What is it to have a concept? What is it to make an inference? What is it to be rational? On the basis of recent developments in semantics, a number of authors have embraced answers to these questions that have radically counterintuitive consequences, for example: • One can rationally accept self-contradictory propositions (e.g. Smith is a composer and Smith is not a composer).• Psychological states are causally inert: beliefs and desires do nothing. • The mind cannot be understood in terms of folk-psychological concepts (e.g. belief, desire, intention). • One can have a single concept without having any others: an otherwise conceptless creature could grasp the concept of justice or of the number seven. • Thoughts are sentence-tokens, and thought-processes are driven by the syntactic, not the semantic, properties of those tokens. In the first half of Conceptual Atomism and the Computational Theory of Mind, John-Michael Kuczynski argues that these implausible but widely held views are direct consequences of a popular doctrine known as content-externalism, this being the view that the contents of one's mental states are constitutively dependent on facts about the external world. Kuczynski shows that content-externalism involves a failure to distinguish between, on the one hand, what is literally meant by linguistic expressions and, on the other hand, the information that one must work through to compute the literal meanings of such expressions. The second half of the present work concerns the Computational Theory of Mind (CTM). Underlying CTM is an acceptance of conceptual atomism – the view that a creature can have a single concept without having any others – and also an acceptance of the view that concepts are not descriptive (i.e. that one can have a concept of a thing without knowing of any description that is satisfied by that thing). Kuczynski shows that both views are false, one reason being that they presuppose the truth of content-externalism, another being that they are incompatible with the epistemological anti-foundationalism proven correct by Wilfred Sellars and Laurence Bonjour. Kuczynski also shows that CTM involves a misunderstanding of terms such as “computation”, “syntax”, “algorithm” and “formal truth”; and he provides novel analyses of the concepts expressed by these terms. (Series A)

Conceptual atomism claims that most concepts cannot be decomposed into features, so that the conjunction of the features is equivalent to the concept in question. Conceptual atomism of this type is incompatible with many other semantic approaches. One of these approaches is justificationist semantics. This book assumes conceptual atomism. Justificationist semantics in its pure form, therefore, has to be wrong. Nevertheless, its epistemological approach to questions of evaluations and semantic rules could still stand. The main question is how conceptual atomism can be combined with some justificationist ideas. This new synthesis centres on the representational theory of mind and 'internalist' semantics, but ties these to ideas which stress the epistemic commitments that accompany successful assertions.

Preface p. ix 1 Introduction p. 1 2 The Central System as a Computational Engine p. 27 3 Jerry Fodor's Globality Challenge to the Computational Theory of Mind Kirk Ludwig p. 65 4 What LOT's Mental States Cannot Be: Ruling out Alternative Conceptions p. 91 5 Mental Symbols p. 111 6 Idiosyncratic Minds Think Alike: Modes of Presentation Reconsidered p. 135 7 Concepts: A Pragmatist Theory p. 159 8 Solving the Frege Cases p. 183 9 Conclusion p. 229 References p. 233 Index p. 249.

This volume features the complete text of the material presented at the Twenty-Fourth Annual Conference of the Cognitive Science Society. As in previous years, the symposium included an interesting mixture of papers on many topics from researchers with diverse backgrounds and different goals, presenting a multifaceted view of cognitive science. The volume includes all papers, posters, and summaries of symposia presented at this leading conference that brings cognitive scientists together. The 2002 meeting dealt with issues of representing and modeling cognitive processes as they appeal to scholars in all subdisciplines that comprise cognitive science: psychology, computer science, neuroscience, linguistics, and philosophy.

Papers on the philosophy of mind and philosophical logic. Topics covered include probabilistic causation, the nature of formal truth, the role of language in thought, conceptual atomism, simulated vs. actual intelligence, and the nature of emotion.

A philosophical refashioning of the Language of Thought approach and the related computational theory of mind. The language of thought (LOT) approach to the nature of mind has been highly influential in cognitive science and the philosophy of mind; and yet, as Susan Schneider argues, its philosophical foundations are weak. In this philosophical refashioning of LOT and the related computational theory of mind (CTM), Schneider offers a different framework than has been developed by LOT and CTM's main architect, Jerry Fodor: one that seeks integration with neuroscience, repudiates Fodor's pessimism about the capacity of cognitive science to explain cognition, embraces pragmatism, and advances a different approach to the nature of concepts, mental symbols, and modes of presentation. According to the LOT approach, conceptual thought is determined by the manipulation of mental symbols according to algorithms. Schneider tackles three key problems that have plagued the LOT approach for decades: the computational nature of the central system (the system responsible for higher cognitive function); the nature of symbols; and Frege cases. To address these problems,] Schneider develops a computational theory that is based on the Global Workspace approach; develops a theory of symbols, "the algorithmic view"; and brings her theory of symbols to bear on LOT's account of the causation of thought and behavior. In the course of solving these problems, Schneider shows that LOT must make peace with both computationalism and pragmatism; indeed, the new conception of symbols renders LOT a pragmatist theory. And LOT must turn its focus to cognitive and computational neuroscience for its naturalism to succeed.

In this book Jerry Fodor contrasts his views about the mind with those of a number of well-known philosophers and cognitive scientists, including John McDowell, Christopher Peacocke, Paul Churchland, Daniel Dennett, Paul Smolensky, and Richard Dawkins. Several of these essays are published here for the first time. The rest originated as book reviews in the Times Literary Supplement, the London Review of Books, or in journals of philosophy or psychology. The topics examined include cognitive architecture, the nature of concepts, and the status of Darwinism in psychology. Fodor constructs a version of the Representational Theory of Mind that blends Intentional Realism, Computational Reductionism, Nativism, and Semantic Atomism.