« Smart City, a radio show about cities | Main | Mix your Feeds with FeedDigest »
July 29, 2005
Activity Theory and User-centered Design
Peter Merholz has a post today responding to Don Norman’s Human-centered Design Considered Harmful essay, in which “The Don” seems to rediscover “Activity Theory” as it can be applied to interaction design.
I’ll admit I wasn’t aware of Activity Theory and have only now just read the Wikipedia entry on it. To oversimplify, it seems to be saying that “users” are individuals, with knowledge and practice habits gotten from membership in a community (or communities) of practice. “Activities” are object-centered and frequently social. Wikipedia quotes Nardi’s definition of AT: “all the mental functioning including remembering, deciding, classifying, generalising, abstracting and so forth, as a product of our social interactions with other people and of our use of tools.”
Elsewhere, Nardi’s paper here includes a checklist (“Capturing the context of HCI”), with items such as “focus on the structure and dynamics of interaction—internal vs. external components of activity and support of their mutual transformations with the target technology.” So…what aspect of the human use of technology doesn’t fall into that statement? Is this just “it depends” on an academic scale?
These sorts of Grand Unified Theories of Everything, franky, scare me. To me, Activity Theory seems to almost defy practical application. One could read into it that pinning down any specifics about technology use, user behavior, or user context would in fact ruin the sweeping descriptive power of the theory. You can’t measure the speed of the quantum particle and also know its position, so to speak.
In looking for specific descriptions of Activity Theory applied to interaction design, I see Beyond the Interface: Encountering Objects in Use, a paper that teaches me only that the definition-of-the-discipline argument’s was going on in the 1980’s, too. That it references not a single specific example of a product developed using Activity Theory is dissapointing.
Activity Theory and System Design: A View from the Trenches also has virtually nothing about the specifics of designing an application. It’s got pages and pages on the process and results of extensive interviews with users. But the absence of even a simple prototype put in front of users makes it impossible to judge the effectiveness of the approach.
I wonder to what extent UCD’s and Activity Theory’s expectations of design inputs (deep examination of the formations of user goals, dissection of community and social network relationships) are really artifacts from an era in computing that’s passed. Specifically, it’s now easier than ever to iterate working versions of a product quickly, to stay in perpetual beta, and to build quickly stuff people really love to use, in other words, to “Get Real.”. These days, it seems easier—and perhaps safer—to evolve certain products through many small changes as you learn about user behavior. Perhaps this is the best way to “design for practice”: to be truly responsive to practice as it changes. Are we still basing our design practices on ideas largely dating from time when that sort of approach wasn’t possible?
Posted by Andrew at July 29, 2005 10:18 AM