Friday, June 16, 2006

Ethnography in HCI -- Comments on Dourish CHI paper

In his CHI2006 paper “Implications for Design”, Paul Dourish gives a portrait of ethnography that helps me (and hopefully others) to understand its role in HCI. I have for a long time been looking for a kind of research in HCI that would provide me (and others) with substantial and solid ways of thinking about interactive systems and how they are appropriated and used. I would like such research to be focused on creating challenging new theoretical constructs that could help us to see invisible structures and processes that influence the complex interactions between people and technology. My constant disappointment with many attempts is that when I read so called "ethnographic" studies, they are usually interesting and worth reading as long as they do the "scenic fieldwork" (concept from Dourish paper), but they completely let me down at the end, since they don't move to what Dourish calls the "analytical" level. There is no theoretical or conceptual outcome that inspires my thinking that challenges me. There is, in other words, no learning on a general level. I usually get this, almost bodily reaction (and not a nice one) when I read these papers and I come to the conclusion (or "implication for deisgn") and I realize that the outcomes are highly intuitive, everyday common sense, not surprising, not challenging, and I react with "didn't we already know that".

I have usually no problems with the way designers "use" ethnography (or its simplified versions), I agree with Dourish that they should be more informed about what they actually do and label their work appropriately. I do have problems, however, with the way it is done as a way to confirm an already planned and developed design idea (as is almost always the case in HCI). Coming from a deep understanding of design (as presented in my book "The Design Way"), interpretation and measurement of the existing reality is in a design process something very different from interpretation and measurement of the existing reality with a research intention. Design demands immersion into the full complexity of reality and from that full immersion you have to come out (within a limited time frame) with an understanding of something not-yet-existing that transcends the existing. This is to me not the same thing as being involved in deep ethnography with the purpose to come out of "immersion" with a deep understanding of the existing. These two goals are very different and I suspect that is one reason why ethnographers don't like the notion of "implications for design", since it is the wrong kind of outcome. They do create deep and insightful understandings of what-exists. This outcome is always important in design, but it does not tell the designer what to design. The creation of the not-yet-existing demands something more than a deep understanding. Ethnographers should therefore be allowed to do their work, i.e. create deep understandings not for design, but for designers. Knowledge that will change the designer’s way of approaching the specific design situation with its demands, needs and desires. The field of HCI needs really good and insightful ethnographic work, without demanding results easy to apply in design.

So, I really hope that Dourish paper will help the field in recognizing (1) the need for "ethnography" in a form that is intended, aimed, and designed to support in a design process (i.e. design oriented ethnography). I think this is not something we are good at today. And the need for (2) ethnographic research in our field that can create a deep understanding of the intricate relationships between people and technology, and that gives us new theoretical and conceptual "tools" to think about these relationships, not focused on a specific designs but creating a solid understanding that every good designer should be knowledgeable about.

No comments: