Implications for Design by Paul Dourish (CHI 2006 paper)

Central points:

• The requirements that ethnography provide an "implications for design" creates a relationship in which knowledge is produced for the benefit of the designers and administrators -- those who wish to create some goal state in a system that happens to have people ("users") in it. Instead, Suchman proposes ethnography as providing an account of the partnership in which producers and consumers co-constitute one another, in part through technological artifacts.
  
• A key part of ethnography often lost in the way ethnography gets thrown about as a design methodology is the self-consciousness of the observer as a part of the social relations richly described. The observer is not a positivist instrument for transmitting impressions of the observed. (However, a lot of SCOT could be construed as including this subjectivity in non-human observation instruments' construction as well.)
• Ethnography can have empirical outcomes, which inform others of the simple facts observed, as framed and understood by the observer, but we should not overlook the analytical outcome -- those that, when revealed, call into question or qualify assumptions tacit in the design, the specification, the technologist-consumer relationship...

I read this article because I was writing an email to PD and I wanted to make sure I didn't misuse ethnography and ethnomethodology when I really meant interview. I got more than I bargained for, but that's good.