The MIT Encyclopedia of the Cognitive Sciences (Bradford Book) (Bradford Books) [Paperback]

The good news: There are some truly excellent articles in this book. Microcolumns and macrocolumns, cerebellar chips, the pathways of the visual system - you can read this book and find out a hundred amazingly cool things that you never even realized you desperately needed to know. Oddly enough, MITECS is also a pretty good as an encyclopedia - if you suddenly need to know more about vision, you'll find what you need to know in "Visual Anatomy and Physiology". (Or "Visual Processing Streams". Or "High-Level Vision". Or "Computational Vision". Or "Mental Rotation". You do need to do a certain amount of hunting, if it's a sufficiently broad subject. More than half the cerebral cortex is devoted to vision - see "Mid-Level Vision" - and MITECS reflects this fact.)

MITECS *excels* as an authoritative reference; you'll almost never need to quote anything else. If you're familiar with cognitive science, you'll often laugh when you get to the end of an article and see the author's byline: "Columns and Modules" by William Calvin, "Chinese Room Argument" by John Searle, "Evolutionary Computation" by Melanie Mitchell, "Evolutionary Psychology" by Leda Cosmides and John Tooby.

The bad news: If you try to read MITECS linearly, you will find that many of the articles, perhaps even a majority, are eminently skippable. (For the record, I read them anyway.) As all of the articles were written by independent individuals - none of whom could read the book first, since it didn't exist yet - there is understandably a great deal of duplication of information. Every third author feels the need to inform you that the mind is a computational information-processing system. (If I had one request to make of the hundreds of authors who write the next edition, it would be: "Skip all the introductory material and the philosophy and try to pack in as much useful detail as you can.") There are also some understandable problems with depth of coverage, made worse by the aforesaid tendency to write introductions; whenever I read an article about a topic that I had earlier studied in more detail, it really brought home the realization that each of these 471 articles tries to cover a topic about which *multiple* entire books have been written.

There are several things I'd like to see in future editions of this book. First and foremost is *less philosophy* and more focus on concrete details, particularly *surprising* details, or details that have something substantial to say about how the mind works. I don't want to know what David Hume thought about causality; I want to know if anything interesting happens when research subjects are asked to reason about causality. (I must also confess myself uninterested in most of the biographical articles that form much of MITECS - but then, that's probably because I'm not using it to study history.) Finally, I would like to see a neuroanatomical index as well as a table of contents. It's already a big book, but they can afford another six pages to show a detailed neuroanatomical map, with names for the areas, and references to the appropriate sections of the book. Such a map would be an enormous help to those of us trying to build up a concrete visualization of the brain.

Conclusion: This is a *really good* book. It's not so much "a good book with a few drawbacks" as "an excellent book with tremendous potential for *even more* improvement", and I mean this in all seriousness. If you're a cognitive scientist, you have basically no choice but to buy this book. If you're a student of the mind or a cognitive hobbyist, then this may not be the *first* book you buy, but you will buy it sooner or later.

It's just such a great book.