Meaning and mental representation /

The primary aim of this lucid and instructive study is to discover what representation must amount to if orthodox computationalism is to be true and explanatory. Orthodox computationalism, "the computational theory of cognition," views cognition as the systematic manipulation of symbols, but it offers no account of the nature of representation. Cummins's original account explicates representation as a kind of simulation. The concept of representation required by the computational theory of cognition, Cummins argues, is the concept invoked when we say that "a graph or equation represents a set of data" or that "a parabola represents the trajectory of a projectile." Much of the book is devoted to exposition and criticism of influential work on representation, including that of John Locke, plus Jerry Fodor's Psychosemantics (CH, Jan '88), Fred Dretske's Knowledge and the Flow of Information (1981), and Explaining Behavior (CH, Oct '88), and Ruth Millikan's Language, Thought, and Other Biological Categories (1984). Cummins's book will be widely used by philosophers of mind and their graduate students. It is accessible as well to advanced undergraduates, who will appreciate the excellent overview of both the literature and the philosophical terrain. -A. R. Mele, Davidson College