Reconsidering Kripke: Descriptivism's Enduring Whisper in Naming and Necessity — Epoche B2
What the Modal Argument Did Not Refute: Reference-Fixing After Naming and Necessity A death certificate issued too early The received account of Naming and Necessity is that Saul Kripke destroyed descriptivism about proper names [1] , and that the modal argument did the destroying. Both halves need qualifying. The modal argument refutes one thesis: that a description gives the meaning of a name. It leaves untouched the thesis that a description can fix which object a name picks out, and Kripke says as much himself [2] , at length, and relies on the distinction in his own positive account. The arguments that damage descriptivism as an account of reference are different arguments, made in the second lecture, and they hit one particular version of the view. The claim defended here is that the standard story misidentifies both the casualty and the weapon. What died was the synonymy thesis. What survived, and was reformulated in direct response, is a descriptivism about reference-fixing whose best statement is John Searle's in Intentionality (1983 [3] ) — a position the modal argument was never aimed at and does not touch. Someone can deny this by showing that reference-fixing descriptivism inherits the modal problem, and that is the shape a rebuttal would have to take. The view Kripke attacked Frege held that a proper name has a sense as well as a reference [4] , and that the sense is a mode of presentation which can be given by a description. Russell held that ordinary proper names are disguised definite descriptions [5] . By the 1950s the descriptivist family had absorbed an obvious difficulty — different speakers associate different descriptions with the same name — by loosening the requirement. In 'Proper Names' (1958) Searle proposed that a name is associated with a cluster of descriptions and refers to whatever satisfies a sufficient but unspecified number of them. No single description is necessary; the cluster does the work jointly. P. F. Strawson's 'On Referring' (1950 [6] ) is worth keeping in view here, because it makes a point that will matter at the end. Strawson's target was Russell's analysis, and his complaint was that referring is something a speaker does on an occasion, not a property of a sentence. Expressions do not refer; people refer by using them. That thought will return when we ask what determines which causal chain a name-use belongs to. The modal argument, and the distinction Kripke insists on Kripke's modal argument is short. Take the name 'Aristotle' and the description 'the teacher of Alexander'. It is true that Aristotle might not have taught Alexander; he might have died young, or taken up another trade. It is not true, in the same way, that the teacher of Alexander might not have taught Alexander. The name and the description therefore behave differently when we describe how things might have gone, so they cannot be synonyms. Names are rigid: they designate the same individual in every counterfactual situation in which that individual exists. The argument is decisive against the synonymy thesis. It is silent about a second use of descriptions, which Kripke separates from meaning in the first lecture and returns to repeatedly. A description may be used to fix the reference of a name without giving its meaning. His own examples make the point: the metre was fixed by reference to the length of a particular bar, and 'Hesperus' by reference to a body seen in the evening sky, and in neither case does the name mean what the fixing description means. Once the distinction is drawn, the modal argument's reach is exactly as wide as the synonymy thesis and no wider. It is also worth recording what Kripke says about tracking individuals across possibilities, since the worry recurs. He denies that a counterfactual situation is a distant place to be inspected, and denies that we need a criterion for recognising Aristotle when we get there. Possible situations are stipulated. We say "suppose Aristotle had not taught Alexander" and thereby specify who we are talking about. An example from Japanese naming practice The distinction between fixing reference and giving meaning can look like a technicality. A real naming practice makes it concrete. Japanese emperors are designated after death by the name of their era: the emperor whose personal name was Mutsuhito, and who reigned from 1867 to 1912, is known as Meiji. The designation was introduced by a description — the emperor of the Meiji era — and could not have been introduced any other way, since it was not available during his lifetime. Now run the modal test. "Meiji might have died in infancy and never reigned" is plainly true. "The emperor of the Meiji era might never have reigned" is not true in the same way. The name and the description that introduced it come apart under counterfactual supposition, exactly as Kripke says. And yet the description did the introducing. A practice that names people by describing them produces rigid designators, which is precisely what the reference-fixing distinction predicts and what the synonymy thesis cannot accommodate. What actually damaged the cluster theory The arguments that bear on reference rather than meaning come in Lecture II, and they are of two kinds. The first turns on ignorance. A speaker may associate with 'Feynman' nothing more than "a famous physicist", which does not pick out one man, and still refer to Feynman. Descriptions the speaker can supply are often too thin to determine anything. The second turns on error, and is the sharper of the two. Suppose everything the ordinary speaker believes about Gödel — that he proved the incompleteness of arithmetic — were true instead of a man called Schmidt, whose work Gödel appropriated. On the cluster theory, that speaker's use of 'Gödel' would refer to Schmidt. It does not. It refers to Gödel, the man at the end of the chain of use, however badly the speaker has been misinformed. This is a direct counterexample to the cluster t