Expert Systems and Knowledge-Based Engineering (1984-1991): Implications for Instructional Systems Research

Expert system technologies are varieties of artificial intelligence (AI) approaches in which decision-making knowledge is codified and modeled. This design case has the challenging task of characterizing this set of technologies during a particularly important period in its development (1984-1991), with an emphasis on a particular system that was used in food production environments by Campbell Soup. It analyzes the social and research impacts of early, pioneering information elicitation and processing strategies that focused on the distillation of the knowledge or know-how of individuals construed as experts in particular arenas, approaches broadly labeled as “knowledge-based engineering” (KBE). Widely-publicized notions of “thinking machines” and “canned experts” provided motivation for a good deal of early expert systems development (Feigenbaum & McCorduck, 1986), with accusations of “hype” often levied (Blair, 2002). This article historically situates these technological strategies in the period from 1984 though 1991, then links them with current instructional systems approaches that more fully involve collaborative elements as well as contextual perspectives. The motivation for this article is to explore how larger technological and social trends and assumptions can influence particular research efforts, especially in the richly interdisciplinary area of information systems. The article also explores the circumstances and consequences of “failures” of system development, with expert systems providing widely-discussed exemplars (Gill, 1995; Oravec & Plant, 1992). This article is rooted in the assumption that historically-informed perspectives can provide some underpinnings to the building of humane and sustainable research projects, particularly in areas that have human subjects and volatile contexts as essential elements. This article also addresses the continuing legacy of university curricula and business training initiatives that were shaped to accommodate expert system and KBE approaches in past decades. Discourse about human expertise generated by expert system efforts in 1984 through 1991 holds insights for current research and development, as well as signals potential sources of dysfunction of, and opposition to, future instructional system initiatives.