Divining a Digital Future: Mess and Mythology in Ubiquitous Computing
Divining a Digital Future: Mess and Mythology in Ubiquitous Computing
by Paul Dourish and Genevieve Bell
The MIT Press, Cambridge, MA, 2011
264 pp., illus. 4 b/w. Trade, $32.00
Reviewed by John Vines
School of Design
Northumbria University, UK
Designers are often preoccupied with the idea that technology can be designed to become seamlessly incorporated into everyday experience. At one point, designers felt it was useful to start measuring the physical dimensions of human beings so that cars and office chair were easier to use. Later on, especially when computers began populating workplaces, some designers attempted to design based upon rather basic computational models of the mind. Over the last two decades, technology designers have become more interested in the social aspects of technology use. It is now very typical to find academics from anthropological and social science backgrounds working in teams with engineers, scientists, and designers working within the major technology organisations and Universities. The co-authors of this book have observed and experienced the bringing together of the social and computer science research communities from both disciplinary perspectives. Dourish, currently Professor of Informatics at UC Irvine, trained as a Computer Scientist and has been a part of research groups at a number of the world’s leading centres for computing research. Bell is a trained anthropologist who has worked at Intel for over a decade, studying the appropriation of new technologies in different cultural contexts. Despite coming from rather different disciplinary roots, both authors find themselves situated within the same frame of argument—that designers, engineers, and scientists within computing-related disciplines would do well to pay more attention to the societal and cultural phenomena surrounding the use of technology.
In this book, Dourish and Bell bring their perspectives on HCI together on the specific subject of Ubiquitous Computing. For the uninitiated, Ubiquitous Computing – or ‘UbiComp’ as it is often referred to by HCI experts (and I’ll refer to it in this way here) – is an area of research that, broadly speaking, develops and studies technologies that permeate beyond the traditional confines of personal computers. As a paradigm of research, the origins of Ubiquitous Computing is typically associated with the work of Mark Weiser in the late 1980s, and in particular an article published in Scientific American where Weiser described his vision for ‘the computer for the 21st century.’  Weiser pictured a future where computers were omnipresent in peoples’ experiences but at the same time disappeared into the background: ‘They weave themselves into the fabric of everyday life until they are indistinguishable from it’. In the article, Weiser invoked that this would be an engineering challenge, one that requires lengthy collaboration with other disciplines; social scientists, ecological psychologists, artists and designers were all referred to in his writing. Dourish and Bell start off their contextualisation of UbiComp with Weiser’s article, not necessarily as it is the first instantiation of the idea of computing ‘everywhere’ but due to its formidable impact on computing research ever since. In the years following Weiser’s article, research groups developing ubiquitous technologies and pervasive computing systems started to emerge in key (mostly American) Universities, and large conferences such as UbiComp and Pervasive  became important venues to disseminate the latest knowledge on the subject. The ubiquity of Weiser’s article lives on in the vast corpus of peer-reviewed papers archived from these conferences that directly cite his original article. Similarly, key ideas that Weiser introduced—small wearable tags, mobile pad-like screen devices, and large tabletop and wall-based interactive screens—are still central focal points of the UbiComp research community.
Dourish and Bell explain the above context in chapters one and two as, from their perspective, it is the legacy of Weiser’s work and more specifically his Scientific American article that has supported the mythology of UbiComp. In chapter two they go on to argue this mythology is founded on the idea of a proximal future where technology will permeate everyday environments and, as Weiser argued, calmly enter the backdrop of our experiences. It seems, within the very Anglo-American centred world of UbiComp research at least, that Weiser’s vision is still not quite being achieved—technologies do not integrate well with one another, computer systems require a lot of human effort to keep maintained, and technological breakdowns are a frequent occurrence. Dourish and Bell find this assertion, which prevails in the UbiComp literature, flawed. Firstly, they highlight how in some regions, such as the Republic of Korea and Singapore, the UbiComp vision is already established and part of everyday life (although those experiencing it would doubtfully recognise the similarities between their own use of technology and those predicted by Weiser in 1991). Secondly, and more importantly from the authors perspective, they argue that many cultures experience UbiComp already—it is just that this experience is distinctly messy, effortful, and chaotic—and against the imagined future that Weiser presented.
In chapters three and four, Dourish and Bell set out their particular approach to understanding the messiness of UbiComp as it is practiced and experienced within certain cultures and settings. Chapter three provides an introductory overview of anthropology and ethnography and the distinctions between ‘the social’ and ‘the cultural’ as they are conceptualised in these disciplines. In chapter four, they discuss in more depth the relationship between ethnography, its methods of study, and the study of computing technologies. As I noted above, the use of social science methods in HCI is far from being unusual (Dourish in particular being a key early member of this part of the community) but, the authors argue, ethnography is typically used primarily as a feed for the eventual outcome of HCI and UbiComp research—the design of a new system. Whilst it is not unusual for a systems design team to include an ethnographer of some description, the outcome of any field study is typically a list of guidelines or heuristics for engineers and designers (who are completely disconnected from the studied context) to implement. The work of ethnographers and anthropologists working within HCI and UbiComp, therefore, tends to be very weakly translated to the scientists and engineers working on the same team. This appears to highlight the great problem of much technology research that claims to be interdisciplinary—that is, at the end of it, the ‘designers’ (who typically tend to be engineers of some description) are the gatekeepers at the end of the project.
It is clear that part of the problem that Dourish and Bell have with UbiComp research is that although it makes use of social scientific methods, often it is done with a complete lack of sensitivity to the subjects of interest for social scientists. They argue that: ‘First, it constructs ethnography as a point of mediation between, on the one hand, a domain of everyday practice and, on the other, a domain of technological design. Second, it implies that people will encounter technology as something just as it was designed and, hence, is appropriated or incorporated into practice’ (p.73). The way in which social science is used in technology design focuses on studying ‘the field’, finding gaps to be filled by technology, and then designing in response to the ethnographic account. Problematically, the very principle of using social science in design in this manner is paradoxical. ‘Seeking to close the gap through the application of ethnographic methods is a contradiction in terms; the gap is where all the interesting stuff happens, as a natural consequence of human experience’ (p.73). The suggestion implicit here is that the subject of study for the social scientist is rather different to that of the design team. Not only is this a significant difference, but also the very idea of using social science contextual work in design to alter a culture is, in itself, counter to the ethnographers cause. Rather, Dourish and Bell argue that ethnographic studies should be used to generate accounts of everyday uses of technologies but not directly influence the design of new systems. This is not to say, however, that ethnographic enquiries are not useful for designers; it’s that ‘the most useful strategy when engaging with ethnographic work is to “read for theory” as much as for empirical evidence, since in the end these may be where the truly significant implications lie’ (p.75). Dourish and Bell argue that whilst UbiComp researchers are continually striving for the seamless integration of technology into being, people from diverse cultures are already enacting technology into being in a ‘ubiquitous’ manner—it is just the process of enaction is messy and in many case unremarkably mundane. Ethnographic studies are useful here in theorising this mess, and by understanding technology use in highly diverse contexts, it is possible to find patterns that allude to the ways people make technology coherent over time.
Chapters five through eight form the second part of the book, where Dourish and Bell deploy their ‘messy’ approach by tackling influential concepts from within the Ubicomp research agenda. In these chapters Dourish and Bell continually juxtapose seemingly disparate ethnographic works with one another in order to bring out tensions in the overriding themes—a process of ‘defamilairisation’. Chapter 5, ‘What lies beneath’, begins by unpacking the concept of space as it relates to UbiComp. Dourish and Bell examine space from the perspective of ethnographic infrastructures . Infrastructure, here, does not refer to technical infrastructures but rather experience of a multitude of space and how this is influenced by a number of infrastructures, such as ‘naming, movement, interaction ... [that] emerge from and are sustained by the embodied practices of people who populate and inhabit the spaces in question’ (p.108). These infrastructures, therefore, are both the shape of and shaped by these embodied, culturally-contextualised, experiences. Chapters 6 through 8 provide similar re-conceptualisations of ‘Mobility and Urbanism’, ‘Privacy’ and the prevailing stereotypes of domesticity in UbiComp research.
Whilst chapters 5 to 8 read smoothly, they do feel somewhat drawn out – particularly to someone who is already familiar with the author’s past work – and each chapter really acts as a reiteration of Bell and Dourish’s claim that ethnography is at its best at unravelling the messy and effortful use of technology rather than food for designers to work with. These chapters are primarily an extended lead-in to the third part of the book, which contains just the one concluding chapter, where a ‘reimagining’ of the next quarter of a century of Ubiquitous Computing research is outlined. They begin by reaffirming their argument that studies of social and cultural meaning making cannot be used to directly inform design to close the gap between technology and practice. By their argument, design cannot act as anything other than a consequence of social action—new technologies are shaped by the encompassing cultural milieu and therefore become incorporated into them, or appropriated in a manner so that they do. As social scientists working within an area that essentially designs new technologies, however, Dourish and Bell appear to be forced to relate the ‘implications’ of their book back towards design in one form or another. Design, rather than responding to studies of culture by trying to smooth over cultural cracks, should instead act as embodiments of sociotechnical theories and arguments. By embodying these arguments, newly designed technologies can act as theoretical lenses to uncover new aspects of the world that were previously hidden. This is not to argue that they support technological determinism—far from it, by embodying cultural conceptions of space, mobility, privacy and the home, designers might move beyond the technical rationality of current systems. By understanding the relationship between social science and design/computer science in this manner, they proclaim the possibility of a future ‘socio-technical practice’.
There is much to be taken from Dourish and Bell’s book. They provide a wonderful summary of the past 20 years of research on the subject of UbiComp with a distinctly ethnographic lens, and add considerably to scholarship on the fringes of the sciences and humanities that focuses attention on collective human agency in the production and consumption of technologies. For designers who want to be more critical of their practice, they emphasise the fallacy of predicting how users will react, appropriate and use new designs and technologies. For those interested in disciplinary intersections, they make us aware of the problems of negotiating ontological and epistemological boundaries. They make us ask: How can knowledge from one domain be both translated and applied in other disciplines without losing meaning or, as Dourish and Bell argue, contradict the basis of the knowledge itself? They also make researchers reflect on the dominance and ‘myth’ associated with key texts, articles, and works, and they do so in a way that is at once in-praise and damning of Weiser’s original work. We should not burn down the archive and forget it ever existed, but we should not be frightened of shaking at the foundations of a disciplines research enterprise.
There is a strange circularity and contradiction to the way in which Dourish and Bell bring their argument to a close however. As social scientists, they understand technology as a social construction. At the same time, in forming their socio-technical hybrid practice, they argue that technologies reveal new dimensions of the world and allow for the reimagining of how people can act in the world. By their argument design is both a passive and active constituent of cultural meaning making—but they offer little to understanding moments where it is one, the other, neither or both. There are also a number of times when I found myself confused at sudden dichotomisation of the social or cultural with the physical or natural – dichotomies that are understandable when we stick strictly to a particular disciplines heritage (social science) but felt at odds with the idea of working in an inter- or trans-disciplinary manner.
I feel there is a more basic issue with the book, however—not to do with philosophical dichotomies or self-contained contradiction. Rather, it is that I worry this will be another example of literature from outside of the computer science and engineering literature that will have very little impact on the actual practices of technology designers. As someone already familiar with the authors’ work and the use of social science in design, there were very few moments where I paused for thought out of confusion. Those with little to no prior understanding of the language of social work, however, may find some sections more inaccessible. Perhaps this is not an issue—perhaps the future of a socio-technical hybrid design practice that Dourish and Bell outline must by a necessity only to be practiced by a small pocket of researchers who have the space to critically reflect on their actions. There is already a community of researchers within studies of human-computer interaction who are fully accepting of ethnographical studies of computer use and appropriation. The case presented in the book, therefore, would be preaching to the already well doctrined. As I read the book, I felt as though Dourish and Bell wanted to impact beyond this well-trodden area—to communicate their reimagined future of ubiquitous computing to a new generation of human-computer interaction scholars, computer scientists and engineers. Weiser’s articles provided the conditions for mythology by cutting across disciplinary boundaries—even if in a rather uncritical and crude manner—perhaps, in a way, Bell and Dourish need to let go of theirs just a little bit more.
1. Weiser, M. 1991. The computer for the 21st century. Scientific American, September 1991, 94-104.
2. www.ubicomp.org and www.pervasiveconference.org
3. Star, S. L.. 1999. The ethnography of infrastructure. American Behavioral Scientist, 43 (3), 377-391.