Roads to Reference

The talks in this series are parts of a potential book on the metasemantics of reference. In them I explore questions concerning the issue of how the referents of a variety of expressions are fixed. The ways reference is fixed are varied, both across kinds of expressions and across examples of one same kind of expression; in some cases some kind of descriptivism is reasonable, while in others some kind of anti-descriptivism holds, for example. However, the talks (and the projected book) will seek to draw some (more or less modest) patterns of unification. For example: descriptivism is only reasonable in some localized cases where the relevant description is relatively simple, and in other cases some kind of anti-descriptivist mechanism is at work; reference to abstract kinds is feasible via ordinary mechanisms of reference fixing, despite popular eliminativist considerations to the contrary coming from the philosophies of special scientific disciplines (mathematics, chemistry, biology...); etc.

Talk 1: Proper Names

Here I will first consider the classical question whether a speaker must know, for an arbitrary name he is familiar with, a reference-fixing description that is “co-referential” with that name as he uses it. I will argue that cases where it is intuitively unclear whether a certain name has a reference or not suggest that extant reference-fixing conventions for names are conventions giving merely sufficient conditions for reference and reference failure in selected types of occasions, leaving many cases referentially indeterminate. And I will argue from this to a negative answer to the classical question. I will also develop a little bit the picture of name reference-fixing conventions as conventions giving merely sufficient conditions for reference and reference failure, and argue that it does a good job of dealing with some test cases.

Talk 2: Demonstratives

Descriptivism has seemed a more natural view of reference fixing in the case of demonstratives than in the case of names. However, I will argue that considerations similar to the considerations made about proper names point to the conclusion that there is no hope that a normal speaker can know, for an arbitrary use of a demonstrative, a reference-fixing description that is “co-referential” with that use of the demonstrative. And I will again develop a little bit a picture of demonstrative reference-fixing conventions as conventions giving merely sufficient conditions for reference and reference failure, and argue that it does a good job of dealing with some test cases.

Talk 3: Arabic Numerals

Arabic numerals and other mathematical symbols appear relatively frequently in writings on reference, but their metasemantics (or even their semantics) is rarely if ever considered in detail in this literature. Descriptivists often take Arabic numerals as obvious examples of disguised relatively sophisticated mathematical descriptions, while anti-descriptivists take them as semantically unstructured terms that get their referents fixed by those same descriptions. I will argue for a different view, on which Arabic numerals are semantically unstructured terms that are intended to get their referents fixed by relatively unsophisticated descriptions. And I will argue that these descriptions do probably fix some referents for the numerals, against popular eliminativist (nominalist) views in the philosophy of mathematics. These referents—the numbers—are probably certain kinds or properties of pluralities.

Talk 4: Terms for Natural Kinds

The points made in talk 1 have analogues for natural kind terms, including the feasibility of a picture of natural kind term reference-fixing conventions as conventions giving merely sufficient conditions for reference and reference failure. A different question is whether this reference-fixing mechanism does fix appropriate referents. A growing literature, especially in the philosophies of chemistry and biology, argues that the classical Kripke-Putnam mechanism cannot do that, and often reaches broadly eliminativist conceptions of (at least ordinary) natural kinds. I will examine some aspects of this literature and I will argue that, while they do point to probable errors in Kripke’s and Putnam’s identifications of the referents of ordinary natural kind terms, they do not threaten a more reasonable picture of these referents as scientifically irreducible kinds.