Simulates Human Thought

This article is concerned with the psychology of human thinking. It sets forth a theory to explain how some humans try to solve some simple formal problems. The research from which the theory emerged is intimately related to the field of information processing and the construction of intelligent automata, and the theory is expressed in the form of a computer program. The rapid technical advances in the art of programming digital computers to do sophisticated tasks have made such a theory feasible. It is often argued that a careful line must be drawn between theattempt to accomplish with machines the same tasks that humans perform, and the attempt to simulate the processes humans actually use to accomplish these tasks. The program discussed in the report, GPS (General Problem Solver), maximally confuses the two approaches—with-mutual"!benefit. GPS has previously been described as an attempt to build a problemsolving program (Newell, Shaw, and Simon, 1959a, 1960a), and in our own research it remains a major vehicle for exploring the area of artificial intelligence. Simultaneously, variants of GPS provide simulations of human behavior (Newell and Simon, 1961a). It is this latter aspect—the use of GPS as a theory of human problem-solving—that we want to focus on exclusively here, with special attention to the relation between the theory and the data. As a context for the discussion that is to follow, let us make some brief comments on some history of psychology. At the beginning of this century the prevailing thesis in psychology was Associationism. It was an atomistic doctrine, which postulated a theory of hard little elements, either sensations or ideas, that became hooked or associated togetherwithout modifica-