The Case for Rules in Reasoning

rules ore not involved in human reasoning, or at best ploy a minor role. Other views hold that the use of abstract rules is o core aspect of human reasoning. We propose eight criteria for determining whether or not people use abstract rules in reasoning, ond examine evidence relevant to each criterion for several rule systems. We argue thot there is substantial evidence that several different inferential rules, including modus ponens, contractual rules, cousai rules, and the law of large numbers, ore used in solving everyday problems. We discuss the implications for various theoretical positions ond consider hybrid mechanisms that combine aspects of instance and rule models. One of the oldest views about the nature of thought is that reasoning is guided by abstract rules of inference. This view has its origins in Plato’s theories of reasoning and education, and was the rationale behind “formal discipline” approaches to education ranging from the medieval scholastics’ teaching of the syllogism to the English “public school” curriculum of Latin and mathematics. In modern times, abstract inferential rules have played important roles in some of the most influential theories of cognition, including those of Newell and Simon (e.g., 1972) and Piaget and Inhelder (e.g., 1958). This blue-blood intellectual history notwithstanding, the role of abstract rules has recently come under attack from a variety of sources. Part of the attack stems from the development of alternatives to rulebased models of thought. One class of alternatives is instance models, which assume that solving a problem involves the retrieval of specific instances The research reported here was supported by a grant from the National Science Foundation (BNS-8705444) to Edward E. Smith and by grants from the NationaI Science Foundation (BNS-8709892 and SES-8507342). the Sloan and Russell Sage Foundations, and the Army Research Institute (MDA 903-89-C-022) to Richard E. Nisbett. We are indebted to Keith Holyoak, Michael Morris, Steven Pinker, Zenon Pylyshyn, Lance Rips, Steven Sloman. an anonymous reviewer, members of the Wednesday night seminars at the University of Michigan. and especially Douglas Medin for critiques of earlier versions of this article. Correspondence and requests for reprints should be sent to Edward E. Smith, Department of Psychology, University of Michigan. 330 Packard Rd., Ann Arbor. MI 48104.