Ask an anthropologist

Bonnie Nardi may not always feel your pain, but she’s willing to listen to you talk about it.

As a research scientist with Agilent Technologies in Palo Alto, Calif., Nardi is among the few IT professionals who bring an anthropological perspective to managing user expectations. This application

of ethnography — the description of races of men — is rare, but Nardi believes it should reshape the way companies approach product development.

Nardi is also the author of “”A Small Matter of Programming: Perspectives on End User Computing,”” and “”Context and Consciousness: Activity Theory and Human-Computer Interaction.”” She spoke recently with Computing Canada about how she has applied her research in the private sector.



CC: What’s an anthropologist doing in IT?

BN: When I was in graduate school, I actually did an unconventional dissertation — for an anthropologist, at least — where I modelled an old problem in social organizations using computer simulation models. I really liked doing that. I like combining the formal methods you get out of computer science with these really sticky human problems you get from anthropology. But I was a little ahead of my time. This was in the late 1970s, and I had to do a very different kind of thing to get a job in anthropology but I retained my interest in technology.

Anthropology has not really picked up technology as a big focus. It’s been really heavily influenced by the whole post-modern thing. Things like race, class and gender really kind of took over what anthropology was about. I figured I wasn’t really going to be able to pursue my interest in technology within the confines of conventional anthropology. So I left and did various things and got back into research by going to research lab at Hewlett-Packard, and since then I’ve worked in various industrial settings in research labs. So it’s sort of a circuitous route, I guess.

CC: Why do you think anthropology has ignored the sort of problems you’ve been focusing on?

BN: (laughs) I wish I knew the answer to that. A lot of it is there’s a real political agenda for anthropology and it’s fairly far towards the left of the spectrum. It’s a fairly small field, also; you can’t have all that many different conversations going on. There is a small subset of people interested in technology but it’s a real small minority.

CC: What about the flip side? How do technology companies see the discipline of anthropology as helpful to their business?

BN: Companies are always looking for new angles on customers, and I think that’s why I’ve managed to survive in the industry. Anthropologists have this bag of tricks — it’s a little bit different than marketing, it’s a little bit different than psychology — and it basically consists of going out and having very deep conversations, in-depth interviews and doing a lot of observing of customers and what they’re up to. And then trying to analyze and distill all that information and distill it in a meaningful way. And that’s something anthropologists are better at, I think, than anyone else.

CC: Can you give me an example of a research project you’ve worked on recently?

BN: Let me tell you about one I worked on at AT&T. I’ve only been at Agilent for a year and a half, so I haven’t done anything that has come to any great fruition yet. But I was at AT&T for about four-and-a-half years, and I worked on a project where I was interested in how people communicate across organizational boundaries. The contrast here is that you’re working within a small, familiar, well-known team. And that tends to become the dominant image that designers hold in their minds.

As I started looking around, I realized that yes, people are spending some of their time in teams, but in fact they’re reaching out far beyond their teams. Talking to customers, consultants, contracts, associates in other companies, government agencies, journalists, venture capitalists. There’s all these other entities that people actually have to interact and work with. That became the problem I became really interested in.

What I learned was people spend a lot of time managing their contacts. Not just putting the names in their address book, but actually getting to know people, keeping up with where people are — and because there’s so much flux in the world today, people are always moving around. Their job functions change, the kinds of things they know change. So, where are my contacts and what are they doing? If I have a question that I want to ask somebody, if I need information, or if I’m recruiting labour, what does my world of contacts look like? That was the problem we wanted to model and help people solve.

We built this system called ContactMap which basically models the individual’s personal social network and really completely changes the computer desktop. So instead of seeing files and folders and hierarchies and lists and all the machine-centric entities, you see a representation of your own social network. It’s very visual; you can even have little pictures of all the people in your network. And then behind every node — that is, every person or group in your network — you have all their contact information and you can contact them in a much more automatic way than you can by dialing the phone or actually having to type their address into an e-mail message. You also have a repository of the documents that have flowed between you and the other person. So I would have everything I’ve sent you to and you’ve sent to me. That’s how people remember their documents. It doesn’t mean the underlying file system has gone away — you can still certainly file that way — but you have this other representation that seemed very promising to us. And since I’ve been at AT&T, they’ve done some experiments and have shown that there is some value in having this representation of people’s communication and information worlds.

CC: We hear a lot about the stuff that doesn’t work well — the software that’s not intuitive, the stuff that’s not human friendly. Is there anything out there you see as a great model for something that does work?

BN: I guess I have three models in my mind about how we get really good products. Spreadsheets and the Internet are my two favourite examples where somebody had a great idea and it just happened to be something that everyone in the world wanted, too. That person said, ‘Okay, I can implement this myself,’ in the case of the Internet, or ‘I can work with a programmer to implement this idea I have that I know is a great idea,’ which was the case with spreadsheets. So you had basically an end user who had a good idea and somehow manages to get it implemented, and the other people can use it.

At the end of the continuum someone you have like Steve Wozniak, who just had this incredible brainstorms. I have no idea where they come from, but they just dream up these amazing things. And then you have someone like me in the middle who tries to mediate between people who can do the implementing and end-users who can’t articulate exactly what they need. The spreadsheet people knew, ‘Okay, I’ve been using this paper ledger thing; I should just put that on the computer and I’ll be done.’ But then you have people like the ones that we talk to that are working across the organizational boundaries and they don’t have a concept for a product, but they can articulate their problems. Then you try to put someone in the middle of that process, and I can extract what I think might work in a software design. So I think we have three avenues of getting great products.


Share on LinkedIn Share with Google+
More Articles