While browsing and working on a recent paper, I mused on the missed opportunity of interaction design. Reading Terry Winograd’s (1997) From Computing Machinery to Interaction Design, I was stunned to see how visionary this was, in the context of contemporary HCI thinking which focused on interactions with computer screen interfaces (still, sadly, the main focus of much HCI work). Winograd saw computing as a “social and commercial enterprise” and saw the role of interaction design as situating technology within social and commercial processes. This thinking is related to Suchman’s (1987) Plans and Situated Actions: The Problem of Human-machine Communication, which saw human-computer-interaction as part of a stream of activity, located in the rationale of a wider sequence of tasks. While HCI theorists were fixated on task-analysis and screen-interface design, Suchman argued that we should see tasks as related to what had gone before and what was to follow. Winograd argued that we should design technical artifacts to be useful in the larger context of social networks and the complexities of interactive spaces.
I was reminded of this when reading a discussion of Don Norman’s (2005) Human-Centered Design Considered Harmful. In this essay, Norman argues that HCI designers focus on “human-centered design,” which he relates to support for tasks and artifact-interactions, when they should focus on “activity-centered design,” related to the larger context of what people do. While I agree wholeheartedly with the sentiment (and applaud the fact that the idea will at last get an audience if Don Norman has taken it up), the concept of activity-centered design still misses the point that we need to understand how actors perceive their stream-of-reality, situated within both a social and a cognitive-processual context, for interaction design to fulfill its potential.
In my 2003 paper, Human-Centered vs. User-Centered Approaches To Information System Design, I argued that human-centered design is not the same as user-centered design. User-centered design sees the human-being as a consumer of technology, whose reality is – somehow, magically – represented by the set of functions accessed via the computer artifact. This tends to be the focus of “traditional” HCI research. Human-centered design, on the other hand, sees the human-being as an autonomous individual, who may want to perform tasks in a different way, or a different order, to other computer “users.” They see the logic of what they do – and therefore the manner of its execution – as part of a socially-situated stream of activity that is meaningful to their understanding of work-processes and not some engineer’s idea of “best practice.” This means that design methods need to deal explicitly with problem inquiry, rather than just focusing on problem closure.
In a new paper (hopefully to be accepted soon!), I have argued that situated interaction-design needs an analysis of two dimensions of the work that people do:
- the formal vs. informal translations that need to take place, to locate work practice in both the social (unstructured-interaction) and organizational (structured-interaction) worlds, and
- the global vs. local translations that need to take place to locate work practice in both the situated and generically subjective worlds.
Most of our design methods focus only on one quartile of this reality: the formal, structured world of data-processing. To really support interaction design, both education and practice need to take on a much wider scope.