We are developing an “action science” approach to human-computer interaction (HCI), seeking to better integrate activities directed at understanding with those directed at design. The approach leverages development practices of current HCI with methods and concepts to support a shift toward using broad and explicit design rationale to reify where we are in a design process, why we are there, and to guide reasoning about where we might go from there. We represent a designed artifact as the set of user scenarios supported by that artifact and more finely by causal schemas detailing the underlying psychological rationale. These schemas, called claims , unpack wherefores and whys of the scenarios. In this paper, we stand back from several empirical projects to clarify our commitments and practices.