Realism, understanding and truth

Khlentzos, Drew M.

Realism, understanding and truth

Date

1986

Authors

Khlentzos, Drew M.

Abstract

Realism, as Michael Dummett understands it, is a thesis about the meanings of sentences of a natural language. The Realist's thesis is that the correct model of meaning for those sentences is a truth-conditional one. Dummett has sought to prove that Realism thus defined faces insuperable objections. Those objections centre around the communicability of the Realist’s truth-conditions - how, when these truth-conditions are ones that can obtain without speakers being aware that they do, could speakers possibly evince in their linguistic behaviour an understanding of those conditions? This thesis is an attempt to meet Dummett's arguments against Realism. In the first chapter, I outline Davidson's form of Realism, ultimately disagreeing with his views on the nature of mind (§§1.1.2, 1.4), but endorsing his holistic approach to meaning, truth and interpretation. Meaning for Davidson is a theoretical notion, underdetermined by linguistic usage. Central to Davidson's position is the belief that truth is primitive, a belief which Frege argued for. I argue that Frege and Davidson are right about this and that this doctrine is inconsistent with a correspondence theory of truth (§1.2). The second chapter sets out Dummett's Anti-Realism, noting Dummett's requirements for acceptable theories of meaning. With one reservation, I accept Dummett's characterisation of Realism (§2.1 ). Dummett thinks that our use of language is guided by implicit knowledge of a theory of meaning for our language (§2.2), our grasp of all sentences, in particular the undecidable sentences, consisting in a grasp of their assertibility conditions. Essentially defeasible statements present a problem for Anti-Realism (§2.4). I then present Dummett's Manifestation Argument against Realism, questioning some central assumptions of that argument (§2.4). I conclude the chapter with a discussion of Dummett's objections to holistic theories of meaning- some of these are cogent, but others are based upon misunderstandings, I contend (§2.5). In chapter three, I critically discuss Dummett's idea that truth is a construct from the more primitive notion of correct assertibility. I argue that Dummett's formulation of the latter notion is unsatisfactory (§3.1.1) and that his most convincing argument for truth's arising from assertibility, which is based on our understanding of time and tense, does not, even if sound, prove what he needs to prove if he is to create problems for Realism (§3.1.2). I then examine Dummett's reasons for holding that a theory of meaning must contain a subpart which pairs truth-conditions with practical recognitional abilities on the parts of speakers (§3.2). I argue that first person avowals or others require a truth-condtional model for their meanings (§3.3 ). With this as a counterexample to a global Anti-Realist semantics, I seek to show that Dummett's Manifestation Challenge can be answered provided one can rebut Wittgenstein's Private Language Argument. Although I do not investigate whether it really does so, I suppose with Dummett that Wittgenstein's argument entails that meaning must be exhaustively manifest in use. I took at the most cogent form or this argument, which is due to Saul Kripke in the final chapter. I contend that Kripke's semantic scepticism is self-refuting (§5.2). The penultimate chapter seeks to examine the intuitionistic foundations of Dummett's Anti-Realism. I begin with a discussion of the intuitionist's philosophical position, attending particularly to his views on quantification over infinite totalities (§4.1 ). Dummett and Dag Prawitz have developed a proof-theoretic approach to the meanings or the logical constants (§§4.2,4.3). Dummett uses a generalised version of the proof-theoretic notion of a conservative extension to press for revision in our Realist-inspired logical practices (§§4.2,4.5). I examine and reject the most persuasive form of the argument for revisionism (§4.5.1 ). I argue that theories of meaning based upon the notions of assertibility or deniability cannot explicate the meanings of the logical constants (§4.4)- to do this, the Anti-Realist must develop an acceptable theory of truth; in this connection, Dummett's suggestions are inadequate (§4.:S). I conclude the chapter with a discussion of Dummett's ingenious attempt to justify deduction (§4.6). This is Dummett at his very best and here I argue that, somewhat in contrast to the overall tenor of my other conclusions, Dummett is absolutely right.