Concept Cartesianism, Concept Pragmatism, and Frege Cases

This paper concerns the dialectal role of Frege Cases in the debate between Concept Cartesians and Concept Pragmatists. I take as a starting point Christopher Peacocke’s argument that, unlike Cartesianism, his ‘Fregean’ Pragmatism can account for facts about the rationality and epistemic status of certain judgments. I argue that since this argument presupposes that the rationality of thoughts turns on their content, it is thus questionbegging against Cartesians, who claim that issues about rationality turn on the form, not the content, of thoughts. I then consider Jerry Fodor’s argument that ‘modes of presentation’ are not identical with Fregean senses, and argue that explanatory considerations should leads us to reject his ‘syntactic’ treatment of Frege cases. Rejecting the Cartesian treatment of Frege cases, however, is not tantamount to accepting Peacocke’s claim that reasons and rationality are central to the individuation of concepts. For I argue that we can steer a middle course between Fodor’s Cartesianism and Peacocke’s Pragmatism, and adopt a form of Pragmatism that is constrained by Fregean considerations, but at the same time denies that concepts are constitutively tied to reasons and rationality. The philosophical debate over the nature of concepts has recently been cast as a battle between ‘Concept Cartesians’ and ‘Concept Pragmatists’ (Fodor 2003, 2004). Pragmatists claim that concepts are individuated in terms of the role they play in the cognitive lives of thinkers, e.g. in terms of their role in inference, perception, and judgment. Cartesians, on the other hand, hold that none of the epistemic properties of concepts are concept-constitutive. To possess the concept DOG, for instance, thinkers needn’t be willing to make any inferences concerning dogs, nor must they be willing to judge (say) that’s a dog when they are looking at a dog in good judging conditions. 1 Rather, Cartesians claim that to possess DOG thinkers must simply possess a mental representation that is ‘nomically locked’ onto the property of being a dog (Fodor 1998). On this view, possessing DOG is not a matter of being able to draw certain inferences or make certain perceptual judgments, but is simply a matter of being able to think about dogs (as such). The Cartesian’s theory of content is thus purely referential: DOG has its content, not in virtue of any of its inferential properties, but rather in virtue of standing in a certain nomological relation to the property of being a dog instantiated in the environment. There are well-known problems with any purely referential theory of conceptual content. Two problems stand out: those raised by Putnam’s Twin-Earth examples, which putatively show that concepts with (what we pre-theoretically think of as) the same content can differ in reference, and those raised by so-called ‘Frege cases’, which putatively show that concepts that 1 I follow the convention of using words written in small capitals to refer to concepts. When quoting, I change the author’s notation to fit mine.