In several works, Ruth Millikan (1998a, 2000, 2006) has developed a ‘teleosemantic’ theory of concepts. Millikan’s theory has three explicit desiderata for concepts: wide scope, non-descriptionist content, and naturalism. I contend that Millikan’s theory cannot fulfill all of these desiderata simultaneously. Theoretical concepts, such as those of chemistry and physics, fall under Millikan’s intended scope, but I will argue that her theory cannot account for these concepts in a way that is compatible with both non-descriptionism and naturalism. In these cases, Millikan’s view is subject to the traditional ‘indeterminacy problem’ for teleosemantic theories. This leaves the content of theoretical concepts indeterminate between a descriptionist and non-descriptionist content. Furthermore, this problem cannot be overcome without giving up the naturalism desideratum. I suggest that the scope of Millikan’s theory should be limited. At best, the theory will be able to attribute naturalistic, non-descriptionist content to a smaller range of concepts.

What distinguishes a meaningless utterance from a meaningful term?  While one might say that, within the context of Ruth Millikan\'s teleosemantics, it is a term\'s having a proper function that distinguishes it from a meaningless utterance, I propose that the distinction can be made with reference to the history of the term.  Using evolutionary game theory, I offer a way to clarify the distinction between the meaningless and the meaningful.  I reject the possibility of correlating meaning with an evolutionarily stable strategy as this does not seem to be consistent with how communication works or with Millikan\'s theory.  Instead, when a term has meaning, the function category of that term corresponds to an evolutionarily stable state composed of both speaker and hearer strategies. / Master of Arts