as tracing the development of complexbiological forms such as flowers. Computer-graphics programs now make useof L-systems to yield realistic three-dimensional images of plants.The significance of Lindenmeyer's contribution is evident in the factthat so-called "genetic algorithms" are now basic to research into a-lifeas well as many other areas of interest. Genetic algorithms, firstdescribed by computer scientist John Holland of the University of Michiganin the 1970s, are comparable to L-systems. A computer worker trying toanswer some question about a-life sets up a system--an algorithm--by whichthe computer itself rapidly grades the multiple possible answers that ithas produced to the question. The most successful of the solutions are thenused to develop new software that yields further solutions, and the cycleis repeated through several "generations" of answers.Evolutionary ModelingLangdon himself picked up on the work of von Neumann by attempting todesign an "a-life" form on a computer screen. In 1979 he finally succeededin developing loop-shaped objects that actually reproduced themselves, overand over again. As new generations spread outward from the initial"organisms" they left "dead" generations inside the expanding parameter.Langdon noted that the "behavior" of these a-life forms genuinely mimickedreal-life processes of mutation and evolution. He eventually proposed thata-life studies could provide keys to understanding the logical form of anyliving systems, known or unknown.One of the most striking a-life simulations of evolutionary processeshas been the work of Thomas Ray of the University of Delaware, who in 1990set in motion a "world" of computer programs that he called Tierra. Theworld started out with a single ancestor, a program containing 80instructions. A-life evolution proceeded as mutations rapidly appeared. Thenew forms included "parasites" that interacted with the original hostforms, producing further mutations of h...
