The Integrity Critique of Utilitarianism

・6 min read

Bernard Williams’s hugely influential (and highly enjoyable) “A critique of utilitarianism”, originally published in 1973, has been called “a tour de force of philosophical demolition” (Chappell 2013). In it, Williams forcefully argues that utilitarianism can’t make sense of some of our basic moral concepts, especially of personal integrity, and hence shouldn’t be regarded as a guide for moral action.

At the core of Williams’s critique, which is part of his larger endeavour of criticizing and deconstructing systematic theories of morality, is the so-called integrity argument: an argument for the thesis that “utilitarianism […] makes integrity as a value more or less unintelligible [because] it cannot coherently describe the relations between a man’s projects and his actions” (Williams 1973, 99 f.). Williams develops and illustrates the argument using two thought experiments: One about pacifist George who would have to act against his convictions and take a job in the WMD business were he to follow utilitarian principles, and one about reporter Jim who is pushed by Pedro to kill an innocent person in order to save more innocent people from being killed by Pedro himself. Especially in the case of George, Williams argues, utilitarianism ignores the fact that special projects and commitments are at the core of our moral self and thus leads to unacceptable consequences—it prescribes actions as obligatory that undermine our very nature as moral agents.

In the course of the essay, Williams presents at least three distinct versions of his thesis:

  1. Utilitarianism ignores the idea “that each of us is specially responsible for what he does, rather than for what other people do.” (Williams 1973, 99, my emphasis)
  2. “Because our moral relation to the world is partly given by [moral] feelings, […] to come to regard those feelings from a purely utilitarian point of view, that is to say, as happenings outside one’s moral self, is to lose a sense of one’s moral identity”. (Williams 1973, 103 f., my emphasis)
  3. To “demand from […] a man, when the sums come in from the utility network which the projects of others have in part determined, that he should just step aside from his own project and decision and acknowledge the decision which utilitarian calculation requires” is “to alienate him in a real sense from his actions and the source of his action in his own conviction.” (Williams 1973, 116, my emphasis)

The third version is clearly the final one, and supported by Williams’s most sustained reasoning. But is it really convincing?

As other authors have noted (cf. Cox et al. 2017), it seems to assume that even though an agent follows utilitarian principles, these principles are not part of his convictions in the same way as e. g., in George’s case, pacifism. But if George really was a utilitarian agent, his utilitarian principles would clearly be convictions at least as important as other, less universal moral beliefs. Williams seems to confuse non-utilitarian George, whose projects would be undermined by utilitarianism, with utilitarian George, who will at least modify his other projects in the light of his more universal convictions, if he is a rational agent at all. So for a real utilitarian, his convictions really are the source of his actions, it seems.

What about the other versions, then?

If the above objection is correct, then the utilitarian’s moral feelings, too, are his own in a very real sense, are part of his moral identity. If that is the case, then Williams‘s critique collapses into the standard complaint that utilitarianism is too rational, too “cold” in its treatment of feelings as a mere input into its utility calculations.

In the first version of the thesis, Williams’ main point is that utilitarianism doesn’t discriminate between an agent’s actions that lead to a certain state of affairs (positive responsibility) and his allowing things to happen due to actions of others (negative responsibility), although these cases do not seem to be symmetrical. Assuming that decisions (not) to act are actions themselves, the following seems plausible: If the agent is a utilitarian, i. e. evaluates his decisions only according to their consequences, then this critique is only striking if his inaction does not really cause the consequences in question, that is, if there is a causal gap between his decision and these consequences. But Williams himself seems to grant that there is no such gap and that the only difference between the cases is that between “one state of affairs brought about by me, without intervention of other agents, and another being brought about through the intervention of other agents” (Williams 1973, 94, my emphasis). So for a utilitarian agent, there doesn’t seem to be a point about Williams’s first version, either.

Thus, in all three versions the critique loses its power once we assume a truly utilitarian agent—for him, there is no incoherence between his actions, commitments, feelings, and responsibilities. Williams’s argument seems to be valid only from an already decidedly non-utilitarian point of view.

On the other hand, there is a lot intuitive appeal in Williams’s examples and behind his explicit arguments. For example, even if we accept that by not shooting an innocent person Jim is morally responsible for more innocent people to die, we still feel that Pedro, who created Jim’s situation in the first place, is somehow more responsible for that outcome. Or to use a more contemporary illustration: In the 7th season of the successful TV show The Walking Dead, villain Negan repeatedly creates situations in which series hero Rick has to decide between different courses of action that will cost more or less people’s lives. We feel that Rick, via his decisions, is morally responsible for these outcomes—but at the same time, we feel that Negan is more responsible for them. We’re repelled by the fact that Rick has to make these decisions in the first place, and we naturally blame Negan for that. The mere fact that we as spectators are forced to morally evaluate these situations makes us at least uncomfortable and sometimes even disgusts us.

So maybe behind Williams’s argument that utilitarianism undermines integrity, there’s a strong intuition that moral responsibility cannot be absorbed in utility calculations; that creating morally charged situations means carrying more responsibility than acting in them; that being forced to decide is an assault on moral autonomy. This emphasis on responsibility and autonomy is certainly at odds with utilitarianism, and it fits even more certainly into Williams’s general uneasiness with systems of morality that abstract away from concrete moral agents, morally ambiguous situations and the complexity hidden in both.