Abstract: Card and Newell, in an influential 1985 paper, described programming languages as 'obviously symmetrical' between programmer and computer, and called for balanced investment on the programmer and computer sides of the design space. But the design space is in fact more complex than that, with important impacts of purpose, as well as of programmer and computer, on effectiveness. Further, each part of the space is fragmented into many distinct areas, reflecting consequential differences in people, purposes and computational setting. Empirical methods face challenges in spaces of this complexity. As suggested by research on the role of mechanisms in scientific thought, cognitive dimensions analysis is better suited to operate in this complex space than are empirical methods, and should be promoted and extended.
PPIG 2017 - 28th Annual Workshop
Methods in user oriented design of programming languages