Design under uncertainty: Engaging the context of use in the design of expert system

This dissertation considers the design of computational artifacts in organizational settings. I uncover the complexities of designing information technology in organizational contexts by drawing from fieldwork on the design of expert systems, although I offer generalizations to designing information technology. The unit of analysis is the process of constructing a computational artifact, rather than the artifact itself. In order to study process, I focused on the activities of designers in organizational contexts, and considered what resources (materials, tools, and methods) they required to accomplish their work. I expand the traditional view of the innovation--an expert system--by focusing on what resources were afforded to the designers adopting the expert system package of materials, tools, and methods, in order to construct these systems. I focus in detail on the developmental trajectories of three expert system projects. Through detailed explanations of events in development, I foreground how the evolving expert systems were projected into contexts of use, who constructed these projections, and what force these projections had on the artifact and the design process. I assume that the design of any technological system relies upon projections to fill the distances between the present in the context of design and the future in the context of use. I conclude by considering the essential needs for projection in the design process and what other contexts should be available in the design process for constructing projections.