R G Frey. New Dictionary of the History of Ideas. Editor: Maryanne Cline Horowitz. Volume 6, Charles Scribner’s Sons, 2005.
Utilitarianism is the name of a group of ethical theories that judges the rightness of acts, choices, decisions, and policies by their consequences for human (and possibly animal) welfare. These theories have been widely influential among philosophers, economists, and political and social scientists, and, in the early twenty-first century, in the general area of applied or practical ethics. It would not be too much to say that utilitarian thinking undergirds one of the contending positions on virtually each of the issues in debate, whether to do with animal rights, euthanasia and physician-assisted suicide, health care coverage, or punishment. This is so, moreover, even as critics of utilitarianism as a normative ethical theory have become more numerous.
Classically, utilitarianism is the view that acts are right or wrong if they produce best consequences—that is, consequences with regard to human welfare that are at least as good as those of any alternative. This is the view normally associated with Jeremy Bentham (1748-1832), James Mill (1773-1836), and John Stuart Mill (1806-1873); with Henry Sidgwick (1838-1900) and G. E. (George Edward) Moore (1873-1958); more recently, with J. J. C. Smart (b. 1920) and Richard Mervyn Hare (1919-2002); and, more recently still, Peter Singer (b. 1946). It began, thus, in the nineteenth century in Britain, though traces of it can be found in earlier thinkers there, such as Francis Hutcheson (1694-1746), David Hume (1711-1776), and William Godwin (1756-1836). This classical version of the theory is also the view that contemporary critics, such as Amartya Sen (b. 1933), Bernard Williams (1929-2003), and Samuel Scheffler, attack. In fact, different versions of utilitarianism have been distinguished from classical-or act-utilitarianism, such as rule-utilitarianism, utilitarian generalization, motive utilitarianism, and so on, though it remains hotly debated how far these versions are ultimately distinct from the utilitarianism (and, to critics, from the objections directed against it).
Act-utilitarianism as it has come down to the present has three main components, and each has generated discussion in its own right. These are the consequence component, the value component, and the range component.
The consequence component maintains that rightness is tied to the production of good consequences (specifically, to consequences better than those of any alternative or best consequences). It may be argued that something in addition to consequences helps to determine an act’s rightness, but consequentialism is the view that consequences alone determine this; act-utilitarians are consequentialists. Consequentialism in the early 2000s is the object of a good deal of criticism, though even in the nineteenth century it was sometimes controversial, as in the debates between Cardinal Newman (1801-1890) and Charles Kingsley (1801-1890). To some it is self-defeating, in that an effort to produce best consequences on each occasion may fail to produce best consequences overall. To others it seems inherently evil because it cannot exclude certain acts as intrinsically wrong or wrong independently of consequences (for example, lying). More recent criticisms have been equally strong. To John Rawls (b. 1921) impersonal accounts of rightness, such as best consequences, fail to take seriously the separateness of persons and so fail to treat people as autonomous individuals with their own individuality, plans, and worth. To Bernard Williams impersonal accounts of rightness run the risk of severing one from one’s integrity, in the sense that pursuit of best consequences may not be compatible with pursuit of one’s own projects, commitments, and relationships. Of course, some act-utilitarians, such as R. M. Hare (1919-2002), have resisted the Rawls-Williams objections, but others influenced by the theory, such as L. Wayne Sumner, have responded by trying to build a scheme of individual moral rights into the theory on the ground that doing so gives one the best chance of producing best consequences. Still more recent criticisms move in an epistemological direction. One cannot know at the time of acting, it is said, what (all) the consequences of one’s proposed act will be. Some have seen this as demanding a role for rules as a guide; others, such as Peter Railton, have viewed it as a reason for not trying to act upon consequentialism in the first place and so for not using this account of rightness as if it were a decision procedure. He suggests a role for the concept of (good) character in determining how individuals shall act.
The value component has from Bentham onward spurred debate. It maintains that consequences are to be assessed by some standard of intrinsic goodness, the presence of which in the world is to be maximized. In the early twenty-first century this good in the case of act-utilitarianism is construed to be human welfare, but how “welfare” is to be understood is contentious. Thus, Bentham is a hedonist, maintaining that all and only pleasure (which his felicific calculus was supposed to be able to calculate) is intrinsically good; not surprisingly, Friedrich Nietzsche (1844-1900) referred to utilitarianism as a pig philosophy. John Stuart Mill spoke of happiness and pleasure and tried to introduce a distinction between higher and lower pleasures, and the mentor of the Bloomsbury Group, G. E. Moore, maintained that other things, such as beauty and friendship, as well as pleasure and/or happiness, were good in themselves. Most actutilitarians have followed Moore in moving away from hedonism and from the emphasis upon happiness.
Three types of issues have dominated recent critical discussion of the value component. First, there has been a movement away from the old mental-state accounts of goodness (e.g., pleasure) to desire-or preference-satisfaction accounts. Economists have helped spur this development. It is unclear, however, upon which desires one is to focus. Problems to do with present or future desires have led theorists to an emphasis upon informed desires—that is, those desires one would have if one were fully informed, detached, free from bias and pressure, and so on. The thought seems to be that, in the appropriate circumstances, informed desires become actual, and so those desires upon which one (rightly) acts. (This thought, so construed, can then be wedded, in economics, to “revealed preference” accounts of goodness, which are widespread.) Second, there has been a movement away from agent-neutral values toward agent-relative ones. Values, it is said, are subjective, in the sense of being agent-relative; they are the values of agents. Utilitarianism, however, requires that desires be aggregated, weighed, and balanced in terms of, for example, agent-neutral concerns to do with the amount of desire-satisfaction produced, irrespective of which individual agents obtain that satisfaction. The question then arises whether any particular agent has reason to value pursuit of overall desire-satisfaction, in addition to or in place of his or her own. Third, if one focuses upon desires, then a question about moral filtering devices obviously arises. If one takes into account the desires of all those party to a situation, does one filter out the desires of evildoers? This seems contrary to the spirit of utilitarianism and its emphasis upon agent-neutral values. But if one does not filter out the desires of evildoers, does this mean that the utilitarian weighs and balances their desires, according to strength, with the desires of those against whom evildoers act?
The range component maintains that all those affected by the act are to have their desires taken into account. This has led notoriously, from the nineteenth century onward, to the problem of interpersonal comparisons of (pains and pleasures) or desire-satisfaction. One can only maximize desire-satisfaction across all those affected by the act if one can compare the effect of the act upon the desire sets of each individual involved, judge the strength and extent of that effect, and then compare the different results. How one does this, what measures or scales of comparison are used, is not obvious. Another problem with the range component that stirs debate in the early twenty-first century has been whether to include animals within the scope of utilitarianism. Bentham did, on the grounds that animals could suffer. But if informed desires are the focus of the value component, can animals have informed desires? The point is not obvious (just as it is not obvious that all humans can have informed desires). Some utilitarians want to keep a pain/pleasure standard of goodness for animals, but endorse desire-satisfaction accounts for humans. Still another problem for the range component has to do with the emphasis in utilitarianism upon maximization of what is deemed to be intrinsically good. Something short of this, of increasing the amount of good to some extent less than the greatest extent possible, is both possible and less problematic, at least if one treats Derek Parfit’s “repugnant conclusion”—that one can increase the greatest total happiness (or desire-satisfaction) in the world by eliminating those at the bottom of the happiness ladder—as tied to the general thrust of maximization of the good.
To a large extent, discussion of utilitarianism in the late twentieth to early twenty-first century has moved away from earlier concerns. It has centered around three developments. First, R. M. Hare and others have urged a kind of indirect utilitarianism, wherein one does not employ consequentialism at the level of practice, in order to decide what it would be right to do. Hare urges a two-level account of moral thinking, which is rule-utilitarian at the level of practice but actutilitarian at the level of theory or rule/institutional design. In his hands, act-utilitarian thinking at the critical level will select as guides at the intuitive or practical level those rules whose general acceptance will give one the best chance of producing best consequences. In this way, it is only in exceptional circumstances that one’s practical guides are exposed to consequentialist thinking. Thus, Hare hopes to avoid many of the problems that are held to beset the act-utilitarian, through the employment of consequentialism as a way of deciding what it would be right to do. Hare’s two-level account of moral thinking can seem, then, a way of developing a further case for actutilitarianism; other two-level theorists, it should be noted, do not see the split-level innovation as furthering this particular case. In fact, on Hare’s account, since the focus of his theory is no longer acts, on a case-by-case basis, the name actutilitarianism is something of a misnomer.
Second, because so many of the objections to act-utilitarianism (and consequentialism) have always centered around clashes with “ordinary moral convictions” or “commonsense morality,” the question has arisen of whether our ordinary moral intuitions have probative force in ethics. To those who feel that they do have probative force, the problem has been to make it appear that certain of one’s intuitions are more secure than others, so that they are believed to be more “true” or “correct” than the dictates of any normative ethical theory. Accordingly, one needs to identify which these crucial intuitions are. Different ways of doing this have been suggested, including Rawls’s reflective equilibrium method. But it is clear, even in Rawls, that some intuitions survive intact, such as the wrongness of slavery. This intuition of his needs no revision. Other theorists privilege other of their intuitions about particular acts or classes of acts. Over privileging in this way, serious doubts can arise, no matter how secure one feels one’s intuitions to be. Yet, it is from this very privileging that condemnations of act-utilitarianism typically begin. It is wrong to lie, and the act-utilitarian mounts a case for lying on this occasion; it is wrong to kill, but the act-utilitarian mounts a case in favor of active euthanasia or abortion or suicide. In this way, arguments about the probative force of certain of one’s intuitions have taken on a life of their own and are often used in moral debate in the early twenty-first century. This is especially true in medical ethics, where some want to draw a distinction between killing and letting die and where act-utilitarians on the whole deny that such a distinction is morally significant or where some want to distinguish, say, between giving someone a pill which, if they take it, will kill them and giving someone an injection of a sufficiently large dose of morphine to hasten their death while intending only to relieve their pain.
Third, the enormous growth of applied or practical ethics, especially medical ethics, has brought act-utilitarianism and consequentialism into the public domain. Almost without exception, virtually every issue, whether to do with killing, the allocation of resources, or the case for animal experimentation, features a utilitarian line of argument, and this line of argument is also, typically, act-utilitarian in character. This in a way is odd: while theorists are developing more and different types of indirect utilitarianisms, most of the examples of applied ethics feature the direct application of consequentialism on a case-by-case basis, with, it is held, morally shocking results. To be sure, not everyone will be shocked any longer by a case in support of a doctor assisting a competent patient who voluntarily requests assistance in dying, but the case can be discussed from the level of rule or institutional design, as well as from the consequentialist realities of the situation in question. Moreover, there are more general questions that arise in the various areas of applied ethics, where act-utilitarians are taken to adopt a particular kind of stand. Thus, one can intentionally kill a patient or one can knowingly bring about or cause a patient’s death: are these morally different? One can directly bring about a person’s death or one can indirectly bring it about: are these morally different? One can bring about a person’s death by action or one can bring it about by omission: are these morally different? One can bring about a person’s death actively or passively: are these morally different? If the consequences appear the same or very similar, the actutilitarian will be held to think there is no difference among these things. Because Hare and others favor an indirect form of utilitarianism, it is not uncommon in the early 2000s to treat these practical issues as if they involved direct consequentialists and leave until another occasion whether these consequentialists are also act-utiliarians. In this way, applied ethics has helped consequentialism to come to dwarf the other components of act-utilitarianism.
For Hare, moral education plays an important role in one’s moral thinking. He rejects the employment of consequentialism on a case-by-case basis for deciding what to do and rejects as well the thought that consequentialism indicates the kind of thinking one should do at the intuitive or practical level. Instead, he emphasizes the role of character and character development: by education, individuals turn themselves into people whose actions flow from their character, in which the traits and dispositions that comprise character are inculcated in them overseen by the utilitarian goal of maximizing human welfare. This further reduces clashes with certain privileged moral intuitions, but it does not require that one treat those intuitions as having probative force. In this way, the moral thinking for Hare is much more intimately connected with one making one’s self into the sort of creature who behaves out of certain dispositions than into the sort of creature who acts only out of consequentialist concerns.
Utilitarianism, then, has in latter years undergone a significant transformation at the hands of theorists. It is no longer the relatively simple, straightforward rubric that Bentham and John Stuart Mill took it to be; its statement, even by those who remain sympathetic to it, such as Hare and those influenced by him, is complicated and layered. An account of its historical development in terms of ideas, as can be seen here, shows that it has become sufficiently encumbered with distinctions and technicalities that it no longer really resembles the earlier view. Yet, in the twenty-first century, it is still common to find thinkers objecting to some particular view in moral, political, or social policy as “utilitarian,” where the view they have in mind is the direct application of consequentialist thinking to decisions on how to behave. Thus, in modern medical ethics, there is little concern to get right the niceties of utilitarian theory; rather, the point is to protest about the “shocking” results that the direct application of consequentialist thinking can appear to produce in some instances.