The Ethics of Possible Worlds

Moral philosophers often distinguish between normative ethics and metaethics. While the former aims to identify the criterion of right action, the latter studies the conceptual and ontological problems raised by moral practice. Some go so far as to claim that the two domains are perfectly independent. In a sense, this split is self-evident: ethics enjoys a certain autonomy; its propositions cannot be derived from non-moral propositions. On the metaethical proposition The ethical reasons are categoricalfor example, one cannot deduce the moral proposition Onanism is morally wrong.

Others, however, believe that the study of moral concepts imposes serious constraints on normative ethics. The British philosopher Richard Hare (1919-2002) illustrates this approach well. In Thinking morally (1981), a now classic work in English that has just been translated, he argues that as such, philosophers should confine themselves to examining the concepts of ethics, but that it is possible to draw substantial moral lessons from the latter. In particular, Hare claims to demonstrate that a form of utilitarianism arises from the logic of moral concepts alone. It is on this aspect of the work that I will focus for the purposes of this review.

Moral thinking

Hare’s metaethics combines two theses about moral language: the prescriptivism and theuniversalizability. According to prescriptivism, moral statements serve not to represent any moral reality but to prescribe behaviors. In short, they imply imperatives (p. 105). The statement “You must wear a mask,” for example, implies the imperative “Wear a mask!” Similarly, the statement “Blasphemy is immoral” implies the imperative “Let us not blaspheme!” Hare bases this thesis on a linguistic intuition: the statements “You must wear a mask, but don’t!” and “Blasphemy is immoral, but let us blaspheme!” seem contradictory to him.

According to universalizability, moral statements imply identical statements for all cases that are identical in their universal properties (pp. 226–227). In the sense that interests us, a property is universal when it can be attributed via a predicate that does not contain a rigid designator, which would refer to an individual in any way other than description. The property of having small hands is therefore universal, unlike that of being Melania Trump’s husband. Suppose that you instantiate the universal properties F, G, …, Z. The statement “You must wear a mask” will then imply the statement “Anyone who instantiates the properties F, G,…, Z should wear a mask.” Like prescriptivism, universalizability relies on a linguistic intuition: the statement “You should wear a mask, but I shouldn’t if I shared your universal properties” seems contradictory.

To these two theses on moral language, Hare adds two conditions on the subject of prescriptive thought. First, a constraint of sincerity: one can sincerely state an imperative only if one prefers that it be followed by effect, because “to have a preference is to accept a prescription” (p. 199). Thus, I must desire that my mother pass me the salt in order to be able to sincerely ask her “Pass me the salt, Mom!”

Second, the “conditional reflection principle”: necessarily, if I imagine a possible world in which I desire that P and I know what it would do to me then to desire that P, I desire, in the actual world, that P be the case in this other possible world (p. 206). In all transparency, I do not like Rafael Nadal. There is nevertheless a distant possible world in which I desire that he win Roland Garros. The conditional reflection principle then implies that, if I imagine this possible world and manage to know what it does to me there to want Rafael Nadal to win Roland Garros, then I now desire that he win Roland Garros in this possible world.

A proof of utilitarianism

Based on these four theses, Hare defends a theory close to preference utilitarianism, according to which an act is right if and only if it maximizes the satisfaction of desires (or preferences – the distinction is unimportant here). Here is the essence of his argument. Suppose that:

Let us also suppose that Donald instantiates the universal properties F, G, …, Z. By virtue of universalizability, it follows that:

Under prescriptivism:

And by virtue of the constraint of sincerity:

So far, so good. But there is one more step – the most complex one, to be honest.

To make things easier, let’s assume that there are only two people involved: Donald and me. Donald instantiates the properties F, G, …, Z in the actual world; I instantiate them in another possible world. According to proposition (4), I can therefore desire that he and I wear a mask, each in the world where he has the properties in question. Let us see what this implies.

I already desire that Donald wear a mask in the actual world. It is therefore appropriate that I form a desire for the other possible world: either that I wear a mask there, or that I do not wear a mask there. Now, in this other possible world, I instantiate the properties F, G, …, Zamong which is the desire not to wear a mask. So it’s about forming a desire for a possible world where I don’t want to wear a mask and, in order for that desire to be informed, I form it knowing what it would be like to not want to wear a mask.

Then comes the principle of conditional reflection. Suppose that I represent this other possible world faithfully enough to know what it is like for me to not want to wear a mask. Then I necessarily want not to wear a mask there. I now have two desires: a desire for the actual world (that Donald wears a mask there) and a desire for the other possible world (that I do not wear a mask there). And my preference that anyone instantiate the properties F, G, …, Z wearing a mask is nothing other than the sum of these two desires.

Consequently, I can desire that anyone possess the properties F, G, …, Z wears a mask only if my desire for the actual world (that Donald wears a mask in it) is stronger than my desire for the other possible world (that I do not wear a mask in it). Since my desire for the other possible world is identical to Donald’s desire, this will be the case only if my desire for the actual world is stronger than Donald’s. In other words, only if:

In sum, I can honestly claim that Donald should wear a mask only if doing so would maximize preference satisfaction. More generally, I can honestly claim that an act is right only if it maximizes preference satisfaction.

Objections

This defense of utilitarianism has the merit of elegance, but it poses several problems. I will mention two of them, which concern respectively prescriptivism and the principle of conditional reflection.

Hare’s argument for prescriptivism is rather implausible. Reading the famous article “Famine, Affluence, and Morality,” you might be convinced by Peter Singer’s thesis but think, “Sure, we should give 30% of our income to charity, but let’s keep 80%! We’ll eat more caviar.” You might be wrong, but your failure would be ethical rather than logical. For, while it would betray your moral weakness, this statement nonetheless seems coherent. According to our linguistic intuitions, moral statements do not imply imperatives; they are not prescriptive.

This is also what the famous “Frege-Geach problem” suggests. Consider this modus operandi :

The argument is clearly valid. That is, the meaning of the phrase “blasphemy is immoral” is constant from one premise to the other. Since prescriptivists claim that it is prescriptive in the first premise, they must therefore say the same of the second. But clearly this implication is absurd: one can clearly subscribe to proposition (7) without prescribing anything.

The principle of conditional reflection is no more plausible. It raises two difficulties. First, it is open to counterexamples. So I can well imagine what it feels like to vote RN. However, because I consider it immoral to give one’s vote to the extreme right, I do not wish to vote. RN even for the counterfactual case where I would like to do so.

Second, the principle assumes that a certain belief guarantees the presence of a certain desire: if you know what it is like to have a certain preference, you will necessarily desire that preference to be satisfied if you had it. But this assumption contradicts the idea, to which Hare subscribes elsewhere, of a gulf between beliefs and desires.

Hare is aware of the second problem. Unfortunately, his solution to it leaves something to be desired. It consists in saying that the pronoun “I” and, therefore, the statements that contain it are prescriptive, so that knowing what it would do to me if I wanted to vote RN is not a belief but a desire (pp. 207-208, 402). It is not surprising then that this state engenders a preference. Do not hesitate to write to me if you see what imperative arises from the statement “Yesterday, I ate an apple”.

A sometimes approximate translation

The objections to Hare’s theory are sufficient to explain its relative neglect in contemporary debates. Its historical importance, however, amply justifies this French translation of Moral Thinking. Overall, the mission is accomplished. This must be emphasized, because it was not made easier by Hare’s immense attention to language – and therefore to the English language in particular.

Inevitably, however, the translation includes some inaccuracies, some of which could hamper the understanding of an already difficult theory. In particular, three implications of the principle of conditional reflection are incomprehensible once translated. These errors are all the more regrettable since this principle, which plays a crucial role in Hare’s argument, is often misinterpreted.

In addition, the translators’ copious notes, which will provide welcome insight to French-speaking readers, also contain some surprising blunders. As when act utilitarianism is opposed to preference utilitarianism, then, a few lines later, confused with a form of egoism (p. 122). This is a lot for a note that is supposed to respond to “philosophers who criticize utilitarianism without knowing it.”

Finally, a mystery remains: why translate the original title, Moral Thinkingby the pompous Thinking morally ? The most literal translation Moral thinking Wouldn’t it have done the trick? But let’s put things into perspective. In a possible world that is perhaps close, we find on the shelves of French bookstores the work Thinking about morality.