Description and necessity : towards a cognitive science of word meaning

• Main Author: Braisby, Nicholas Richard
• Published: University of Edinburgh 1990
• Subjects: 150
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.641888

In this thesis we present a cognitive-scientific view of word meaning. We begin by spelling out some general enterprises of which this thesis may form a small part. Enterprises such as the entwining of facts concerning the cpistcmology of language use and semantic analysis; the orientation of a study of word meaning as part of a study of thought; and the restoration of a study of thought to the psychological domain. We also indicate some of the tools we employ in the rest of the thesis. One tool is that provided by Situation Theory and particularly its notion of conditional constraints. Another is a prc-thcorctic point that encourages us to distinguish certain semantical questions from related ontological ones. Another tool is a set of puzzles of word meaning an analysis of which forms one objective. These and other tools allow us to isolate some points at which theories of sense and word meaning may diverge. And this prepares the ground for a discussion of competing theories. Our discussion is informed by psychological concerns and it is these that we take to motivate a general account of word meaning, Sense Generation, and a specific version of this, the Relational View. Sense Generation, we argue respects linguistic constraints concerning ambiguity as well as psychological constraints concerning concepts. We argue, further, that the details of the Relational View are supported by psychological argumentation. And we meet our first objective which is to present an analysis of the various puzzles. We then turn to the philosophical literature in order to combat some well established positions and arguments which count against the efficacy of Sense Generation views. Sense Generation views hold that the meanings of words arc dcscriptional; opposing this view are the arguments of Kripke and Putnam. We consider first the literature on proper names whose meanings, it is argued, arc not dcscriptional. Our position with respect to this argument is that it be treated as an inadvertent reduction of the traditional conception of analyticity. We then turn our attention to the psychological literature on natural kind concepts and suggest that these do not standardly address the problems that Kripke and Putnam raise. Again, our suggestion is that with an alternative conception of analyticity, one resulting from Situation Theory and Sense Generation, the Kripkc-Putnam argument concerning natural kinds can similarly be viewed as a reductio. We spell out the Relational View of natural kind concepts. Our attention then turns to a major psychological theory of concepts, prototype theory. We argue that it fails to meet various constraints of a psychological nature and, further, that these constraints are met by the Relational View, which we re-label the Family of Constraints view. We attempt to adduce reasons for prototype theory's shortcomings by giving a detailed comparison between it and the Family of Constraints view.