Reply to Currie

Abstract

In 'Frege's Context Principle'1 I explained Frege's adoption and later abandonment of the Context Principle (CP) by relating it to Frege's views on identity statements. CP says that words only have meaning in the context of a sentence. Frege's early theory of identity (El) claimed that in the context of an identity statement words refer to themselves rather than their usual 'content'. This theory was supplanted by the treatment of identity that issues from the theory of sense and reference (SR) contained in Uber Sinn und Bedeutung (SB).2 Gregory Currie3 argues that my explanation fails. In doing so he misconstrues my meaning on a number of occasions due in large measure to too concise expression on my part. Currie also makes substantive claims that require answer. In his conclusion, for example, he betrays a fundamental misunderstanding of Frege's attitude to the complexities introduced into SR, as it applies to natural languages, by failures of intersubstitution salva veritate of terms with the same (customary) reference. First let me point out that contrary to Currie's contention (p. 543) I nowhere claimed that CP expresses, in part or otherwise, El-they are obviously logically independent. What I did do was suggest that El motivated Frege's adoption of CP, for El entails that word-reference (word-meaning) changes, depending on context. There is simply no question of inferring CP from El or vice versa. As Currie notes, I myself, following Michael Dummett, observed that CP does not entail that word-meaning may change from sentence to sentence. It is of course correct to say, as Currie does (p. 543), that El is not 'stated or alluded to' anywhere in Grundlagen. But what weight we should attach to that fact is unclear. Frege held onto a version of El until perhaps as late as i 89o.4 It is therefore quite in order to invoke it in explaining Frege's adoption of CP. Currie implies that the whole purpose of CP is to combat psychologism when he states that there is no textual evidence suggesting otherwise (p. 543). At least one passage in Grundlagen indicates that CP helps to overcome both psychologism and physicalism: 'only by adhering to [CP] can we [. . .] avoid a physical view of number without slipping into a psychological view of it'.5 This is borne out by CP's legitimation of contextual definitions: the latter are incompatible with any atomistic semantics, not just the ideational variety Frege discusses under the heading 'psychologism'. (As Michael Resnik has pointed out,6 a non-atomistic