Steven Pinker is professor of cognitive science at MIT and author of The Language Instinct (Harper Perennial, 1995) and Language Learnability and Language Development (Harvard University Press, 1996).

"How does the mind work? Many books give you airy-fairy musings, with lots of talk about self-organization and the ascent of man, but little insight into concrete puzzles. Instead of this cosmic mumbo-jumbo, read Ray Jackendoff's Consciousnes s and the Computational Mind (MIT, 1987). Jackendoff, a linguist with interests in every aspect of the mind and a professional musician, tells a coherent, plausible story on the nature of language, music, vision, and awareness. Jackendoff conveys clearly some of the most interesting ideas that cognitive scientists have come up with on how we enjoy music, talk, and see. His discussion of consciousness is one of the best around, with straightforward insights-and no delusions of grandeur."

Douglas Hofstadter is professor of cognitive science at Indiana University and author of the Pulitzer Prize-winning Godel, Escher, Bach: An Eternal Golden Braid (Basic, 1979) and Fluid Concepts and Creative Analogies (Basic, 1995).

"The idea of Holes and Other Superficialities, by Roberto Casati and Achille Varzi (MIT, 1994), is wonderfully counterintuitive: The authors want us to think of absences as full-fledged cognitive entities. The book describes and discusses a grand variety of holes-holes in doughnuts, tunnels through blocks, flowerless gaps in regularly-spaced flowerbeds, and hundreds more. There are an enormous number of beautifully-rendered illustrations of every imaginable (and often never-before-imagined) type of hole, and provocative analyses of prior attempts to make holes into 'second-class' entities, rather than full-fledged beings in and of themselves. The overlap with philosophical issues of every sort is marvelous, and the authors have a delightful sense of humor."

Stephen Stich is professor of philosophy and cognitive science at Rutgers University and co-editor of Mental Representation: A Reader (Oxford, 1991).

Stich says that the central problem of cognitive science is showing how the mind works like a computer. "The job of the cognitive scientist," he claims, "is to do 'reverse engineering' on that computer in order to uncover the algorithms the mind is employing. Until recently, however, few scholars asked what kinds of algorithms it would have made sense for natural selection to provide us with. The bible of the new evolutionary approach is The Adapted Mind: Evolutionary Psychology and the Generation of Culture (Oxford, 1992), edited by Jerome H. Barkow, Leda Cosmides, and John Tooby. This book asks: What sorts of algorithms would have contributed to the fitness of our hunter-gatherer forebears? Asking this question generates a cascade of hypotheses about algorithms that we use to select suitable sex partners, to provoke jealous rage, to trigger depression, and more."

Peter Kramer, Clinical Professor of Psychiatry at Brown University and author of Listening to Prozac (Viking, 1993) and Moments of Engagement: Intimate Psychotherapy in a Technological Age (Norton, 1989).

Kramer particularly enjoys books that stand on the boundary of psychiatry and philosophy. He says Lessons from an Optical Illusion: On Nature and Nurture, Knowledge and Values (Harvard, 1995), by Edward Hundert, is a stunning and overlooked example. "Hundert begins with the question of how we construct reality. He walks readers through the contributions of Kant, Descartes, and Hegel before admixing the findings of neurobiology. Hundert shows how clarifying cognitive science can be for central problems of Western philosophy. Along similar lines, Harvie Ferguson's Melancholy and the Critique of Modernity: Soren Kierkegaard's Religious Philosophy (Routledge, 1995) opens with a breathtaking chapter on the role of depression in Renaissance and Reformation philosophy."

Esther Thelen is professor of psychology and cognitive science at Indiana University and co-author of A Dynamic Systems Approach to the Development of Cognition and Action (MIT, 1994).

Thelen says that the best cognitive science books put real-world constraints on models of the mind. One is Gerald Edelman's Neural Darwinism: The Theory of Neuronal Group Selection (Basic, 1987). "Edelman's insights into neural architecture show how cognition may arise from processes such as perceiving, moving, learning, and remembering-without invoking implausible devices in the brain. I also like Barry E. Stein's and M. Alex Meredith's The Merging of the Senses (MIT, 1993); it demonstrates that the so-called 'binding problem' isn't a problem. The binding problem concerns how different sensations (say, the sight, smell, taste, and crunch of peanut brittle) are integrated in the brain to provide a unified perception ("I am eating peanut brittle!"). The authors show that sensory convergence is not a higher-order brain process, but widespread and fundamental. The implications: Reality need not be constructed from isolated precepts, but selected and refined through experience."

Charles Taylor is professor of biology at UCLA and co-editor of Artificial Life ii (MIT, 1992).

"My daughter, Amy, is eight years old," says Taylor. "When her father was born, in 1945, there were almost no computers-and nearly all computation was by slide rule. If advances in cognitive science and computers continue as they have, then one can surmise that Amy's children will grow up surrounded by robots, taking them for granted as Amy takes computers for granted. Hans Moravec provides an intelligent projection of what we can expect from robotics during the next thirty to forty years in Mind Children: The Future of Robot and Human Intelligence (Harvard, 1988). Some of his projections are alarming, especially in regard to the ethical issues that we are only now beginning to confront."

Annette Karmiloff-Smith is senior researcher at London's Medical Research Council and author of Beyond Modularity: A Developmental Perspective on Cognitive Science (MIT, 1992).

"Biology and Cognitive Development: The Case of Face Recognition, by Mark H. Johnson and John Morton (Oxford, 1991), tells fascinating stories about the development of mother-recognition in baby chicks and human infants. We've long known that newly hatched chicks will 'imprint' on their mothers. Johnson and Morton's study extends this insight, showing that chicks have a spontaneous tendency to look at 'hen-like' things right from hatching, ensuring that they will successfully imprint on their mums. Similarly, the human newborn infant knows something about what faces look like (three blobs in the place of eyes and mouth) and prefers at birth to stare at face-like stimuli rather than anything else. That's good news for mothers."

--Dan Zalewski

Are there any great books you think we missed? Let us know.