Section 3
Is the Universalizability True on Logical Grounds?
Let us recall that Hare is trying to show the universalizability as a logical thesis: it is meant to be true on conceptual grounds alone, or true by virtue of the meaning of 'ought' (or any other evaluative word). Thus, if Hare is right, the universalizability is analytically true, given the meaning of 'ought'; or it can be established by the conceptual truth contained in the notion of 'criterion' (sect.2, (1)) or 'reason' (sect.2, (2)) or 'meaning-rule for a descriptive word' (sect.2, (3)).
However, we also have to recall that Sidgwick insisted that all of his three principles are non-tautological (see Okuno 1998b, 6.2). One of the easiest interpretations of his Principle of Justice (sect. 1, (1)), for instance, is that it merely expresses the universalizability of 'ought' or 'right action' (I myself endorsed this interpretation for sometime). But if Hare is right, this interpretation makes the Principle of Justice a tautologous or analytically true principle, and Sidgwick persistently tried to avoid such a principle for his 'self-evident' principles for ethics. Then it is obvious that both cannot be right; so which is wrong, Sidgwick or Hare?
Actually, I once argued (a long time ago, Uchii 1974) that Hare is wrong; and although I have not changed my mind, I now see the reason more clearly why he is wrong, because I have now realized the important differences between Sidgwick's three principles. Let us first concentrate on the universalizability of 'ought'.
No one will deny that an 'ought'-judgment has a criterion. No one will deny that an 'ought'-judgment must be made on some reason. And, again, no one will deny that an 'ought'-judgment has a descriptive meaning. However, it is not so obvious that the criterion must be universal, in the sense that it does not contain any reference to an individual. Likewise, it is not so obvious that the reason must be universal in the same sense. And, finally, it is not so obvious that the descriptive meaning must be explicable without any reference to individuals.
Since Hare seems to be emphasizing the universalizability as the common property between descriptive judgments and value-judgments (in Hare 1963), let us consider the descriptive meaning of a descriptive judgment, such as
(1) This is one meter long.
I presume no one will question that this is a descriptive judgment. Indeed, Hare (1955) mentioned 'one meter long' as an example of a universal expression (Hare 1955, 306). But how can we explain the descriptive meaning of 'one meter long'? You know that the unit of the length 'meter' was determined historically, and there is the standard of 'meter' in Paris. Thus, 'this is one meter long' means that the length of this is identical or at least approximately identical with that standard; more specifically, it would mean that 'if this is transported to Paris and compared with the standard, the two will coincide'. As Hare rightly points out, any descriptive word depends on the similarity or the identity of this sort, and we can say that
(2) anything similar to the Paris standard in the relevant respect (i.e. the length) is 'one meter long'.
This statement certainly has a universal form, and it defines an open class. But the syntactical form of universality is only a necessary condition for the universalizability; and likewise an open class may not correspond to the extension of a universal word, since you can define an open class by referring to an individual, as is the case with 'a citizen of the United States'.
Notice that, with the last example, we can construct a similar statement to (2) as follows:
(3) anyone similar to John F. Kennedy in the relevant respect (i.e. nationality) is 'a citizen of the United States'.
This statement also has a universal form, but no one will deny that it (implicitly) contains an essential reference to an individual (the United States). Thus it is clear a statement with a syntactically universal form may not be properly or semantically universal.
And (2) is such an instance as containing a reference to an individual. I know Hare claimed that the expression 'similar to X' can be replaced with a universal word (Hare 1955, 306-7; 1963, 11) even if X is a singular expression; but his claim is without a proof, and refuted by our example (3). May I also point out that the Paris standard is a unique individual? Now, can we eliminate from (2) the reference to the Paris standard? You might think that the reference is inessential because you can substitute a reference to another standard (many countries have their own copies of the standard); but you must recall that such substitute standards work as a standard precisely because of their connection with the Paris standard; and (2) is an instance of what Reichenbach called a 'coordinative definition', a definition correlating a concept to a particular object (Reichenbach 1958, 14). This is a convention indispensable for making 'meter' a universal word; thus a meaning-rule for a descriptive word is not as simple as Hare supposed.
I know that more recent methods of determining a meter are more complicated and refer to the wave length of a spectrum of a certain atom or of light; still, unless these complicated methods retain their reference to the original standard, the meaning of 'meter' will change. But we do not have to get into messy details. The point here is that the meaning-rule which determins the meaning of 'meter' did historically contained a reference to an individual, and nothing was wrong, logically, with this meaning-rule; statement like (2) is not semantically universal, because it contains an essential reference to an individual, but it works perfectly well as a rule for determinng the meaning of a descriptive word.
Physical knowledge is characterized by the fact that concepts are not only defined by other concepts, but are also coordinated to real objects This coordination cannot be replaced by an explanation of meanings, it simply states that this concept is coordinated to this particular thing. . . . these first coordinations are therefore definitions which we shall call coordinative definitions.
To 2. Hare's Analysis of the UniversalizabilityTo 4. The Weak and the Strong Universalizability
[back to Sidgwick&Hare index][back to Uchii index]
July 20, 1998. Last modified April 17, 2006. (c) Soshichi Uchii
webmaster
