This article uses a variety of principles of cultural-historical activity theory to extend Herbert Simon's (1996) insight into the inherent linkage between the creation of artifacts and design. We argue that design research must grapple with the doubly artificial, as the classrooms in which many educational designs are implemented are themselves already artificial and contingent--the products of design--and the learning that is the focus of investigation is already an adaptation to the classroom environment and so artificial. Focusing our discussion on the mesogenetic character of the temporal characteristics of typical educational design-based intervention research, we present an example of an 18-year-long life span of an intervention that was initially expected to last 3 or 4 years. Crises late in the life of the system, 1 of which rescued the system, 1 of which terminated it, documented through field notes written by undergraduate participants, provide evidence for the dynamics of the system's internal functioning in relation to events occurring in the larger ecology of which it was a part.