Learning Cybernetics
Steve Heims, a physicist turned historian of science, explored two influential figures in early cybernetics in his excellent book, John von Neumann and Norbert Wiener (MIT Press, 1980, 547 pp.). Now he has turned his attention to a wider group of cybernetic forebears, those who gathered at the annual Macy conferences 1946 through 1953.
The Cybernetics Group (MIT Press, 1991, 334 pp.) is as quirky and enjoyable a piece of science history as the previous book, presenting in turns the intellectual biographies of the key people involved (McCulloch, Mead, Bateson, Kubie, Savage, Frank and more) and the debates at the conferences, as revealed through the extensive recorded proceedings.
Heims' interest is with science as a process occurring through dialogue and group exchanges. Thus, in true cybernetic spirit, he ends the book with a phrase, "The conversation continues," that mirrors the title of one forerunner of the present journal, the American Society for Cybernetics newsletter "Continuing the Conversation."
Archival work of similar historical importance has been performed by Rodney E. Donaldson, tender of the Bateson archives, who published a collection of previously unpublished, little-known or hard-to-get papers by Bateson: A Sacred Unity: Further Steps to an Ecology of Mind. A. Cornelia & Michael Bessie Book. An Imprint of Harper Collins Publishers, 1991, pp. 364.
The younger Bateson, Mary Catherine, who recently enjoyed bestseller status in the US with her multi-woman biography Composing a Life (Atlantic Monthly Press, 1991), is reportedly at work on a manuscript that develops her father's thinking on learning and adaptation. She argues for an empathic and ethical way of knowing, according to which knowledge arises out of action and participation in the world, rather than from the detached position of the "objective" observer. Tentative title: Peripheral Visions: Learning Out of the Corner of Your Eye (Harper Collins, 1994?).
Complexity
The next two years are going to see a handful of semi-popular books on the science on complexity. All of them lean heavily on work done at two New Mexico institutions that have dedicated themselves to the study of complex, self-organizing phenomena, such as nerve cells in the brain, genes in the genome, species in an ecosystem and consumers in an economy.
One institution is the Center for Non-Linear Studies at the federally funded Los Alamos National Laboratories, the other is the independent Santa Fe Institute (SFI). The former is a long-time leader in the physics and computer-science aspects of systems with simple elements and complex overall behavior, while the SFI is only half a dozen years old and harbors broader, interdisciplinary ambitions.
One of SFI's trustees is Murray Gell-Mann, Nobel-Prize winning discoverer of the quark, who has since developed an interest in the vagaries of complex systems. His 32-page proposal for a book on this topic, The Quark and the Jaguar, was well handled by his New York literary agent: within a few weeks it was sold to Bantam and over fifteen foreign publishers, fetching advances at well over a million dollars (this according to the trade magazine Publishers Weekly).
Judging by these amounts, which are right up there in the stratosphere with Stephen King and Judith Krantz, this book is poised to continue the mega-bucks bestseller tradition of Hawking's A Brief History of Time, Gleick's Chaos and Penrose's The Emperor's New Mind.
Stuart Kauffman, a University of Pennsylvania biochemist and sometime paricipant the Gordon conference on cybernetics, is a regular at Santa Fe Institute seminars. Drawing on his experience there and his own extensive research on genetic networks, he is under contract to write a book titled Complexity: The Discovery of Anti-Chaos. Unlike the semi-retired father figure Gell-Mann, who is promoting a field somewhat alien to his training, the younger Kauffman is an active researcher in the midst of things.
Two science writers have cast their eyes on complexity research. Complexity is the title of a Macmillan publication by Roger Lewin, whose previous books include Bones of Contention and, with Richard Leaky, Origins. The forthcoming volume presents a story in the manner of Gleick' Chaos, that is, a story of the people, ideas and places that come togehter to make a new science. Presented are Stephen Wolfram and cellular automata, Danny Hillis and the Connection Machine, Manfred Eigen and life as a dynamic system, Kauffman's new Darwinism and likely many more.
The other journalistic account is by Kevin Kelly, hacker organizer and editor of the Californian eco-techno magazine Whole Earth Review (formerly Co-Evolution Quarterly). Emergent Worlds: The Rise of Artificial Evolution examines the unintentional self-evolution of organic and man-made systems, computer simulations, cybernetic worlds and virtual realities -- such dreams as stuff in the 1990's is made of. The book is due out in early 1993, from Addison-Wesley.
Lastly, rumor has it that Christopher Langton, a Los Alomos biologist and computer expert, is at work on a book on Artificial Life (AL), a research program he has helped launch by hosting conferences on the topic. AL promises to do for biological order and evolution what Artificial Intelligence promised (but failed) to do for the mind.
While we wait for Langton to write his probably serious treatment and some science journalist to elevate AL to the next New Science in the making, a book by a Consulting Editor to the present journal, the Danish biologist Claus Emmeche, seems to be the only popularly accessible title on AL on market: Det levende spil: Biologisk form og kunstigt liv ("The Living Game: Biological form and AL") (Munksgaard, Copenhagen; so far only available in Danish).
Neural Philosophy
The 1980's saw a surge of interest in neural networks, the parallel computers modelled on the architecture of the nervous system. Just as the conventional von Neumann-computer led countless researchers and philosophers to ponder the von Neumann-like aspects of the mind, so the neural computer is beginning to inspire serious reflection on the part of philosophers.
Patricia Smith Churchland, a current MacArthur Fellow and professor at one of the major centers of cognitive science and neural network research, University of California at San Diego, got off to an early start in 1987 with her 550-page Neurophilosophy: Toward a Unified Science of the Mind-Brain (MIT Press). The book features a presentation of elementary neuroscience, a review of functionalist philosophy of mind and a sketchy 80-page section on "A Neurophilosophical Perspective" that presents odds and ends from recent research in neural modelling.
Since then, Churchland has prudently teamed up with a neural network expert, Terrence Sejnowski, Director of the Salk Institute's Computational Neurobiology Laboratory. Their book, The Computational Brain (MIT Press, May 1992, 576 pp.) presents more neural network research with a view to what is biologically relevant. The publisher's preview speaks of a focus on visual perception, learning and memory, and sensory-motor integration, while general principles applicable to other levels of neural and mental organization are abstracted along the way.
A promising-looking compilation of texts is the forthcoming The Philosophy of Mind: Classical Problems, Contemporary Issues (eds. Brian Beakley and Peter Ludlow, MIT Press, 464 pp., June 1992). Included are classics from Plato and Hobbes to James and Ryle, as well as a dozen more recents ones from either side of the AI vs. neural networks debate: Chomsky, Fodor, Dennett, Kripke, vs. McClelland, Rumelhart, Hinton, Smolensky and Patricia S. Churchland.
Somewhat distanced from this fray we find the most recent philosophical effort by Francisco Varela, The Embodied Mind: Cognitive Science and Human Experience. It was co-written with Evan Thompson, a young philosopher and son of the cultural critic Willian Irwin Thompson, and with Eleanor Rosch, a Berkeley psychologist (MIT Press, 1991, 308 pp.). Previously circulated in xerox copies as "Worlds Without Ground," the book says that although the neural network approach to modelling the human mind may include much more of the emergent processing involved in learning, perception and intuition so ignored by rationalistic AI, something very important is still left out: the body.
To include the body, the authors propose an "enactive" approach, which derives from Varela's previous work with Humberto Maturana. This centers on the bodily actions performed by the agent and their role in shaping experience and intelligence. One prominent example and source of inspiration used throughout the book is the authors' personal experience with a particualr brand of Buddhist meditational practice. -- The book is a very unusual and also accessible attempt to bring a spiritual perspective to cognitive science.
Return to the content of this issue
Return to the Cybernetics and Human Knowing Homepage