You do not currently have permission to view the full text for this document. Please sign in to an account with an active subscription or contact support.
Since Freud's descriptions of primary and secondary process, psychoanalysts have recognized that the logic by which ideas are processed is at least as important as their content. Various forms of logic have been described, starting in the Middle Ages. However, largely motivated by problems in computer science, logicians have only recently explored processes in ways that suggest models of human information processing of interest to psychoanalysts. Massive parallel processing and neural network models seem particularly promising. In this series of abstracts I examine some books and papers that address other models which may be useful to the psychoanalyst.
Darwinian natural selection can be viewed as a means for solving problems. Using the tools of genetic transmission, mutation, genetic recombination and selection, biological systems evolve to become sufficiently fit in solving the problem of surviving in particular environments. If we create a situation in which fitness is not measured by natural physical survival but by the ability to solve some set of problems rapidly and accurately, we have a method for computer programs to evolve by the same fundamental process as natural selection. First described in detail in John Koza's massive Genetic Programming: On the Programming of Computers by Means of Natural Selection (MIT Press, 1992), genetic or evolutionary programming has rapidly become a major field in computer science. It deserves the attention of psychoanalysts because it presents yet another, quite distinctive, model of how problem solving can develop. Mitchell's volume is a brief, clear introduction to the field. It requires essentially no formal mathematical background, but the reader must proceed carefully to understand each step in the discussion. The McDonnel volume will bring the reader up to the cuttin