Context and Kaplan's Demonstratives — Epoche C1
The Enduring Shadow of Descriptivism in Referencing The prevailing view in the philosophy of language holds that Saul Kripke's arguments in Naming and Necessity — three lectures delivered in 1970 and published as a book in 1980 — dealt a decisive blow to descriptivism as a theory of how proper names refer. On this telling, names are rigid designators whose reference is fixed by an initial baptism and a causal chain of use, not by any cluster of descriptions, and descriptivism belongs to the annals of philosophical history. A closer examination, particularly through the lens of David Kaplan's work on demonstratives, suggests that the relationship between direct reference — the doctrine that some terms reach their object without going through any identifying condition at all — and descriptivism is more intricate and less adversarial than this. But the claim can only be assessed if the two combatants, and the exact points at which Kripke's arguments strike, are first set out plainly. What descriptivism claimed Descriptivism is the theory that a proper name refers to whatever object satisfies the descriptions speakers associate with it. Its classic sources are Gottlob Frege and Bertrand Russell. Frege (1892) distinguished an expression's reference — the object it stands for — from its sense — the way that object is presented. The distinction was forced on him by a puzzle: 'Hesperus is Hesperus' is trivial, while 'Hesperus is Phosphorus' was an astronomical discovery (both names denote the planet Venus, one as evening star, one as morning star). If the meaning of a name were just its bearer, the two sentences should be equally trivial; so, Frege concluded, a name must also carry a sense, something like a descriptive condition — 'the brightest body in the evening sky' — through which the bearer is picked out. Russell (1905) went further: an ordinary proper name is a disguised definite description, so that 'Aristotle' abbreviates something like 'the teacher of Alexander who wrote the Metaphysics '. A refinement due to John Searle (1958) replaced the single description with a loose cluster: the name refers to whatever satisfies enough of the associated descriptions. In all versions, the mechanism of reference is satisfaction : the name reaches its object because the object uniquely fits an identifying condition carried by the name. What Kripke actually refuted Kripke's attack consists of three distinguishable arguments, and it matters for what follows which target each one hits. The modal argument. A rigid designator is a term that designates the same object in every possible world in which that object exists — 'possible world' here being simply a complete way things could have gone. Names are rigid: 'Aristotle might not have gone into philosophy' is true, and it is a claim about that very man in a counterfactual situation. Descriptions are typically non-rigid: 'the teacher of Alexander' picks out different people in different possible circumstances, or no one. If 'Aristotle' simply meant 'the teacher of Alexander', then 'Aristotle might not have taught Alexander' would amount to 'the teacher of Alexander might not have taught Alexander' — which, read as a claim about whoever fits the description, is false — and yet the original sentence is plainly true. Since the name and the description behave differently under 'might', they cannot be synonymous. The epistemic argument. If the name meant the description, 'Aristotle taught Alexander' could be known just by reflecting on meanings, without evidence — known a priori . It cannot; it is a contingent historical claim we might discover to be false. The semantic argument. Suppose the one description most speakers attach to 'Gödel' is 'the man who proved the incompleteness of arithmetic', and suppose (fictionally) the proof was really the work of an unknown man named Schmidt. Descriptivism predicts that when such speakers say 'Gödel', they refer to Schmidt — the satisfier of the description. Intuitively they still refer to Gödel. So satisfaction of associated descriptions is not, in general, what determines reference. Kripke's alternative picture: a name is attached to its bearer in an initial baptism, and each later user refers to whatever the person they acquired the name from referred to — a historical chain of transmission, not a fitting of descriptions. One of Kripke's own distinctions is indispensable here, and it is where the earlier version of this essay went wrong. Kripke distinguishes a description that gives the meaning of a name from one that merely fixes its reference . Le Verrier introduced 'Neptune' by stipulation as a name for the planet causing the observed perturbations in Uranus's orbit — a description did the introducing. Yet 'Neptune', once introduced, is rigid: in a counterfactual world where some other body caused those perturbations, 'Neptune' still names Neptune. The earlier version of this essay stated that Kripke's arguments 'primarily targeted the idea of descriptions fixing reference in a way that made names non-rigid'. That formulation conflates the targets, and the Neptune case shows why it cannot be right: reference-fixing by description does not make a name non-rigid at all. What the modal and epistemic arguments refute is the claim that descriptions give the meaning of names; what the semantic argument refutes is the claim that description-satisfaction is the general mechanism of reference. What survives, by Kripke's own account, is the possibility that descriptive information fixes reference in particular acts of introduction and identification — while contributing nothing to what the name thereafter means. It is exactly this surviving role that Kaplan's framework systematises. Kaplan's two tiers: character and content Kaplan's monograph 'Demonstratives' — circulated from 1977 and published in 1989 — concerns the expressions whose reference shifts with the occasion of use: pure indexicals such as 'I', 'here', 'now' and