A reconstruction of Mill’s argument for utilitarianism (David Brink)

David Brink says that there are various interpretations of what Mill’s argument for utilitarianism is.

Brink then presents the following as a reconstruction of the argument on a traditional interpretation:

1. Utilitarianism is true iff happiness is the one and only thing desirable for its own sake (and not for the sake of something else).
2. The only proof of desirability is desire.
3. Each person desires his own happiness for its own sake (and not for the sake of something else).
4. Hence, happiness, as such, is desired for its own sake (and not for the sake of something else) from the point of view of humanity (= the aggregate of persons).
5. Hence, happiness, as such, is desirable for its own sake (and not for the sake of something else).
6. Happiness is the only thing desired for its own sake (and not for the sake of something else). Other things—such as virtue, health, music, money, and power—can come to be desired for their own sakes, but then they are desired as parts of happiness.
7. Hence, happiness is the only thing desirable for its own sake (and not for the sake of something else).
8. Hence, utilitarianism is true.

Brink also presents eight concerns about Mill’s argument. These are his words:

1. (1) is plausible only if “desirable” means worthy of being desired, not if it means capable of being desired. But (2) is most plausible if “desirable” means capable of being desired (see (iii) below). But then there is a real worry that the argument trades on a tacit equivocation between these two different senses of “desirable” and that the argument is, as a result, invalid.
2. Even so, (1) is false. Even if happiness were the one and only thing desirable for its own sake, this would establish only a claim about the good or “ends.” It is not a claim about duty or right action. Utilitarianism not only claims that the good is human happiness but goes on to define the right in terms of promoting the good. The second claim does not follow from the first.
3. For the argument to be valid, “desirability” in premise (2) must mean worthy of being desired (as it does in premise (1)). But then (2) is false. Desire is not proof of desirability. People can and do have mistaken desires about what is good. Indeed, if Mill is either a hedonist or a perfectionist he must think that people can and do have desires that fail to track the good.
4. It is not clear that (3) is true. It seems as if masochists or selfless altruists might fail to desire their own happiness for its own sake.
5. (4) may be incoherent and certainly does not follow from (3). It is not clear that aggregates of persons have desires. Perhaps under special circumstances groups of people might form a corporate agent or person. But aggregates of persons, as such, are not persons and do not have desires. Even if they did, it is doubtful that one could infer what the aggregate desires from facts about what its members desire. That would involve a compositional fallacy.
6. (5) is presumably equivalent to the claim that happiness is good. But is it good simpliciter or good for the aggregate? The analogy between individuals and groups would suggest that happiness should be a good for the aggregate. But presumably the intended conclusion requires that happiness be good simpliciter.
7. It is not clear how to understand (6). One would think that the aim is to make claims that parallel (4) and (5). But then (6) needs to be understood as making another claim about aggregate psychology. And this raises some of the earlier questions about aggregate psychology. However, much of the discussion in IV 5–8 seems to be about individual psychology. Mill seems to be saying that insofar as individuals do have intrinsic desires for things other than their own happiness the objects of intrinsic desire are desired as parts of their own happiness. Perhaps this is Mill’s initial claim from which he then hopes to infer, as he did from (3)–(4), that the general happiness is the only thing desired by the aggregate for its own sake (and not for the sake of something else). This inference would, of course, give rise to the same sort of worries we raised about the inference from (3)–(4). In particular, we might doubt that aggregates of persons have any aims, much less ultimate aims. And even if we conceded that they did, it is not clear that we could infer facts about the desires of aggregates from facts about the desires of its members. That, we said, would seem to involve a compositional fallacy.
8. Even if we accepted this defense of (5) and (7), this would only establish that happiness as such was the only thing desirable or good for the aggregate. It looks like we could have parallel claims about the agent’s own happiness being the only thing desirable or good for the individual. But this might seem to imply that while the aggregate should pursue or promote the general happiness individuals should pursue or promote their own happiness. That would not be a defense of utilitarianism.

Reference

Brink, D. 2018. Mill's Moral and Political Philosophy. Forthcoming in E. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Fall 2018 Edition).

ß