When system developers design a computer system (or other information artifact), they must inevitably make judgments as to how to abstract the worksystem and how to represent this abstraction in their designs. In the past, such abstractions have been based either on a traditional philosophy of cognition or cognitive psychology or on intuitive, spontaneous philosophies. A number of recent developments in distributed cognition (Hutchins, 1995), activity theory (Nardi, 1996), and experientialism (Lakoff, 1987) have raised questions about the legitimacy of such philosophies. In this article, we discuss from where the abstractions come that designers employ and how such abstractions are related to the concepts that the users of these systems have. In particular, we use the theory of experientialism or experiential cognition as the foundation for our analysis. (David Benyon and Manuel Imaz)