Practice and the social turn: the object of design is a practice, not a user

This seventh of ten units takes the social turn, which replaces the individual user and their task with shared practices and meanings. It argues that ethnography, probes and critical methods become design resources if that holds, but that the field has not shown they improve a design. The singular user becomes a convenient fiction the field keeps defending.

The move: the site of the social is practice, not the mind

Reckwitz (2002) sets out practice theory as an alternative to culturalist mentalism, textualism and intersubjectivism, traced to Bourdieu, Giddens, Taylor and the late Foucault. Shove, Pantzar and Watson (2012) give the form HCI later imports. A practice is an integration of materials, competences and meanings that exists in the moment of doing. Practices recruit or lose carriers as their elements travel between them. Two design consequences follow. The material a designer shapes is an element of a practice, rather than a screen in front of a user. A practice changes when it gains or loses carriers, not when an individual is persuaded. Neither claim predicts what one person will do next Tuesday, which is the price of the move.

The turn arrives in HCI and changes what a method is for

Kuutti and Bannon (2014) name the arrival, contrasting an emerging practice perspective with the prevailing interaction perspective. They set out the tenets of the newer frame, but argue that it complements existing HCI theories rather than replacing them. HCI's study of emerging practices can sharpen the practice approach in return. If the practice lens sits alongside the interaction lens, the user and the task are not abolished; they become one element among several.

Dourish (2006) attacks the convention that an ethnographic paper must close with implications for design. His claims are that the focus misconstrues the enterprise, and that it misses where ethnography helps HCI. Boehner, Vertesi, Sengers and Dourish (2007) show what happens when a method migrates without that background. Tracing how cultural probes were adopted and adapted, they find divergent uses and derivatives. They read that as disagreement about valid interpretation, and about how methods relate to methodology.

Probes and ambiguity refuse to average the field

Gaver, Boucher, Pennington and Walker (2004) describe probes sent to twenty volunteer households and the fragmentary returns they produced. They argue that the approach is being rationalised, and list four losses that follow. Unambiguous questions return what the researcher already knows, summarising produces an average picture that fits nobody well, analyses blunt the designer's contact with the material, and the demand for justifiable accounts constrains engagement. The study behind the argument sent packages to 20 volunteer households recruited through London periodicals and notices in newsagents' windows, a route that assumes a household at home in English, over a month. Each held a disposable camera relabelled with requests that were open-ended or absurd. Among them were something you would like to get rid of, the spiritual centre of your home, and something red. The pack also held a friends-and-family map printed over a cricket pitch and Dante's heaven and hell. The returns, the authors write, were incomplete, unclear and biased, and impossible to analyse cleanly. Gaver, Beaver and Benford (2003) move the same stance from method to artefact, treating ambiguity as a resource for close engagement rather than a defect. They sort it into ambiguity of information, of context and of relationship.

Reflection and affect: the values are inside the artefact

Sengers, Boehner, David and Kaye (2005) give the programme that follows from treating meanings as part of the system. Reflective design combines analysis of the ways technologies reflect and perpetuate unconscious cultural assumptions with building and evaluating devices that embody alternatives. Their two cases differ in audience. One added a layer to a museum visit so that visitors could contribute to an exhibit, not only receive it. It found that some visitors felt no right to author part of it. The other built a small desktop object through which couples in long-distance relationships signalled that they were thinking of one another. The authors piloted it with 5 couples. Sengers, Boehner, Mateas and Gay (2007) press the same case for emotional experience. Much work on affect, they argue, disenchants it by defining, categorising and controlling it, and so reduces its richness to the categories a computer can hold. Their Affector study yields strategies for enchantment, and the claim that enchantment must be approached obliquely. A space for it can be created, but not forced. The object of design becomes an interpretation a person can revise.

The user as a convenient fiction

Bardzell (2010) argues that feminism is a natural ally of interaction design. She proposes six qualities that a feminist interaction design carries: pluralism, participation, advocacy, ecology, embodiment and self-disclosure. None is unique to feminism, she says; it is their appearance together, in critical mass, that characterises the paradigm. Bardzell and Bardzell (2011) move from qualities to method. They assess the epistemologies, methodologies and methods of feminist social science against HCI's work on social change, and propose a feminist HCI methodology that pursues scientific and moral agendas together. Adamu (2023) asks whose work the field reads. She reports that issues of race and racism figure little in narratives about appropriation and localisation, and that a frame centred on race can be dismissive of other forms of recognition. The user is a placeholder for a person whose position and membership the definite article erases.

Visions of the future are claims about practice

Bell and Dourish (2007) observe that ubiquitous computing is driven by the possibilities of a future rather than the problems of a past. Its founding vision is more than a decade old, and the everyday outcome has been messier than it allowed. Drawing on cross-cultural studies of technology adoption, they argue for a ubicomp of the present that takes the messiness of everyday life as a central theme. Rogers (2006) makes the matching move from inside the field. She finds that context awareness, ambient intelligence and monitoring fall short of Weiser's world, because the specifics of context are more subtle and idiosyncratic than theories of context suggested. She proposes a shift from proactive computing to proactive people.

Contested question: does the turn yield design knowledge, or only criticism?

The first position holds that the practice perspective is a usable research agenda. Kuutti and Bannon (2014) present it as a lens. It organises issues emerging in recent work, directs research programmes and gives something back to the practice approach. What would settle it is a body of results in which work directed by the practice lens reached conclusions a task-centred study could not have reached. The comparison would have to be visible rather than asserted.

The second position holds that the turn's value is analytic, and that the demand for design outcomes is the error. Dourish (2006) makes the evaluative criterion the target, and Gaver and colleagues (2004) treat uncertainty as the point of the method rather than a defect in it. On this reading a good study makes a setting intelligible and changes what the designer notices. What would settle it is a study recognised on those terms rather than on its implications.

The third position holds that a descriptive practice account is not enough, because values and the choice of what to study are already inside it. Bardzell (2010) and Bardzell and Bardzell (2011) require the commitments to be named and embedded in method. Adamu (2023) shows that the field's reading canon decides which practices are studied. What would settle it is a methodology that travels with the method, so that a study is accountable for the conception of knowledge that generated it.

Boehner and colleagues (2007) document the technique travelling widely while the methodology did not. Gaver and colleagues (2004) accept that probes may be put to many ends. The test is not another argument. It is a case in which a method and its conception of knowledge travelled together, in two unrelated settings, and produced a decision the field's default study would not have produced. The literature records the migration and the disagreement; it does not yet record the comparison.

The evidence is strong in one place and thin in another. The design side rests on an agenda paper and three position papers, none of which reports a built system evaluated against a task-centred alternative. The best evidence in the unit points the other way, since Boehner and colleagues (2007) analyse how a method travelled while the methodology did not. While the dispute is open, a designer who uses probes, ethnography or critical methods should carry the conception of knowledge that produced them. The study should write down in advance what it expects of the setting. The strongest objection is to the unit's own claim that these methods are design resources. A record of migration shows they can be used, not that they improved a design.

Where the argument is heading

Dourish (2006) doubted that his was the last paper in the field to be forced to end with implications for design. He hoped it might be possible to stop. Bell and Dourish (2007) push towards a ubicomp of the present. Sengers and colleagues (2007) push towards enchantment approached obliquely, so the outcome is a space rather than a feature. Adamu (2023) presses on whose practice is read, and on whether the canon is itself part of the design. A practice account earns its standing when it can say what it expects of a setting before the study, and report what the design did when that expectation was tested.

The sources this unit rests on

• Adamu, M. (2023). What Are You Not Reading? Critical Race Theory for HCI. interactions 30(3). doi:10.1145/3589938 Reflects on what critical race theory offers HCI and on whose work the field reads. It reports that issues of race and racism figure little in narratives about appropriation and localisation, and warns that a frame centred on race can be dismissive of other forms of recognition.
• Bardzell, S. (2010). Feminist HCI: taking stock and outlining an agenda for design. CHI '10. doi:10.1145/1753326.1753521 Surveys feminism in HCI and proposes six feminist interaction design qualities: pluralism, participation, advocacy, ecology, embodiment and self-disclosure. It argues that they characterise the paradigm when they appear together, making values available as design criteria.
• Bardzell, S. and Bardzell, J. (2011). Towards a feminist HCI methodology: social science, feminism, and HCI. CHI '11. doi:10.1145/1978942.1979041 Assesses the epistemologies, methodologies and methods of feminist social science against HCI's engagement with social change, finding that scientific and moral agendas can be pursued together. It proposes an outline of a feminist HCI methodology.
• Bell, G. and Dourish, P. (2007). Yesterday's tomorrows: notes on ubiquitous computing's dominant vision. Personal and Ubiquitous Computing 11(2). doi:10.1007/s00779-006-0071-x Argues that ubiquitous computing is driven by the possibilities of a future rather than by the problems of a past, and that the everyday outcome is messier than the vision allowed. It argues for a ubicomp of the present centred on the messiness of everyday life.
• Boehner, K., Vertesi, J., Sengers, P. and Dourish, P. (2007). How HCI interprets the probes. CHI '07. doi:10.1145/1240624.1240789 Traces how cultural probes were adopted and adapted across HCI, finding divergent uses and derivatives. The pattern reveals disagreement about valid interpretation and about the relation between methods and the methodology behind them.
• Dourish, P. (2006). Implications for design. CHI '06. doi:10.1145/1124772.1124855 Attacks the convention that ethnographic papers must close with implications for design, working through the marginalisation of theory, power relations, a restricted model of technology and practice, and representation. It argues that the analytic work is where the achievement lies, and that the criterion misses where ethnography helps HCI.
• Gaver, W. W., Beaver, J. and Benford, S. (2003). Ambiguity as a resource for design. CHI '03. doi:10.1145/642611.642653 Presents ambiguity not as a defect but as a resource that encourages close personal engagement, sorted into ambiguity of information, of context and of relationship. It moves the practice stance from method to artefact.
• Gaver, W. W., Boucher, A., Pennington, S. and Walker, B. (2004). Cultural probes and the value of uncertainty. interactions 11(5). doi:10.1145/1015530.1015555 Reports a probes study with twenty volunteer households and the fragmentary returns it produced. It argues that the approach is being rationalised, lists four losses that follow, and names the wider stance probology.
• Kuutti, K. and Bannon, L. J. (2014). The turn to practice in HCI: towards a research agenda. CHI '14. doi:10.1145/2556288.2557111 Names an emerging practice perspective and contrasts it with the prevailing interaction perspective, setting out its tenets. It argues that the frame complements rather than replaces existing HCI theories.
• Reckwitz, A. (2002). Toward a Theory of Social Practices. European Journal of Social Theory 5(2). doi:10.1177/13684310222225432 Sets out practice theory as an alternative to culturalist mentalism, textualism and intersubjectivism, traced to Bourdieu, Giddens, Taylor and the late Foucault. It shows that the vocabularies differ in where they locate the social and how they treat body, mind, things and knowledge.
• Rogers, Y. (2006). Moving on from Weiser's Vision of Calm Computing: Engaging UbiComp Experiences. UbiComp 2006, LNCS 4206. doi:10.1007/11853565_24 Finds context awareness, ambient intelligence and monitoring falling short of Weiser's calm computing, because the specifics of context are more subtle and idiosyncratic than theories of context suggested. It proposes a shift from proactive computing to proactive people.
• Sengers, P., Boehner, K., David, S. and Kaye, J. (2005). Reflective design. Critical Computing '05. doi:10.1145/1094562.1094569 Proposes reflection on the unconscious values embedded in computing and in the practices it supports as a core principle of technology design. It combines that analysis with building and evaluating devices that embody alternatives, illustrated through two case studies.
• Sengers, P., Boehner, K., Mateas, M. and Gay, G. (2007). The disenchantment of affect. Personal and Ubiquitous Computing 12(5). doi:10.1007/s00779-007-0161-4 Argues that much work on affect disenchants experience by defining, categorising and controlling it. The Affector case study yields strategies for enchantment and the claim that enchantment must be approached obliquely, since a space for it can be created but not forced.
• Shove, E., Pantzar, M. and Watson, M. (2012). The Dynamics of Social Practice: Everyday Life and How it Changes. SAGE. doi:10.4135/9781446250655 Defines a practice as an integration of materials, competences and meanings, held together only in the moment of doing. It analyses how practices recruit and lose carriers and how elements travel between them.