Prof. Uchii's Page

Section 2

Hare's Analysis of the Universalizability

On the other hand, it is also well known that Richard M. Hare has propounded another version of utilitarianism based on the logical properties of moral words and the requirement of rationality. The logical properties of moral words he mainly appeals to are: the universalizability and the prescriptivity of a value-judgement. Since the main point of this paper is concerned with the affinity and the difference between Sidgwick's three principles and the applications of Hare's universalizability, I shall ignore the prescriptivity, and concentrate only on the universalizability. Now, what is the universalizability of a value-judgment?

Hare gives, as we shall shortly see, several kinds of explanation, but the essential content of the universalizability seems to be clear. He has been maintaining "that the meaning of the word 'ought' and other moral words is such that a person who uses them commits himself thereby to a universal rule" (Hare 1963, 30). "Universal" means that it does not contain any reference to an individual, such as a particular person, a particular time, or a particular spatial location. Thus, according to Hare, an 'ought'-judgment like "He ought not to smoke in this compartment"---- although it is a singular judgment referring to an individual person 'he' (whoever it is)----, depends on, or implies, another 'ought'-judgment which does not contain any reference to an individual, and hence can be expressed only in terms of universal quantifiers and universal words.

But why does the word 'ought' or any other moral (evaluative) word have this property of the universalizability? Hare gives the following reasons (roughly in a chronological order, as Hare's view develops).

(1) First, an 'ought'-judgment (and a value-judgement in general) must be made on a criterion; and this implies that if the same criterion is satisfied, the same judgment must be made. Thus as long as a value-judgment is made on a criterion, it is implicitly universal (Hare 1952, ch.6).

(2) Second, an 'ought'-judgment (and a value-judgement in general) must be supported by a reason; and this implies that the same judgment must be made whenever the same reason holds. Thus there must be a universal 'ought'-judgment behind the individual 'ought'-judgment (Hare 1952, 176; Hare 1963, 21, etc.).

(3) Third, an 'ought'-judgment (and a value-judgement in general) has a descriptive meaning; and a descriptive meaning presupposes a universal rule which determines it. Although evaluative words and descriptive words differ in their essential function, they do share this feature as long as they have a descriptive meaning as an element of their meaning (Hare 1963, ch.2). And since the meaning-rule which governs the use of a descriptive term is a universal rule (dependent on the similaritiy of objects in a certain respect ), a singular descriptive judgment is universalizable, and in the same way, a singular 'ought'-judgment (and an evaluative judgment in general) is universalizable (Hare 1963, 13).

These are the major lines of arguments in favor of the universalizability. But our question is: are they right, or is any of the three reasons good enough to establish the universalizability?


? 1. Sidgwick's Three Principles

? 3. Is the Universalizability True on Logical Grounds?

[back to Sidgwick&Hare index][back to Uchii index]



July 20, 1998. (c) Soshichi Uchii

suchii@bun.kyoto-u.ac.jp