CRITICISM Bernard Williams on Utilitarianism

Williams on Utilitarianism - a discussion by Martha Nusbaum

Boston Review Oct/Nov 2003

Utilitarianism, with its tremendous influence on public life, came in for a variety of different criticisms at different times in Williams's career. His most influential critique concerned its impoverished notion of the person and of human agency. According to the utilitarian the right choice is the one that maximizes total or average welfare. Because of this exclusive focus on the results of our conduct, the theory forces us to consider the world from a perspective outside our own lives. We are not permitted to ascribe particular importance to a consideration simply because it plays a salient part in our own life, or to an action simply on the ground that it is we who do it.

In a famous and much-debated example, Williams imagines a scientist, Jim, doing research somewhere in Latin America. Suddenly confronted by the henchmen of a brutal dictator, he is shown a group of captive rebel Indians and told that all 20 will shortly be executed-unless Jim, as an honored guest, agrees to kill one of them himself. In that case the other 19 will be let go. For utilitarianism there is no dilemma here, because the only relevant consideration is the outcome. For Jim, Williams argues, there ought to be a serious dilemma (though in the end he thought Jim should probably accept the offer), because the fact of his own agency in wrongdoing-the fact that he will kill someone, and not simply that someone will die-makes an ethically relevant difference. In general utilitarianism makes the person only an engine to produce outcomes, but this is an impoverished conception of the person and cannot suffice for a plausible theory of human conduct. Williams made it clear that his argument was intended as an objection not only to utilitarianism but to all forms of consequentialism-all views, that is, that hold that an action is right just in case it promotes the best consequences.

A later discussion coauthored with Amartya Sen, the Introduction to their edited collection Utilitarianism and Beyond, usefully dissects utilitarianism into three parts: consequentialism; sum-ranking (a theory about how different people's welfares are aggregated, namely that they are simply added together); and welfarism (the view that the value of consequences is a function solely of human welfare). All three parts are independent, all contain serious problems, and all the troubles converge on a "narrow view of the person" as merely a container of satisfactions. No doubt because Sen did not accept Williams's critique of consequentialism (he has argued that agency and the personal point of view can be worked into a complicated account of which consequences are in fact best), Williams did not repeat his earlier arguments. Instead, the coauthored piece focuses on difficulties with sum-ranking and utilitarian accounts of welfare.

Sum-ranking fails to give sufficient salience to the worth of each person: it in effect permits one person's large misery to be overbalanced by small benefits to a large number of others. If A has a miserable life full of pain and indignity, but B and C and D have extremely happy lives, then the sum of happiness may be greater than if A's life were improved at the cost of small losses to B, C, and D. But politics should not only consider the total, treating a person as nothing but a function in a larger social system; it should also consider the claim that each individual has to be treated with decency and have his or her rights respected. In making this criticism, Sen and Williams were in strong agreement with John Rawls's Kantian account of justice. These views about the importance of equal respect for persons were central to Williams's engagement with politics; he was a lifelong social democrat and egalitarian. (His article "The Idea of Equality" remains one of the most important discussions of that political value.)

Sen and Williams also criticize the major theories of welfare in utilitarian theories: pleasure and satisfaction of preferences. These accounts assume that the good is single when in fact it is heterogeneous; they rely on preferences, which are notoriously malleable, adjusting to unjust background conditions, rather than adopting some more robust account of the good.