We will not actually discuss Searle's paper in class. Rather, I will spend our class time giving students a Very Brief Introduction to Modal Logic. But you should nonetheless read Searle's paper and post about it here, as usual.
Searle aims to give an account of the senses of proper names. What he ends up with is a version of what has come to be called the "cluster of descriptions" theory. It is similar to Russell's view in that Searle regards names as associated with descriptive conditions. But his view is more subtle than Russell's in several ways.
Searle begins the paper by revisiting the puzzle about identity that Frege uses to motivate the claim that names have sense. He first considers the suggestion that even a statement like "Tully = Cicero" or "Mark Twain = Samuel Clemens" might be analytic and that the information conveyed by it might be about words, even if the statement itself is not about words. But this turns out to be a bit of a red herring, since this shows at most that such identity statements can be analytic and yet informative, not that all such identity statements are analytic. As Searle says, "...[P]eople who argue that Shakespeare was Bacon are not advancing a thesis about language" (p. 167).
Searle then gives a short argument that names must have sense as well as reference: In teaching someone the use of a name, we must establish for them a connection between the name and its bearer, and we can only do this if we can identify the bearer for them in terms of its "particular characteristics". But, as Searle's interlocutor points out, this is to assume that these characteristics are rules for the use of the name, as opposed to a one-off technique for establishing a connection with its bearer. On the other hand, however, Searle notes that, if a statement of existence or non-existence is to make any sense at all, it seems as if there need to be some descriptive conditions associated with the name.
Many of Searle's arguments, both in this section and later, turn on the fact that "Aristotle", say, may be the name of many different people. How might his opponent resist Searle's appeal to that fact?
But, Searle suggests, this sort of view, in assimilating names to descriptions, seems contrary to the apparently "subject-predicate structure of language", and it leads to the view that "variables of quantification [are] the only referential terms in the language" (p. 169). This is an allusion to an extreme form of the Description Theory developed by Quine, which eliminates names in favor of corresponding predicates (e.g., "Socrates" is replaced by "socratizes") and then treats names as descriptions (e.g., "the one who socratizes").1
In effect, then, Searle's problem becomes to explain why we have names in our language at all if every name is associated with one or more descriptions that determine its reference. Why not just use the descriptions instead? Searle's answer has two parts.
- First, Searle insists that, when one uses a proper name, one does not assert that the object in question satisfies the descriptive conditions that are associated with the name. Rather, "referring uses [of proper names] presuppose that the object to which they purport to refer has certain characteristics" (p. 171, my emphasis).
- Second, Searle thinks that there is no very definite set of such characteristics, and that what is presupposed is not that the referent has all of them but only that it has "a sufficient but so far unspecified" number of them.
Names are thus associated with "clusters" of descriptions, whose boundaries are vague and need not be determined in advance, but can be made more precise when and if the need arises.
Searle regards the vagueness just indicated as not a bug but a feature of his view. It is what allows him to explain why we have names in our language and not just descriptions. What is that explanation? And what does he mean when he says that names "function not as descriptions, but as pegs on which to hang descriptions" (p. 172)?
Searle notes that, if we were simply to replace "Aristotle" by "the teacher of Alexander", then this would have an odd consequence: One would normally think it a contingent fact that Aristotle ever taught anyone. But if we make this substitution, then "Aristotle taught someone" would be transformed into "The teacher of Alexander taught someone", and that looks analytic. Searle does allow, however, that "it is a necessary fact that Aristotle has the logical sum, inclusive disjunction, of properties commonly attributed to him" (p. 172). But does that answer the objection or concede it? Will a similar problem arise in that case?
In the last paragraph, Searle sketches an account of how a (true) identity sentence could be used to make a synthetic statement. What is that account? Why might one think that, on Searle's view, the content of the statement "Twain = Clemens" will actually be the same as that of "Twain = Twain"? And if that is right, what does that mean about what sorts of beliefs such a statement might be used to convey?
1 The whole truth about Quine's views on these matters is, as Delia Graff Fara notes, more complicated. See her "Socratizing", American Philosophical Quarterly 48 (2011), pp. 229-38 (JSTOR, Scanned PDF)