Two views are important, among others, at M.I.T. and Harvard. The classic M.I.T. view looks upon the AI Lab as a microcosm for examining how architects think. Underlying this view is the position of 'lets examine the way architects think about design and build tools which can reflect that process'. Another point of view, as expressed at Harvard, is speculative on what architects seem to do in design practice and education, rather than speculative on the nature of thinking per se. Both views seem ultimately to be concerned with representing architectural design knowledge within computers. And in the rob of computers as a design medium. This paper examines how the M.I.T. view and the Harvard view have superficially been associated with separate research directions. As these contrasting points of view incorporate many common themes. The author finds that it may be possible to take an eclectic position in teaching computer aided design.