Connectionist models provide a promising alternative to the traditional computational approach that has for several decades dominated cognitive science and artificial intelligence, although the nature of connectionist models and their relation to symbol processing remains controversial. Connectionist models can be characterized by three general computational features: distinct layers of interconnected units, recursive rules for updating the strengths of the connections during learning, and “simple” homogeneous computing elements. Using just these three features one can construct surprisingly elegant and powerful models of (...) memory, perception, motor control, categorization, and reasoning. What makes the connectionist approach unique is not its variety of representational possibilities or its departure from explicit rule-based models, or even its preoccupation with the brain metaphor. Rather, it is that connectionist models can be used to explore systematically the complex interaction between learning and representation, as we try to demonstrate through the analysis of several large networks. (shrink)

The field of neuroimaging has reached a watershed. Brain imaging research has been the source of many advances in cognitive neuroscience and cognitive science over the last decade, but recent critiques and emerging trends are raising foundational issues of methodology, measurement, and theory. Indeed, concerns over interpretation of brain maps have created serious controversies in social neuroscience, and, more important, point to a larger set of issues that lie at the heart of the entire brain mapping enterprise. In this volume, (...) leading scholars -- neuroimagers and philosophers of mind -- reexamine these central issues and explore current controversies that have arisen in cognitive science, cognitive neuroscience, computer science, and signal processing. The contributors address both statistical and dynamical analysis and modeling of neuroimaging data and interpretation, discussing localization, modularity, and neuroimagers' tacit assumptions about how these two phenomena are related; controversies over correlation of fMRI data and social attributions ; and the standard inferential design approach in neuroimaging. Finally, the contributors take a more philosophical perspective, considering the nature of measurement in brain imaging, and offer a framework for novel neuroimaging data structures. Contributors: William Bechtel, Bharat Biswal, Matthew Brett, Martin Bunzl, Max Coltheart, Karl J. Friston, Joy J. Geng, Clark Glymour, Kalanit Grill-Spector, Stephen José Hanson, Trevor Harley, Gilbert Harman, James V. Haxby, Rik N. Henson, Nancy Kanwisher, Colin Klein, Richard Loosemore, Sébastien Meriaux, Chris Mole, Jeanette A. Mumford, Russell A. Poldrack, Jean-Baptiste Poline, Richard C. Richardson, Alexis Roche, Adina L. Roskies, Pia Rotshtein, Rebecca Saxe, Philipp Sterzer, Bertrand Thirion, Edward Vul The hardcover edition does not include a dust jacket. (shrink)

After people learn to sort objects into categories they see them differently. Members of the same category look more alike and members of different categories look more different. This phenomenon of within-category compression and between-category separation in similarity space is called categorical perception (CP). It is exhibited by human subjects, animals and neural net models. In backpropagation nets trained first to auto-associate 12 stimuli varying along a onedimensional continuum and then to sort them into 3 categories, CP arises as a (...) natural side-effect because of four factors: (1) Maximal interstimulus separation in hidden-unit space during autoassociation learning, (2) movement toward linear separability during categorization learning, (3) inverse-distance repulsive force exerted by the between-category boundary, and (4) the modulating effects of input iconicity, especially in interpolating CP to untrained regions of the continuum. Once similarity space has been "warped" in this way, the compressed and separated "chunks" have symbolic labels which could then be combined into symbol strings that constitute propositions about objects. The meanings of such symbolic representations would be "grounded" in the system's capacity to pick out from their sensory projections the object categories that the propositions were about. (shrink)

The claims that learning systems must build causal models and provide explanations of their inferences are not new, and advocate a cognitive functionalism for artificial intelligence. This view conflates the relationships between implicit and explicit knowledge representation. We present recent evidence that neural networks do engage in model building, which is implicit, and cannot be dissociated from the learning process.