EDITOR Gary Marsden [e-mail]

Electronic Tablecloths
and the Developing World
Gary Marsden

University of Cape Town| [e-mail]

Within the HCI community, many of us are working on how to create appropriate technologies for people living in developing countries. Some people are involved in deep ethnographies—studying and living with users in remote communities—while more technology-driven projects seek to establish appropriate digital infrastructure (for example, WiFi networks). Reporting on many of these projects has appeared in this forum over the past few years.

In doing this work, researchers can sometimes feel isolated or at odds with colleagues who are working on more mainstream problems. Certainly, mainstream-HCI techniques do not always make a smooth transition to a developing-world context. Pure participatory design is an example: How do you get someone to suggest a design for a computer system when they have never seen a computer? We get locked into thinking that we need special methods to deal with users from the developing world. But we lose sight of the fact that the users aren’t different; it’s the environment in which they reside.

This was brought home to me recently as I sat through Bill Gaver’s closing keynote at DIS 2008. For those of you not familiar with Gaver’s work (shame on you), he creates devices, such as the History Tablecloth, that explore new ways of interacting with digital technology. The History Tablecloth uses weight sensors to detect objects being placed on cloth’s surface. Over time the cloth will emit light around the objects, almost like a halo, the intensity of which increases with the length of time the object rests on the table.

Gaver’s main point was that devices were not designed to serve a particular goal or solve a problem, but instead offered new opportunities and resources to people. That meant it was up to the users to reason about what they were for or how they might fit into their lives. When users eventually found a use for the product, it was usually not what the designers had expected. This led Gaver to conclude he was unlikely to ever predict all the

possible uses of his designs—and that this was a good thing. For example, in the case of the History Tablecloth, users deliberately placed objects on the cloth to create interesting patterns, whereas the research team had anticipated that it would be used as a history mechanism, recording household interactions as a pattern on the cloth. The best one could do, he concluded, was to recognize that users’ stories were as valid as those of the researchers. So, in his current work he ensures that some aspect of the design is flexible, allowing users to impose their desires on it. For example, his Plane Tracker system uses aircraft transponders to display a Google Earth visualization of the route onto a screen in a user’s living room as the plane flies overhead. However, no geo-political data was layered on top of the visualization in an effort to see how the users would interact with the little data that was presented. Would they augment the experience by looking up the route in an atlas, or be content to see the trip as purely an aesthetic experience?

Coming back to users in the developing world, the issues are surprisingly similar. Here, we also have users being exposed to novel technology, which make it impossible for them to intuitively understand how that technology might fit into their lives. Many of the researchers I have talked to have stories of how their work failed because they had not properly understood the user’s needs and the social impact, economic impact, and so on, of the context and culture in which they were working. Indeed, I have many of these stories myself. However, the message I took from Gaver’s talk is that failure (in the sense of accurately predicting usage) is inevitable and must be built into the design process.

This is subtly different to the argument advocating the use of prototypes. Prototypes support “fast fail”—they let you get the bad ideas out of the way so you can get on with refining the good