Can Russell Avoid Frege's Sense?

Abstract

In On Denoting Russell writes that Frege's distinction between sense and reference is an 'inextricable tangle'.' Russell had no misgivings in not postulating sense so long as there can be an adequate theory of meaning that does not suppose it, and Russell believed his theory of meaning was such a theory. Frege felt that any adequate account of meaning must include sense, and he provided some arguments why we need to postulate sense. His basic argument comes from his puzzle over identity. In this paper we argue that Russell's theory of meaning can deal with Frege's main reason for postulating sense. By Russell's theory of meaning we have in mind his view that the only intensional entity one need suppose is propositions. But though Russell's theory of meaning can handle the linguistic phenomena which prompted Frege to postulate sense, Russell's epistemology blocks his using his theory of meaning so as to avoid Frege's sense. Frege's basic reason for the postulation of sense is the puzzle over identity.2 Suppose that 'a' and 'b' are names for the same object. A statement of the form ra = a' is uninformative, while a statement of the form ra = b' can be informative. So the difference is the mode with which the denoted object is presented to the user of the name. Frege postulated sense to explain this difference. Thus sense is the mode of presentation associated with an expression which accounts for the informative value of statements. It is also, for Frege, the way of determining the reference of an expression and what an expression denotes in oblique contexts. For Frege the way to determine whether 'a' differs in sense from 'b' is to ask whether a statement of the form Fa = b' can be informative. It seems that the primitive way to determine whether such a statement is informative is to ask whether someone can believe that a is a and not believe that a is b. In other words, failure of substitution of coextensive names