Many philosophers hold that all thoughts are conceptually structured--that they are composed of concepts in much the way that a sentence is composed of words. My dissertation explores and ultimately challenges this view, drawing on empirical results from the cognitive sciences to argue that thoughts come in a variety of structures, many of which are nonconceptual. I begin the dissertation in chapter 1 by providing a broadly functionalist account of thought according to which thoughts are contentful mental states of a subject that causally and inferentially mediate between perception and action, are modifiable through learning and are stored in memory. In chapters 2 and 3 I then discuss the thesis that thoughts are conceptually structured--i.e., that their contents or vehicles are structured like sentences. One consequence of this thesis is the Generality Constraint, which holds that the thoughts one can think are closed under recombination of the constituents of the sentences which best express them.

Having generated an understanding of the thesis that thoughts are conceptually structured, I turn in the second half of the dissertation to evaluate its truth. Chapter 4 considers several arguments that philosophers have marshaled in its favor. I contend that while these arguments tend to show that some thoughts must be conceptual, they leave open the possibility that other thoughts might be nonconceptual. In chapter 5 1 argue that this possibility is actualized by showing that so-called analog magnitude thoughts --which represent magnitudes such as number, time, distance and rate--engender violations of the Generality Constraint. In chapter 6 I then argue that two further types of thoughts--imagistic and cartographic--also exhibit properties which make them nonconceptual. Thus, just as we use various representational kinds in everyday life--including sentences, pictures, maps and thermometers--our brains employ various mental representations in thought. I conclude chapter 6 with a discussion of how these various kinds of thought interface with one another. One benefit of distinguishing different varieties of thought, I argue in the appendix, is that it has the potential to illuminate the continuities and disparities between human and animal minds.