Abstract

The objective of this dissertation is to examine Millikan's teleosemantics as an attempt to explain intentionality in a naturalistically way. Naturalistic semantics in general tries to explicate intentionality as a physical or 'natural' relation between mental representations, understood as physical structures, to things in the world. The essence of the problem, from a naturalistic perspective, is to explain how meaning arises out of purely physical structures. The main difficulties for naturalists are grouped into two problems, which I will label the 'normative problem' and the 'descriptive problem' respectively. The normative problem is to explain what constitutes semantic normativity. The descriptive problem is to explain how mental representation acquire their particular semantic contents. ;Millikan answers the normative problem by reducing semantic norms to biological norms. In order to exhibit some strengths of Millikan's approach, I contrast two main competing theories about the notion of function , the etiological theory of function vs. Cumminsian functional analysis. I argue that the etiological theory of function gives us a more promising perspective for answering the normative problem. ;The etiological approach to functions leads us to a counter-intuitive conclusion that a 'swampman' that suddenly arises in a swamp by improbable coincidence but otherwise a perfect physical duplicate of a normal person, doesn't have any intentional states. I argue that, if we assume that 'intentional states' are normative, the teleosemantic conclusion about the swampman can be defended as an inference to the best explanation of this normative character. ;For the descriptive problem, Millikan suggests a consumer-based teleosemantics. According to Millikan, the content of a representation is identified with the Normal condition which should be presupposed for the consumer device of that representation to perform its proper function in accordance with the most proximate Normal explanation. In a nutshell, fixation of content is reduced to the matter of giving a univocal Normal explanation of the consumer's proper function. I consider two main criticisms against this approach. The first problem raises questions about the privileged status of consumer-level explanations in fixing representational content over other levels of Normal explanation all of which are evolutionarily and causally adequate. I will argue that teleosemanticists can live with this criticism. The second problem concerns the role of counterfactuals in teleosemantics for answering Fodor's disjunction problem. To answer some disjunction cases that critics have come up with, Millikan should invoke some sort of counterfactual analysis of causal relevance. But, it is very doubtful that such a counterfactual analysis is readily available to Millikan, given her resolute commitment to historical approach to functions. This is where I think that Millikan's theory encounters its most serious problems. ;My tentative conclusion is that, whereas Millikan's theory provides a very fruitful perspective for understanding the normative character of intentional content, it fails to give a satisfactory answer to the descriptive problem