Publication 7560

An evolving theory known as “constructivism” challenges the traditional view of how we generate and revise knowledge. Constructivism helps address a major issue raised by modern scholars of the history and philosophy of science, and decision theory. The question is: How do we reduce the search and solution space of complex and changing environments to “mind size” (i.e., to fit our limited memory and computational capacity)? One emerging answer is that we rely heavily upon robust presuppositions and simplified representations of environmental structure. However, such constructed knowledge is likely to be highly fallible, relying as it must on impoverished data bases in the service of strong expectations or paradigms. In this paper we address two issues: Under what conditions can knowledge be highly fallible and at the same time be highly functional?; Can we make a plausible case, within this constructivist frame of reference, for realism, for knowledge that approximates “reality”?