are explicitly about "ethics" or "morality" will reveal this to be a very unpopular view. Ethics is not commonly conceived of as the summation of all behavior. Ethics is seen sometimes by some as one way of approaching all behavior, a way which may conflict with other ways, with other definitions of best. Thus repairing a bicycle and doing so in a particular way are not good or bad ethical choices, are not intended to maximize happiness or not, are not aimed (or not) at any other end conceived of as ethical. Ethics is restricted to questions of the law, of honesty, of sex, and of the latest breakthroughs (uses of the newest technologies and adaptation to changes imposed on a society from without). Ethics can be applied to bicycle repair if one is charging money for it, but ethics does not encompass every aspect of bicycle repair - not in most people's imagination.
When we speak of "ethics," and ask whether it is best to act so as to achieve the best results, a great many dissenters will almost certainly appear. This seems odd when we consider that in producing "the Good" (producing the Good is a good generic definition of ethics) it ought to be more important than anywhere else (if there is somewhere else) to produce the best results we can, to "make the world a better place," and to make it the best place we can, and not to be satisfied with, advocate, or take pride in something less or other. And there are people who have told me that they do want to produce the best state of affairs possible, but who contradict that statement when specific questions are discussed. Some people assent to the goal of producing the best results before they have thought much about it, before realizing that they do not really mean it.
Other people not only do not act so as to produce the best results, but they do not even aspire to do so. In fact, a lot of people consider themselves ethically superior to those who do so aspire. And in our books and articles on ethics, producing the best results - far from being considered an obvious starting point - is generally debated, if not scorned, as a "theory" known as "consequentialism." Where "the best results" is understood to mean the greatest amount of happiness for the greatest number of people, the theory is called utilitarianism (unless the book or article is about politics, in which case it is called democracy). Utilitarianism is thus a specific type of consequentialism.
Because utilitarianism is virtually the only type of consequentialism ever discussed, the two terms (utilitarianism and consequentialism) are sometimes used interchangeably. In this article I will be comparing what I see as advantages with utilitarianism with what I see as disadvantages with nonconsequentialism. It would be the task of another paper to compare utilitarianism with another consequentialism, one which makes the goal of actions, for example, the maximization of sadness or of the rapidity of space colonization or some combination of objectives, rather than the maximization of happiness. It is important, however, neither to understand "happiness" too narrowly (shading into the the term "pleasure" of which opponents of utilitarianism are so fond), nor quite to see it as shorthand for anything at all that anyone desires. If someone advocates a project for a society to accomplish despite the hardship involved, this can be understood as advancing his happiness. But the wishes of everyone else in the society must be considered as well. If someone says "What makes me happy is obeying the Bible," he is not discussing utilitarianist happiness. He is trying to please a book, and not humans.
Philosophy professors who write about ethics often seem to share many popular views but to diverge from common thinking in some details. I will try to be clear in the rest of this paper when I am talking about views that seem to be common in the general population and when about "common views" that are common only among philosophy professors.
Among philosophy professors, the distinction between consequentialism and nonconsequentialism is widely accepted. Many nonphilosophers deem utilitarianism offensive while also claiming to be aiming for the best results. Often, when asked, such people call their view consequentialist. But they do not specify what consequences they are aiming for. Or, when they do specify some goal (such as protecting liberty or respecting the institution of marriage), it seems applicable only to some small range of actions and liable to conflict with a great multitude of other goals they are bound to advocate as well. Usually these goals can better be described as nonconsequentialist.
The distinction between consequentialism and nonconsequentialism rests on that between actions and results. In many cases this is unproblematic. A utilitarian might decide whether to shoot an attacker based on whether he thinks the harm inflicted and any psychological harm involved is outweighed by the benefits to himself and others. A nonconsequentialist might decide based on a strict ban on violence or a "Right" to self-protection. These bases involve actions, not results.
But in other cases, the distinction is not as clear. In weighing whether to take a break from work in order to play with his dog, a utilitarian may consider not just the resulting health of the dog or effects on his work, but (even especially) the pleasure of playing with the dog. This playing is both action and result.
The same point might be made about the case of deciding whether to "commit adultery." A utilitarian might find himself considering the pleasure and perhaps simultaneous regret potentially involved in the action, as well as the risks of further harm, the damage to his habits of honesty, the damage to trust in his society, potential benefits as well as harms to his marriage, etc. He might even - and why not? - find himself taking into account the displeasure his action could cause in close friends holding nonconsequentialist views about the horrible sinfulness of adultery. More than that, he might recognize certain vestigial nonconsequentialist views of a similar sort in himself. I can see no way for him to avoid taking these into account, even while working to abandon them.
Such views could also be simply utilitarian rules of thumb, habits of regular behavior. If a utilitarian is offered the chance to save a thousand lives if he will agree to torture one little girl, it is entirely possible that he will conclude that he must do it. [But let's make clear that such a situation never arises in real life and has been invented for philosophy courses by opponents of secular thinking.] But the effects this will have on his sanity and his future actions, as well as on the actions of those who learn about it, must be taken seriously. Utilitarianism can require gritting one's teeth and proceeding with a distasteful action to produce good results, but it is important not to understand results too narrowly. Being a good utilitarian does not mean behaving as if everyone else already were a perfect utilitarian. On the other hand, it also does not mean simply deferring to every nonconsequentialist habit of thought one can detect in oneself, allowing the slightest psychological strain to outweigh enormous benefits. I will have more to say on these topics below.
Some people have a hard time thinking in terms of the consequentialist - nonconsequentialist distinction. They see what they do as aimed at producing good results, but also find utilitarianism highly offensive. Often, as I've said, many of the "results" they have in mind are better described as actions. But even when they are clearly results, they are such a variety of goals that any way of combining them into a workable system seems hopeless. Simplicity is a great advantage for utilitarianism, which works with a single goal rather than a collection of rights and claims and duties and obligations, many of which conflict with others in ways nearly impossible for anyone to sort out into a workable hierarchy of trumps and compromises. Such ad hoc and open-ended collections of goals, on close inspection, have little to do with creating a better world, and much to do with performing and abstaining from certain actions, even if their proponents protest: "But I am aiming for results."
In some situations in which a commandment not to lie comes into conflict with a commandment not to kill, happiness will be furthered by letting the first trump the second, in others vice versa. Therefore a rule assigning priority to one or the other will sometimes increase suffering.
Often the same person who will subscribe to the goal of producing the best results, will employ the term "utilitarianism" as if it were a harsh criticism. In part this may be the result of confusing meanings of "utilitarian". "Utilitarian" is used to mean not just an adherent of utilitarianism or an action in accord with that philosophy, but also something that is devoid of beauty or ornamentation, something that serves a function, but does so gracelessly and simplistically. This second meaning has nothing to do with utilitarianism. If we were to employ the term utilitarianist, this confusion might be lessened.
But many people, explicitly or implicitly, will not subscribe to the goal of producing the best results, will not be consequentialists. They prefer to be religious or philosophical, to follow divine guidance, or natural law, or common sense, or instinctive primal truths, or the dictates of their superegos. They do not believe that it is up to people to choose how to behave for the sake of people. They imagine that there is something else to which they need to be obedient. As a result, a discussion of how to produce the best results in any given situation cannot even be begun. There is bound to be substantial disagreement among utilitarianists on any topic, but we cannot devote our energy to resolving such disagreements until the commandments and natural traits and achetypal urges and moral intuitions of the nonconsequentialists have been addressed.
Because both consequentialists and nonconsequentialists can arrive at just about any position, it is tempting to suggest that ethical theory is of no importance at all. But this would be a grave mistake. A consequentialist who disagrees with you can try to convince you, can show you what his concerns are. You can enter into a lengthy discussion with him, each showcasing the concerns you find most moving. A debate with a nonconsequentialist takes the futile and interminable form of "Is not," "Is so," "Is not," etc.
I am overstating the contrast, but only slightly. Someone COULD be persuaded by a nonconsequentialist dogma. It does happen. If it did not, there would be no reason to write against nonconsequentialism. But any two or more dogmas are mutually refuting, and with an amazing frequency interminable debates like that described above result. I will mention a few examples:
A utilitarianist who favors banning gay marriages will explain his position in terms of the health and happiness of the spouses, or of the care of children, or whatever his concerns may be. Each such point can be discussed at length, the facts disputed, the values reevaluated. A nonconsequentialist will ban gay marriages because of a passage in a book somewhere, or because of what is "unnatural." This is a conversation-stopper. I would much rather debate the utilitarianist. I will have more to say about disputing facts and values below.
Similarly, a utilitarianist who favors regressive taxation and industrial deregulation will make his case in terms of overall wealth generated or the imminent mass-migration to the moon, or whatever his motivations may be. Arguments for and against such premises of fact and value can easily be imagined. A nonconsequentialist will appeal to claims of desert, or faith in the market, or rights to free enterprise. What can be said for or against those things? One person claims that "rights to free enterprise exist," and another denies it. What persuasive arguments can be martialled for either side? Such rights can be translated into desires or, to use the term I have begun with, values. But, then, we will almost inevitably ask why such things are valued, and will slip into an overt consequentialism.
A utilitarianist who opposes lying to protect someone will appeal to the corrosive effects of lying on civic trust and on the habits of the liar, or to whatever concerns him. A nonconsequentialist will appeal to Augustine or Kant or to the sloppy skeptical view that because "Truth" can't be known, truthfulness is pointless. It seems clear which type of position makes a more constructive contribution to a public debate. The idea that we simply must not ever lie can either (1) be supported as a rule of thumb on grounds of the great benefit to happiness of living in a trusting environment, in which case the rule must be open to exceptions where lying increases happiness, or (2) such a ban can be supported on some extra-human grounds: some entity can command us not to lie even when not lying causes great harm. (Kant famously refused to condone lying to a would-be murderer who asks if his intended victim is in your house). Some will accept such a theistic dictum, and some will not. No richer or subtler understanding of the dilemmas, and no agreement, will ever be reached as long as such arguments are recognized.
Here is a quintessentially "ethical" topic in the opinion of those who would make ethics a particular academic specialty and call in one ethicist with every dozen scientists or engineers. A utilitarianist will discuss this topic with reference to such considerations as: pain, loss of power, fear of uncertainty, humiliation, and potential for abuse of permissive laws. A utilitarianist may even take into account the likely reactions of loved-ones holding nonconsequentialist views on the topic. A utilitarianist will be appealing to both facts and values, and - in the case of polls of patients' and doctors' views - both at once. Facts such as the recovery history of patients with a particular disease can be disputed and augmented in the usual ways. Values, such as a concern with avoiding humiliation, can be given weight by describing or showing in detail through any medium what humiliation and its alleviation are like. That there should be nothing mysterious about values, that they arise as do any urges or desires (and, in fact, are such) must be the topic for another paper.
A nonconsequentialist will appeal to a "right" to independence, or a categorical imperative not to commit suicide, or something of the sort. Such assertions will close off discussion. Someone who makes such an assertion can read a dozen lengthy accounts of assisted suicides permitted and denied, and will not possibly revise his decision - because it is not his; it belongs to God or Kant or the moral realm or the natural law. Does not that fact itself oppose a common belief? Is it not disconcerting to think that an "ethicist" could learn of, say, the Nazi holocaust, and not change his views on how it is best to behave? Is there not something radically undemocratic in such a refusal to learn, to grow, to make ethics a product of on-going public debate?
Complaints with Utilitarianism
But nonconsequentialists do not view ethics as a matter of public debate. They see it, rather, as an undemocratic authoritarian system. And they have their own complaints with utilitarianism. I'll mention five of these.
1. It's often suggested that, for example, the rich could get a whole lot of happiness out of abusing the poor, or a great many people could get happiness out of seeing one person tortured. This is supposed to show that utilitarianism reaches conclusions out of accord with our "intuitions," as if the task of ethical theory is to produce some verbiage to back up what we already believe. (It's also claimed that a person brought up in luxury might need greater wealth to maintain happiness than does another, but many nonconsequentialists themselves subscribe to this dishonest position.)
If ethics is just the production of verbiage in accord with what most people already think, then ethics becomes no more than a pretentious argumentum ad populum. The ethicist takes a poll, and then pontificates. Ethics as politics. I prefer to imagine the ethicist, not to mention the politician - and why can't they be the same? - as arguing for new beliefs. Any new belief will have both similarities and differences with old beliefs, and someone will always be able to highlight similarities to claim that the "essence" hasn't changed, or - as in the case of John Rawls - that a balance has been struck between innovation and "intuition." But this is window dressing. The important thing about a change for the better is that it is a change for the better.
A utilitarianist will often advocate radical changes. Such advocacy has produced a lot in the past few centuries. A utilitarianist will not worry about not fitting with someone's "moral intuitions." But the sorts of situations usually cited as complaints with utilitarianism are very rare. And the sort of suffering utilitarianism is imagined to inflict on those not ready for changes is usually overestimated. As I've said, a utilitarianist must take into consideration the mental well-being of nonconsequentialists without advocating the preservation of nonconsequentialist tendencies.
If it were concluded that the greatest happiness really were produced by abuse or torture, then the utilitarianist would, indeed, advocate those practices in the cases having that result. No doubt some utilitarianists do so, just as do some nonconsequentialists. But some basic rules of thumb common to much utilitarianist thinking argue against such positions. These include: Eliminating pain is more important than producing pleasure; Improving the condition of many people is usually preferable to improving that of a few or one; Slight pain or pleasure in many cannot outweigh tremendous pain or pleasure in a few or one.
The poor usually outnumber the rich, and their suffering needs to be alleviated more than the happiness of the rich needs to be augmented. And, of course, many of the rich will probably want to give priority to helping the poor. If a rich person is a utilitarianist, he will not see this "ethical choice" as in conflict with his "personal preference," any more than he will see it as justifying utilitarianism by according with preexisting "intuitions." In addition, the political and cultural advantages of an egalitarian society keep this from being a zero-sum game. A democratic culture produces happiness above and beyond the sum of its members financial incomes.
Similarly, many will suffer upon seeing someone tortured for sport. The tremendous pain and humiliation suffered by the object of sadists' entertainment can be seen as outweighing a good deal of pleasure in the sadists. But there are larger utilitarianist reasons to oppose such torture. Creating a society that will find happiness in it is undesirable on utilitarianist grounds. Creating a society in which someone is chosen to be tortured will be discomforting even to selfish sadists who don't want to be next. And this sadism will have undesirable repercussions in the behavior of all involved. People in such a society will be more open to treating each other cruelly.
2. Another way of putting this complaint is: "There are some things one simply does not do." There are those who argue that one should not kill a person to save ten people (even including that one), on the grounds that one just plain must not ever kill. Here we have a failure of nerve elevated to an armchair theory. This "non-action" is the equivalent of nine murders. I don't know how to make that point any clearer than it is on its own. It's called a "non-action" because it is not a particular action, but it is another action, and one with serious results.
3. It is also suggested that utilitarianism is simply too hard. Every moment you spend watching a movie you could spend working to help the poor. Every present you buy for a family member could have gone to someone more needy. Utilitarianism demands that you make such sacrifices. This is considered too demanding by many people, and therefore "wrong."
Certainly few if any people who try to produce the best results possible are content with their success. I call myself a utilitarianist, and am certain I could do far better. But, unfortunately, the purpose of ethical theory is not the building of self-esteem. Doing ethics is determining what would be best. If the results indicate that only one person in every billion can reasonably be called decent or civilized, so be it. To introduce a line-item veto for whatever we find challenging, is simply to no longer be doing ethics. Ethics is the creation of an ideal. A career of philanthropy so rigorous and demanding that it kills people by age 30 is probably not an ideal.
On the other hand, it could be. If the results appear to be great enough, someone may conclude that he needs to give his life in that manner if he is to maximize happiness. There is no mathematical calculation for determining such things. Utilitarianism amounts only to the assertion of a desire to think seriously about them, and to be brave enough to accept a difficult conclusion. Claiming that. for example, leaving home (to go philanthropize) is always wrong (or right) is an avoiding of the question, and will not convince or enlighten or set at ease even many of those who maintain it. The answer is always a rough approximation. Calculating numerical quantities in units called utils, as some utilitarianists have done, exhibits a foolish physics-envy or longing for certainty that utilitarianists usually criticize in nonconsequentialists.
A life that contributes the most possible happiness to the liver AND to others is a utilitarianist ideal. No individual is worth extra, but each individual is in a peculiarly advantageous position to improve his own happiness. Such a utilitarianist ideal should be aimed for. Unless ethics develops such ideals, it will just describe what we currently do. Ethics will then be distinguishable from journalism only by its lack of perception.
Imagine a friend has given you a large loan of money. Now you discover an opportunity to do a great many people a great deal of good, though you will be unable to pay your friend back. You will also be doing something you greatly enjoy, so that your philanthropy cannot be (nonsensically) praised to your friend as a great sacrifice, in order to convince him of your regret. You will have to consider the damage your action will do to your friend, to his faith in people, to others' faith in promises. You will have to guess at his ability to understand and believe and forgive you (if that would still be needed once he understood). If the good you will do is great enough, you will have to do the evil of betraying your friend. Such decisions are the difficult stuff of human life. Even more difficult is the decision the betrayed and uncomprehending friend must make of whether to forgive.
On this topic I recommend the discussion of Abraham's near-sacrifice of his son, Isaac, in "The Gift of Death," by Jacques Derrida, by no means a utilitarianist book. Derrida points out that your sense of duty to your friend is itself a betrayal of strangers:
"How would you ever justify the fact that you sacrifice all the cats in the world to the cat that you feed at home every morning for years, whereas other cats die of hunger at every instant? Not to mention other people? How would you justify your presence here speaking one particular language, rather than there speaking to others in another language? And yet we also do our duty by behaving thus. . . . "
And how, we might add, does one know when one is acting in good faith in avoiding stress or anxiety that will possibly debilitate one, in allowing for one's vestigial nonconsequentialist leanings, in preferring the near and ordinary tasks to the extreme and unusual, the easy to the heroic? Needless to say, the answer is that one does one's best.
Abraham's story (or Isaac's, since I believe he has the harder role) can be secularized by treating God's Will as Abraham's belief in some good he can do for many strangers, or by treating it - as Derrida does - as the maintaining by Abraham of a secret desire to do some good without any desire to be praised for it. I prefer the former way, since - following David Hume - I see very little harm, if any, and often much good, in taking pride in good behavior. Derrida writes of Abraham:
"Things are such that this man would surely be condemned by any civilized society. On the other hand, the smooth functioning of such a society, the monotonous complacency of its discourses on morality, politics, and the law, and the exercise of its rights (whether public, private, national or international), are in no way impaired by the fact that, because of the structure of the laws of the market that society has instituted and controls, because of the mechanisms of external debt and other similar inequities, that same 'society' PUTS TO death or (but failing to help someone in distress accounts for only a minor difference) ALLOWS to die of hunger and disease tens of millions of children (those neighbors or fellow humans that ethics or the discourse of the rights of man refer to) without any moral or legal tribunal ever being considered competent to judge such a sacrifice, the sacrifice of others to avoid being sacrificed oneself. Not only is it true that such a society participates in this incalculable sacrifice, it actually organizes it. . . ."
4. "We need to respect each and every individual." If a utilitarianist doctor can murder one healthy patient and thereby save five patients in need of organ transplants, and if he knows that he will never be found out, that no one will fear visiting his office, that the victim has no friends or loved-ones or valuable talents, that the doctor and his future actions will not be damaged greatly by this action or by the keeping of the secret, he may be bound to do it. In this it is said that the doctor lacks respect for the individual. I beg to differ. He is demonstrating respect for FIVE individuals. What he lacks respect for is the "natural course of things". He fails to adopt the usual accepting view of natural death. The more praise to him! But it should be recognized that we are talking about a near-impossibility. Establishing a PRACTICE of such actions would be highly destructive of trust and peace of mind. A guarantee of secrecy can probably never be had. And, in addition, the doctor needs to take into consideration possible deleterious effects on his own character. Will this action make him arrogant in future dealings with people's fates? All such psychological results must be carefully weighed before a final decision is made, even should such a nearly-impossible situation arise.
The same applies to the thinking of a judge who is considering convicting an innocent person for the good of society. Even should he become convinced that society would benefit sufficiently, he would have to consider the difficulty of guaranteeing the secret, and the negative effects on himself of taking such an action and keeping such a secret. A utilitarianist would not advise eagerly looking for an opportunity to perform such an heroic act, but would rather subscribe to the idea that one should never convict the innocent, labeling this a rule of thumb instead of a divine commandment or categorical imperative.
As quickly as possible I will now perform the almost obligatory task of explaining where I stand in relation to John Rawls' well-known paper "Two Concepts of Rules." At issue here is a debate between retributivist punishment and punishment aimed at reducing crime and the ill-effects of crime through such specific concerns as deterrence, restitution, protection, and rehabilitation.
Rawls makes two excellent points in this paper to which I will come in a moment. With his central thesis I cannot agree. Rawls distinguishes between "justifying" (and shaping) an institution and justifying and shaping a particular case, ie a small piece of said institution. I take this distinction to be as misguided as it initially looks. If all the little bits of some practice are given one justification, how can all the little bits of that practice (this time thought of together as a whole) be given a contradictory justification? Rawls' argument appears more convincing than I've made it sound because he quotes a perfectly believable conversation which seems to support it:
"We might try to get clear about this distinction by imagining how a father might answer the question of his son. Suppose the son asks, 'Why was J put in jail yesterday?' The father answers, 'Because he robbed the bank at B. He was duly tried and found guilty. That's why he was put in jail yesterday.' But suppose the son had asked a different question, namely, 'Why do people put other people in jail?' Then the father might answer, 'To protect good people from bad people' or 'To stop people from doing things that would make it uneasy for all of us; for otherwise we wouldn't be able to go to bed at night and sleep in peace.'"
Rawls believes that he can reconcile retributivism with utilitarianism, both of which he finds valuable, by allowing retributivism to explain the particular case and utilitarianism to explain the institution. (And it might be that a "rule utilitarian" could be satisfied with this. A rule utilitarian is a hybrid creature who favors strict rules that maximize happiness as much as any strict rules can, but which must be followed even when they do harm.) It would seem to follow that the legislator should be a utilitarianist and the jurist a retributivist. I, of course, would like them both to be utilitarianists (act utilitarians, that is: not rule utilitarians), and I believe the difference to be meaningful.
(Of course no set of strict rules, however encyclopaedic, can cover more than a small percentage of situations, and then conflicting rules need to somehow be reconciled.)
Let's look at another conversation.
Son: Why was J put in jail yesterday?
Father: Because he robbed the bank at B.
Son: And why do we put someone in jail if he robs the bank at B?
Father: Because we don't want him to rob any other banks or commit any other crimes whatsoever, and because someone else may think twice about robbing a bank if he hears about what happened to J, and because we want to make J regret what he did, we want to convince him to change his ways, maybe by employing him in jail and later out of jail and giving some of what he earns to the bank-tellers and customers he frightened, and to the bank which he put to so much trouble, and to the government which he put to even more.
The key difference between the two conversations (No, I don't mean the father's loquacity!) is that in this new version it is made clearer that the father's first answer is a partial one, and his second answer a more complete one. The first answer assumes that the son knows the second. With the second question the father is disabused of this assumption. The son might go on to ask a third question, about why we don't want banks robbed, why we want J to change his ways, why we don't want to be frightened, etc. Neither Rawls nor I extends the conversation that far because we both take the answers to be uncontroversial and irrelevant (though I do think rehabilitation and restitution are undervalued). But if the second answer is utilitarianist, as it is in both conversations (in Rawls' version protection or deterrence, in mine protection and deterrence and restitution and rehabilitation), then so is the first. There is no chasm here between two discourses, but rather a variation in the degree of detail requested. A policy of putting bank robbers in jail can be perfectly utilitarianist. I don't think anyone has ever doubted that. But it should be understood that, for a utilitarianist, such a policy is only a policy and not a final answer to any normative question, since a bare statement of the policy refers to no general normative good. Certainly at one level putting bank robbers in jail is a good, while at another deterring potential criminals is a good, and at the very end of the story (the question to end all questioning) happiness is the all-encompassing good. The mistake is to suppose that (for a utilitarianist, at least an act utilitarianist) the buck stops anywhere other than at happiness. If jurors and judges mechanically fulfilled the exact will of legislators, I would be willing to ask that legislators be utilitarianist and jurists be whatever they liked. But sentencing, and convicting (which cannot be fully detached from sentencing), are not mechanical. A retributivist may accept the explanation "because he robbed the bank at B." But what will he do if J is a minor or mentally deficient or a recidivist? What will he do if J only attempted to rob the bank? The retributivist's theory will tell him to alter his sentence or eliminate it (and do something other than punishment) in view of such complications. It is not possible that this will involve only some extra words (and let us assume they are coherent, self-consistent ones) and perhaps some extra litigation, but no other final results than the utilitarianist. Rather, the retributivist's decision will fail to promote rehabilitation and restitution in many cases in which the utilitarianist finds them relevant. A utilitarianist jurist of my stripe can certainly accept the explanation "because he robbed the bank at B," but intends by this (implicitly) the second, lengthy, answer of the second conversation above, and will without hesitation modify the sentence in light of the four concerns listed there and any relevant details of the case.
I said that Rawls makes two excellent points. The first is, in fact, the degree to which a retributivist could accept a utilitarianist system. It might even be that a retributivist could approve of a system of punishment (in my sense of the word) designed in accordance with my theory. Without a doubt the vast majority of the actions produced by such a system would be reactions to specific criminal acts. If that is enough, in some opinions, to make it retributivist, then there remains only the glitch regarding forcibly institutionalizing the non-criminal dangerous insane. Rawls does not address this dilemma despite the fact that his utilitarianist is concerned with protection. Instead he phrases this question in terms of the "innocent," without specifying the dangerous and mentally-deficient innocent.
This leads to the second point I want to highlight. Rawls refutes the idea that utilitarianism would punish (in the inflict-harm-on sense, if perhaps not the on-a-convicted-criminal sense, if "convicted" is read as "honestly convicted") the innocent by pointing out how disadvantageous it would be to set up an institution guided by such a policy. Rawls means to tie this in to his notion of "the institution" as opposed to "the particular case." And that is a shortcoming of his way of stating this refutation. He leaves open a doubt about occasional cases, or even just that one extreme exceptional case intended to be kept as secret as can be. For this reason I prefer to make central to this refutation the observation that no secret can be guaranteed. (This is in addition to the fact that punishing the innocent is contrary to rehabilitation, and in itself constitutes a harm.) Even if the ruling class is cleanly separated from the ruled (ignoring the non-utilitarianist nature of such an arrangement) so that even a policy of punishing the innocent does not put any policy-makers (or their loved-ones?) at risk, and even if such a policy is rejected but one particular case considered as a possible exception, the secret could not be guaranteed. And the risk would be heightened by the harm done when confidence in a government's statements is lost. In addition, harm would already have been done to the ethical habits of the policy-makers. This is the same argument I give for why I bother to vote, as opposed to just campaigning and pretending to vote, knowing quite superiorly that my one vote makes no difference. I cannot guarantee that I will keep the secret, and I cannot be sure of avoiding ill-effects from admitting dishonesty into my life. So Rawls is right, I believe, that an open policy of punishing (still in the do-harm-to sense) the innocent cannot be produced by utilitarianism. I maintain, in addition, that a secret policy would only be utilitarianist if, per impossibile, a secret could be strictly guaranteed, and if dishonesty were not so insidious. But perhaps there is still an exception to be found, namely a decision made by a judge who knows he has only hours to live, and who is convinced that the most good he can do with that remaining time is to concentrate all his energy on carefully keeping the secret, a secret which he is certain will die with him. (This last criterion is just barely imaginable.) In a world of perfectly utilitarianist judges, and no shortage of them, a policy could be adopted of removing from the bench any judge known to be near death, and disqualifying any decision made by a judge who had known he was near death (this for reasons of public peace of mind). But we can imagine a clever-enough utilitarianist judge to have known he was near death (a sudden discovery made after hearing all the evidence in the case, leaving him no time to retire) but to have made his death look like an unforeseen accident. His decision, punishing an innocent for the greater good, would stand unquestioned. And so I leave it, and not the least bit shamefacedly. It is better that one innocent suffer than that another suffer worse, or more than one suffer to a considerable degree. (It has reasonably been argued that it is generally worse for one to suffer severely than for many to suffer a tiny bit.) I will add only that spending a significant amount of time thinking about such highly improbable cases is very likely NOT desirable, as the most likely result of such thinking is the erroneous discovery by judges of cases they imagine to call for similar tough-minded heroics.
It is also worth noting that utilitarianism is not alone in reaching the above conclusion. Those forms of retributivism which deal in terms of the satisfaction of grievances or resentment, in so far as they make this a consequentialist concern, must also punish the innocent in such a bizarre situation, for reasons of satisfaction rather than deterrence, if not for both, unless they include a constraint specifically to block such a conclusion.
5. It has also been suggested that utilitarianism is too vague, that it doesn't solve a single ethical question, but simply rephrases them all. This is true, but it's not a telling criticism. Utilitarianism is no more than the defining of a course of inquiry. Utilitarianism is the declaration: We will try to agree on what actions will lead to the greatest happiness for the most people. Of course, there will be rich disagreement in the course of that discussion. Perhaps the only specific point the utilitarianist need stand by is that people can be at least as happy with utilitarianism as with nonconsequentialism. Obviously some people can. Whether everyone can remains in doubt. As long as it remains in doubt, the utilitarianist is subject to charges of paternalism. He must take care not to impose his views on others "for their own good." But surely, in the absence of any third option, even paternalism is preferable to both paternalism and the blocking of any conversation. And surely dogmas without explanations are the more paternalistic than advice as to how to become (and make others) happier.
Consequences of Consequentialism
A lot of nonconsequentialist ethical talk seems to be a cover for a concern with consequences. And often the rules devised can be rephrased as utilitarianist rules of thumb to which exceptions might be made when needed. But in many cases the disagreement between utilitarianists and nonconsequentialists is substantial and serious.
Nonconsequentialists spend a lot of time trying to "ground" their ethical "beliefs" in facts. Utilitarianists see their ethical preferences as things they themselves add to facts. "Bombs are destructive, therefore we should never use bombs," is not a good argument for a utilitarianist who, following David Hume, will tell you that you cannot get an ought from an is. It is missing a preference or value or desire (pick your preferred, valued, desired term), such as a dislike of destruction. The fact that everybody lies does not excuse lying. The fact that gay people don't procreate together does not in itself mean that gay people should not marry. The facts that some people have darker skin than others, and that some books allow slavery, do not make slavery desirable.
Where do values come from? They come from the same place a desire for grapefruit juice comes from. They come from where a fit of coughing or an image of a palm tree comes from. That is: you "make them up." Let me throw back another question: Where else would they come from?
For utilitarianists, actions are to be ranked in order of desirability, preference, justifiability, appropriateness, excusability, acceptability, goodness, and so on, all of these terms being synonymous. For nonconsequentialists, these and many other similar terms all have different meanings, and actions are to be categorized in any number of confusing and conflicting schema.
To begin with, there is the ethical as opposed to the desirable. People speak of what they want to do as being something other than what they see as ethical. In rare cases this may make sense. I may want to go to the movies, but also want - and see as ethical - to volunteer at a soup kitchen. But this way of speaking suggests that ethics is something other than human doing, whereas it is just a particular type of desired action, namely that which involves our desire to maximize happiness (or to achieve some other ethical end).
And when nonconsequentialists begin to talk about "weakness of will," confusion really sets in. What you do simply is what you "will." You have not done what you wanted as opposed to what you "were supposed to do." You have done what you wanted rather than something else you also wanted to do.
Nonconsequentialists believe that some actions are "non-moral," not matters for ethical concern, perhaps (because) "private" or trivial. (A "private" act would be one magically isolated in time, having no effect on later acts.) And nonconsequentialists believe they can view choices either morally or "prudentially" / "strategically." This is a result of thinking of ethics as the concerns of others being imposed on a selfish individual. But individuals are not purely selfish. Often their interests are the interests of others. And for people to act purely selflessly would not often maximize happiness. Not only the desirability, but also the definability, of pure selflessness (or selfishness) is dubitable. No action is too trivial to be an ethical concern, and ethical concern is not the same thing as altruistic concern. We are each in a peculiarly advantageous position to increase our own happiness. There may be cases when we conclude that we must minimize such efforts to further our work for others. But such is not a choice made in advance in order to take up consideration of ethics.
Sometimes "non-moral" actions are called "permitted" as opposed to "required." Going for a walk in the woods is said to be permitted but not required. But a choice to walk in the woods is a choice between that and an infinite number of other actions. Either there is a preferable (more ethical, more optimific) choice available, or there isn't. That some things are important and others usually unharmful make good rules of thumb, but not good divisions for analyzing specific actions.
Nonconsequentialists also like to speak of some things as being good "in advance" or "prima facie" (as opposed to "after the fact"). Again, a rule of thumb can be a useful thing. But the fact that something is usually good does not make it a smidgen better in a case in which it is productive of pain and misery.
Similarly, nonconsequentialists speak of "beliefs" "in theory" or "in principle," as opposed to "in practice." This division ought to be largely abandoned. The words "in practice" can usually be deleted from a sentence without damaging its meaning. And sentences containing "in theory" can usually be deleted in their entirety without damaging a book.
There is a great deal of nonconsequentialist vocabulary of which we could easily be rid. This would include most talk of obligations, duties, and rights, as interacting forces of differing strengths. I find it far clearer to ask what will produce the most happiness, than to sort through various claims about my claims on others, their claims on me, my duties to them, their duties to me, our various rights to do things, our various rights to refrain from doing things, and so forth. All of this "nonsense on stilts" impedes serious discussion with futile and interminable disagreements, and valueless complications.
The myths to be abandoned are legion: that of "rights" as pseudo-factually extant things, that of retribution as the setting straight of some mysterious balance, that of a social contract the breaking of which subjects one to punishments the results of which need not be considered, "human nature" as a pseudo-factual and pseudo-normative thing/commandment both compelling and recommending actions, epistemology as some kind of basis for ethics, determinism as something in some kind of conflict with freedom and responsibility, the conscience or superego as an external judge, etc., etc. I am listing, not arguing, these points, for lack of space. Each has been argued well by many authors. Progress is being made in abandoning these myths. John Rawls has questioned the authority of "intuitions," suggesting that we try to balance our innovations with our intuitions. From there it is a short step to discounting the authority of intuitions altogether.
This is not to say that the utilitarianist should remake all his preferences and values on an hourly basis, or that he should act rashly or innovate for the sake of innovation. It is only to say that what one currently thinks is no more than what one currently thinks. Everything is open to change, including our choice of utilitarianism itself.
Utilitarianism is not as strange or cold or simple as people take it to be who long for nonconsequentialism. Should a group hand over an individual to be unjustly punished by an enemy, in order to save the rest of the group from a similar fate? Possibly. It depends on whether the rest of the group would be miserable after having done so, whether the enemy would impose the same type of bargain again in the future, whether the sort of loyalty which opposes doing so has significant advantages. This is a question to be debated within a rich and psychologically perceptive utilitarianism, not a test of utilitarianism's "validity."
Some divisions of types of actions can be incorporated into utilitarianism, while utilitarianism at the same time suggests diminishing them. For example, we make great distinctions between speaking, promising, and stating under oath. For the most part, we would be better off without these sorts of distinctions. We should almost always expect / require honesty.
The most important divisions are those which attempt to excuse one from responsibility, and utilitarianism can help to dissolve them. People are responsible (that is no more or less than to say that we are best off holding each other to be "responsible") for their actions, and also for who they are and for the emotions they have. There are no ghosts.
Being a good utilitarianist does not mean acting as if others were perfect utilitarianists. I cannot even imagine what that would mean. What's needed is a movement in the direction of perfect utilitarianism, which is not some fixed situation. It will always change, and always in the direction of greater happiness.