Archetypes in Man-Computer Problem Solving

Information systems applied to operational environments have meaning only in what they do for humans performing tasks, whether clerical, technical or managerial. Each person's job-position entails interaction with a limited set of categories of variable data. By 'limited' is meant less than several thousand, and more likely several hundred, categories. A category set associated with a collection of tasks performed by an individual or an organization may be called a category domain. This concept makes possible a practicable (in size) data base responsive to support human tasks in human (psychological) time. An analysis of human problem-solving tasks reveals the following gamut: simple inquiry and update, status inquiry, briefing, exception detection, diagnosis, planning/choosing, evaluating/optimizing, constructing (designing), and discovery. There is no compulsive orderinig of these on a complexity scale. The information processing structure of each is exarnined: some common denominators among this set reveal five underlying archetypes of interaction. By making these archetypes explicit and consistent with concepts of domain, application disciplines and system design can move in parallel and generate a simple, well-defined language structure betweefn system and human user.