In lieu of an abstract, here is a brief excerpt of the content:

ARTIFICIAL LIFE by Christopher G. Langton, ed. Addison-Wesley, Redwood City, CA, U.S.A., 1989. 654 pp., illus. Trade, $43.25; paper, $21.50. ISBN: 0-201-o934~; ISBN: 0-201-09356-1. Reviewed l7y RobertS. Lansdon, 3830 Annapolis Court, South San Francisco, CA 94080, U.S.A. This volume is a wide-ranging compilation of papers from the first of the Artificial Life conferences, held in September 1987. The meeting was sponsored by the Santa Fe Institute, which specializes in "the sciences of complexity". A common thread in the essays is the use of approaches from the physical sciences, computer science and mathematics to mimic and explore properties shared by living organisms and to model genetic and population interactions. Christopher Langton has edited a collection of papers whose quality is consistently high. His preface provides a unifying overview of the papers to come, and he presents in his own paper some thoughts on the conduct of research on the model systems termed 'artificial life' . Langton proposes that because superposition does not hold in nonlinear systemsbut is responsible for the interesting behavior (and life itself?)-a global approach is necessary in nonlinear studies. He further argues that analysis has not gotten far and suggests that synthesis presents better possibilities . This fits in very well with the computational school of physics and with the emerging area of experimental mathematics, which uses computers and usually graphics. The importance of hands-on interactive graphics experimentation as a means of building intuition is asserted by Peter Oppenheimer . Harold C. Morris suggests that work such as his simulated genetic engineering of artificial life with 'typogenetics ' is useful in part as a means to create an atmosphere in which 'doing' allows an atmosphere in which approaches can be refined. Oppenheimer , however, warns that the photographic quality of current computer graphics can be seductive-that a picture can disprove a theory but cannot prove anything to be correct. He also notes that the cross-fertilization of genetic engineering with software is the "dream and nightmare" of every investigator of artificial life. Oppenheimer introduces the notion of numerical genes. These are software attributes that can be manipulated by a chosen set of rules for the genetics of the artificial life model. Eight full-color plates illustrate the results of his methods in constructing images of artificial trees and plants as well as some surrealistic hybrid objects. The applications to interactive or intelligent computer graphics and to kinetic art warrant further exploration. Cellular automata and reactiondiffusion equations with flow graphs result in intricate two-dimensional patterns that illustrate the paper by Tamayo and Hartman. Steen Rassmussen uses graph theory (with and without attached kinetic information) in his study of auto-catalysis, a likely stage in the development of primitive life forms. Work on autocatalysis has also been done by Norm Packard, who presents here a paper on adaptation by model organisms to intrinsic and extrinsic environmental factors resulting from the dynamics of many internal reactions or of interactions between organisms. A fitness function is specified and evolution is studied. Two papers deal with artificial worlds and simulate the population biology of inhabitants. Grammars, linguistics, graph theory and logic are used in various papers as means of structuring environments and studying the development ofartificial life. lrsystems of logic, graphs and delays were used to produce a set of color plates as part of the paper by Lindenmayer and Prusinkiewicz. A slightly stylized T-4 virus is shown assembling itself in the simulations of a movable finite automaton (MFA) by Goel and Thompson . Their paper includes wonderfully detailed diagrams of the MFA interfering with its host in order to have itself replicated. Stewart W. Wilson outlines four steps for experiments on the life cycles of his simulated organisms. Stuart Hameroff et al. present a model of microtubules as intracellular molecular automata, complete with scanning tunneling microscope images of microtubules in vivo. Hence the two papers on nanotechnology demonstrate that art and technology may yet again mimic life. A very stimulating paper by Moravec is filled with information on the behavior, senses, communications and intelligence of creatures in the air, on land and in the seas. He includes charts of the...

pdf

Share