ABSTRACT
One of the most important issues to be considered in system design is how best to convey the functional attributes and possibilities of the system to the user via the human-computer interface. Current thinking suggests that a particularly powerful technique is the use of metaphors for real world, and therefore familiar, objects and activities to represent system properties (Rogers, Leiser, & Carr, 1988; Carroll, Mack, & Kellogg, 1988, Hutchins, 1989; Wozny, 1989). Most of these papers (and those discussed by them) are concerned solely with the selection of appropriate real world domains to be used as representations of a particular system. The conceptual structure proposed in this paper to describe and explain the function of representation at the human-computer interface, suggests that there are in fact two equally important questions to be answered when designing a system interface. This paper will attempt to clarify the differences between these two questions and will explore the implications of the second, and little studied, problem.
