Originally posted: Sun, 02 May 2004 02:47:55
Full disclosure: Activity-Centered Design is the second book in the Acting with Technology series at MIT Press. Mine was the first. The series examines how human activity is mediated through tools and technologies, drawing on frameworks such as activity theory. Like Tracing Genres through Organizations, Activity-Centered Design combines activity theory with other perspectives in order to investigate and design information technologies.
The text is exactly 100 pages, making it short enough to read in one sitting. And Gay and Hembrooke's writing style makes it quite easy to read. Illustrated by a series of studies by the authors at Cornell's HCI Group, the book draws on CMC analysis and configural analysis, as well as insights from the Social Construction of Technology (SCOT), to develop and apply a theoretical framework. Through their studies of information technologies in museums and at Cornell, the authors demonstrate the framework. In so doing, they critique much user-centered design work as too focused on individuals and their tasks, not enough on organizational development. Particularly impressive here is their use of Engestrom's cycle of development.
I'll limit my critique to the issue that has been vexing me these many months. Gay and Hembrooke discuss activity systems as decomposable into a network of activity systems: "the original setting and increasingly broader contexts." They provide a cite that I'll have to look at. But I find this formulation to be suspicious. When they say "decomposable," I envision sort of a pinata that, once struck, yields its contents. But in practice, activities aren't contained but overlapping. Take the university classroom, in which the activity is mediated by many artifacts and practices -- books, chalkboards, chairs, lectures -- each of which is connected with other activities. Sometimes wildly different ones, as Gay and Hembrooke's studies demonstrate: students in computer classrooms found themselves using instant messaging and checking their websites in class, for instance. The activities are interlinked, the artifacts are polysemous, but not in ways that are decomposable or containable.Along these lines, I should note that Gay and Hembrooke attempt to provide activity theory with something that it's long missed: a thoroughgoing account of stakeholders. AT grew up in the Soviet Union, where questioning the Soviet orthodoxy could land you in the gulag, so it understandably tended to focus on harmonious craft activities with little conflict or on overarching confrontations (contradictions) between Marxism and capitalism (see Leont'ev's books for examples of both). Gay and Hembrooke do a nice job of bringing in tools that help to conceptualize, explore, and develop stakeholder perspectives. But so much more needs to be done.