A techno-utopia dismantled: Critical Review of “Ubicomp’s Colonial Impulse”

Dourish and Mainwaring’s paper claims that the research and development agenda of Ubicomp (ubiquitous computing) is essentially a colonial enterprise. In short, the aim of Ubicomp is to integrate computers into every aspect of human life (cp. Figure 1), thus making them essentially invisible, much like written language today is ubiquitous and at the same time invisible as it is taken for granted (Weiser 1991).

As the authors argue, this rhetoric is problematic for a number of reasons. First, like colonial nation-states, it conceives the flow of goods (read: knowledge and technological innovation) from the centre (the research laboratory) to the periphery (the rest of the world), thus establishing a disempowering relationship of deficiency. This redefines the world reductively by what it is lacking. Second, Ubicomp pushes standardisation and uniformity. By removing information’s situatedness in locality, specificity, and plurality it uses patriarchal mechanisms of hegemony and homogenisation. It positions dominant views as universal, marginalises alternatives, and erases difference. Third, the most important universalising mechanism is quantification. Like colonial nation-states (or modern-day bureaucracy), Ubicomp uses statistical analysis to manage everyday life at scale and thus reducing it to those aspects that are computable. Fourth, Ubicomp has a techno-deterministic view of the future. The laboratory acts as a microcosm where future techno-life is conceived and later flows to the less advanced periphery. By this it privileges the researcher’s perspective and equalises it with the world of the “users”. Furthermore, it shifts focus away from understanding current conditions and their political and historical circumstances.

These conditions pose an ethical problem by denying agency to alternative voices. This partial and narrow view is also a pragmatic problem as it is not recognised or reflected upon. Finally, it is a conceptual problem, as technological innovations should not just be ends in themselves.

To go forward, the authors propose to avoid the rhetoric of centre and periphery, as is e.g. Participatory Design by questioning the traditional design/use relationships. In particular, they propose to ban the word “user” and evaluate ethnographic work not just for its “implications for design”. Furthermore, a feminist techno-science agenda should be adopted to recognise historic specificities of sites of technology design and use, aim for coordinated autonomy instead of centralised control, and embrace the presence of multiple perspectives without privileging any.

I really enjoyed reading this article as it articulated to the point the concerns I have with many techno-sciences. While its focus on Ubicomp requires some background knowledge from the reader, the criticism is – as the authors state themselves – not unique to Ubicomp. The idea of “solutionism”, to fix social problems with technology is ubiquitous 😉 in HCI and other engineering sciences. This is not to abandon thinking of possible (desirable) futures in science. In fact, I always appreciated HCI for being not just largely descriptive like most social sciences but incorporating a prescriptive design dimension. However, these two aspects should go together and the latter should be informed by the former. The paper illustrates convincingly how we as researchers and designers can approach this.

Related Article

The following paper is a case study of a Ubicomp application:  Kasteren, T. Van, & Noulas, A. (2008). Accurate activity recognition in a home setting. In UbiComp (pp. 1–9).

It describes a sensor system to detect various activities in households, namely Idle, Leaving, Toileting, Showering, Sleeping, Breakfast, Dinner, and Drink. It illustrates very clearly Dourish and Mainwaring’s criticism. (1) it conceives the flow of innovation from the centre (the experimental home setup (see Figure 2) to ‘standard homes’ lacking these features; (2) It generalises from an individual’s home to any home thus homogenising different living; (3) it quantifies human activities to sensor-detectable movements, ignoring individually relevant activities (the list of activities above exemplifies this mechanistic view of everyday life dramatically) (4) The experimental setup is more concerned with future techno-living than with present everyday practices of living.

Further References

Dourish, P., & Mainwaring, S. D. (2012). Ubicomp’s Colonial Impulse. In UbiComp ’12 (pp. 133–142).

Vines, J., Clarke, R., & Wright, P. (2013). Configuring participation: on how we involve people in design. In CHI ’13 (pp. 429–438).

Weiser, M. (1991). The computer for the 21st century. Scientific American, 265(3), 94–104.