ROBERT A. WILSON and FRANK C. KEIL (Eds.) The MIT Encyclopedia of the Cognitive Sciences Cambridge, MA: The MIT Press, 2001, 1,096 pages (ISBN 0-262-73144-4, US$65, Softcover)
Reviewed by DANIEL BUB
Cognitive science is the banner under which the combined forces of philosophy, psychology, the analysis of computational intelligence, linguistics, and anthropology have rallied in an attempt to make headway against the ultimate scientific question: How do the capabilities of the human mind emerge from physical matter? Recently, another formidable ally, the field of neuroscience, or more specifically, cognitive neuroscience in its new setting, has added brain-based methodologies to the alliance. It is an exciting mixture of disciplines, its activities driven by the common goal, at least in principle, of advancing our understanding of human cognition. The MIT Encyclopedia of the Cognitive Sciences (MITECS) is an attempt to put together a summary of the "...full range of concepts, methods, and results derived and deployed in cognitive science over the last twenty-- five years" (p. xiii), a description given by the editors in the preface of this ambitious undertaking.
An encyclopedia can be evaluated according to two criteria: How are the entries organized? and What has been left out, given the (necessarily) abbreviated exposition of concepts, methods and results? MITECS has 471 articles relevant to cognitive science presented in alphabetic order, beginning with one entitled "Aboutness" and ending with one on "X-bar Theory." Each article is no longer than the length of this review and includes at the end a list of additional entries that the reader can peruse to further explicate the topic. To organize this impressive collection, the volume begins with an overview (average length about 20 pages) of each of the six domains that make up the content of MITECS, written by one or two consulting experts. As summary statements, part of their rationale is to provide a key to the linkages between different articles. This goal is difficult to accomplish if one is seeking an introduction that makes transparent the threads connecting the diversity of ideas summarized in MITECS, and in many instances the reader is simply told in effect that if he or she wishes to know more about topic X, he or she should also consult topics A, B, C, etc. This limitation one must accept as inevitable given the logistics of the enterprise, and the overviews are valuable tools that anyone who has taught a course in the foundations of cognitive science will relish.
More problematic in reading these overviews of the subcomponents of the field is the impression that the different domains, while showing signs of some integration at the margins, often make little contact with each other at the core. For example, the review section on "Neuroscience" includes a discussion of the neural mechanisms involved in translating a decision into action but makes no reference at all to the problem of the homunculus (what part of the brain decides, in the end, to act?), a major topic analyzed in the corresponding section on "Philosophy." The section on "Computational Intelligence" has little use for the evidence from neuroscience that human cognition is made up of functionally independent modules. For detailed theoretical modeling of cognitive architectures, we learn in this section that the evidence from neuroscience is too subjective and informal, and the nature of the connections between modular components remains obscure. Thus, "...the basic organizational principles of intelligence are still up for grabs" (p. lxxvi).
The summary of "Psychology" is perhaps less perfidious in connecting up with the other members of the cognitive alliance, but only because its style is more a synopsis of history and empirical methods than an analysis of core methodological issues in the field. …