«Abstract. The paper explores what exactly it is that users participate in when being involved in participatory design (PD), relating this discussion to ...»
First, while the concept of ‘decision’ itself (which we take to mean commitment to action) may imply distinct, identifiable choice, in fact many decisions cannot easily be pinned down, in time or in place. […] Third, even when a decision can be isolated, rarely can the process leading up to it. (p. 261) Hence, analyzing decision-making in design may require going back to the fieldwork material we have collected, in a ‘reflective attitude’ (Schütz, 1954) or an act of ‘reflection-on action’ (Schön, 1983, 1987), revisiting the arguments that have been brought forth by all involved participants. We may find out that some of the decisions that we identify in this way are not the result of explicit deliberation. We will show that, in design work some choices are made in the process of making.
The second complication has to do with the fact that in this process of revisiting some decisions will turn out as more important than others, as they are interlinked with other choices. In the process of designing ‘decisions typically become inextricably intertwined with other decisions’ (Langley et al. 1995, p.
261). The fact that decisions interact with each other may make identifying a decision and those involved in taking it even more difficult. Langley et al. propose to look at decision-making as ‘a complex network of issues involving a whole host of linkages, more or less tightly coupled’ (p. 275). They have identified different types of decision linkages: sequential, that link decisions over time;
precursive, that frame later decisions; and lateral, that share resources and context. The different types of decision linkages are useful for seeing how decisions are linked and intertwined and, ultimately understanding why some choices are more important than others. Our analysis shows that while some of the choices made in the four PD projects framed later decisions, some small choices ‘snowballed’ into a major one and that a choice prevented the designers from seeing other options.
To sum up, the decision-making framework we use in our analysis has several
• The notion of ‘design-moves’ captures the experimental, step-by-step character of design work, where evaluation and reflection of a move open up for new choices (while closing others);
• More explicit choices emerge through processes of deliberation – the (collaborative) imagining and ‘projecting’ into the future;
• The notion of decision-linkages captures how choices are interrelated.
3.2 Power and influence Looking at power (and influence) helps understand why decisions in a PD project may be taken in a more or less participatory way, as participation does not necessarily make power issues disappear. In the early days PD was about empowering workers. The power of different participating stakeholders was described in terms of an unequal distribution of organizational resources, knowledge and skills (e.g. Bjerknes and Bratteteig 1995). Articulating and challenging power relations formed an important part of the PD process. Joan Greenbaum (1996) describes this tradition of early PD that focused on political
struggle and empowering workers:
The early action-research projects in Scandinavia which were part of the Critical tradition realized that designing for workplace democracy was not something that could be done within the profession alone. Using political coalitions they fought for and won laws which gave workers the right to co-determination in decisions involving technology. They also realized that workers as users needed more and better training in order to participate in design. This gave rise to a second generation of Scandinavian research projects in the 1980s which took an analysis of labor-capital relations a step further, using labor process analysis to explicitly design for increasing skill […]. These projects were designed for specific groups of workers supporting their interests and including the workers tacit skills in design principles for future systems (p. 232).
While power relations remained a recurrent theme in the PD literature, little effort has been made to more systematically examine the different aspects of a concept that, as Crozier stated in his 1973 essay ‘The problem of power’ sometimes borders confusion. Our interest in using power as a conceptual tool has led us to revisit the vast literature about power, carefully selecting what we think is useful for understanding power issues in design (see Bratteteig and Wagner 2014). In her book ‘Wittgenstein and justice’ (1973) Pitkin distinguished between ‘power over’ and ‘power to’, stressing that these two uses of the word ‘power’ express rather different phenomena. ‘Power over’ another person (‘by his getting the other to do something, but also by his doing something to the other’) is a relational concept. ‘Power to’ denotes ‘capacity, potential, ability, or wherewithal’ (ibid, p. 276). Pansardi (2012) also examined this distinction adding that it ‘is nothing more than an analytical distinction between two aspects of a single concept of power, and, since they always occur together, an investigation of the former is always also an investigation of the latter’ (p. 86). We think that both positions are valid and helpful: the analytical distinction that helps identify power as the ability to contribute; while also keeping in mind the relational aspect of any kind of power. PD is about sharing power with users; hence both aspects need to be examined.
A related concept is empowerment, which ‘is frequently theorized as power to’ in contrast to ‘power as domination, largely characterized as power over’ (Haugaard 2012, p. 33). The discourse about participation (e.g. Gaventa and Cornwall 2006) traces the notion of empowerment back to the pedagogical work of Paolo Freire, who believed that community empowerment starts when people listen to each other, engage in a dialogue, identify their commonalities, and
develop solutions for their own problems:
In truth, the opposite of manipulation is learners’ critical, democratic act of participation in the act of knowing that they are also subjects. The opposite of manipulation, in brief, is people’s critical and creative participation in the process of reinventing their society […] (Freire and Macedo 1987, p. 43) ‘Power’ is only one way of looking at decision-making in PD. Pitkin also insists on the difference between ‘power’ and ‘influence’, which are often treated as (almost) the same (notably by Robert Dahl, 1957). These concepts ‘are not strictly comparable. They are of different kinds, or move in different dimensions’ (Pitkin 1973, p. 279). The difference between the two concepts has been described by
While power represents an intervention in the action space of others, influence has to begin with one’s interaction partners’ own dispositions to act: how they develop opinions and take decisions and not—as in the case of power—with pushing through what already has been decided. (p. 38) Other relevant concepts for understanding decision-making, such as ‘trust’ and ‘loyalty’. Trust includes the feeling that one can somehow rely upon others: it is the ‘confident expectation of the benign intention’ of others (Dunn 1990, p. 74).
Loyalty is ‘the feeling of confidence that trust between others […] can be maintained in the long run and therefore restored in the future if absent at any given time’ (Barbalet 1996, p. 80). More generally speaking, the sharing of power in PD is a complex interplay of mechanisms, in which different resources and multiple dependencies and loyalties come to work together. We have shown elsewhere how influence as a regulating mechanism is very common in decisions requiring highly specialized (mostly technical) expertise. A large number of decisions are based on trust: delegating power to people who have the expertise to solve a problem competently (Bratteteig and Wagner 2014).
In this paper we are particularly interested in the ‘power to’ of the different participants in a PD project: the agency and capacity to shape action, which partly depends on access to organizational resources, partly on ‘power/knowledge’ in the Foucauldian sense. Foucault (e.g. 1982) has argued that depersonalizing power is important if we want to understand the most effective forms of it. What makes power so effective are particular ‘technologies’ or regimes that construct knowledge, bodies and subjects. The term ‘power/knowledge’ refers to the power of defining issues, ‘normalizing’ them so that they can be recognized and resolved; translating them into a language that makes them amenable to particular interventions (at the expense of others).
Both, ‘power over’ and ‘power to’ come to play together in decision-making.
They help explain what makes participation possible and what its limitations are.
Some of the ‘power to’ in collaborative work is inscribed in the division of labor, as well as in other structural aspects, such as authority; control of scarce resources; use of organizational structures, rules, and regulations; or control of boundaries (Morgan 1986). Some ‘power to’ is ingrained in the ‘logic’ of a practice; some derives from expertise or power/knowledge.
4 Participation in what Studying the dynamics of decision-making in design is important for understanding what users actually participate in. For this purpose we introduce a set of distinctions that are based on Schön’s notion of ‘design moves’: between creating choices, selecting a choice, concretizing a choice, and ‘seeing’/ evaluating the result of a choice (see also Bratteteig and Wagner 2014). While we find these distinctions useful, we need to stress that they should not be interpreted as stages in a design process, although they indicate a certain sequence. They stand for different sets of practices, depending on the kind of design work we are looking at. For example, ‘creating choices’ may correspond to rather different practices in a PD project, from getting design ideas from an ethnographic study to involving participants in all kinds of creative-experimental exercises (Bratteteig et al. 2012); this also applies to the other types of activities we talk about.
How did we proceed? We chose the starting and end point of the timeline of each PD project, working our way back and forth. As it turned out, two of the projects (Sisom and IPCity) had started out with choices that proved to be ‘big’ decisions: decisions that strongly framed further choices. While some of these were about values and concepts: the visions (Bratteteig and Stolterman 1997), others were technological choices how to implement these visions. When reconstructing how some of these decisions had come about we had to extend the projects’ timeline well before their official start, recognizing that they were based upon and inspired by previous work that in both cases dated back several years.
Hence, our first round of analysis looked at the dynamics created by these ‘big’ decisions, trying to understand how they influenced the many other choices that came up in a project.
However, not all PD projects start out with a vision; the vision sometimes emerges slowly as part of many smaller choices. This is why we, in our second round of analysis, chose to look closer at the final design result, aiming to identify the choices that became part of it. In a PD project not all design ideas or choices that participants create will or can be pursued. Some of the ideas will never be represented and explored, for a variety of reasons: they may be discarded as too complicated to be implemented or as not fitting into the project vision, or they may not find the alliances that are necessary to ‘push them through’ (Bratteteig et al 2016). Here, questions of politics and power in a PD project come to the fore.
Identifying choices that are visible in the design result allowed us to work our way back, looking into some of the smaller decisions that can be understood as happening in a series of design moves.
As mentioned above, we chose projects we know rather intimately, since we were involved in them, in different roles. The main reason for this is that the kind of analysis we undertook is not possible without a deep knowledge of the project and the many details of the design process – including the things that participants do not want to talk about. This is a limitation of our method, which we discuss later. The empirical material we have worked with is not ethnographic in that the projects we were involved with were not systematically recorded by any outside observer. We have performed a secondary analysis of written project accounts based on empirical material (some of it ethnographic). Our reconstruction of the decision-making processes is mainly based on documents from each of the projects: reports, memos of different kinds, and video material (in the case of the Sisom project). These documents were complemented by personal memories and, in the Sisom project, interviews with the project leader. In all four cases we have relied on an extensive project documentation.
4.1 The projects In the Florence project (1983-87) designers, together with health care personnel, developed prototypes to support professional nursing practices. The prototypes were based on an extensive mutual learning period that included fieldwork enabling the designers to understand nursing practices, and training sessions enabling the nurses to understand technical possibilities. The Nursing system is a simple, flexible reporting system made for the presentation of patient information on screen and in print, giving overview as well as details. The prototype makes it simple to update and access patient information, and its reports are used as flexible work sheets for the nurses on duty (Bjerknes and Bratteteig 1988). The nurses continued to use the Nursing system and it later served as a requirement specification when the hospital invested in a new IT system for nurses.
The Sisom project (2005-06) aimed at developing a symptom registration system for ill children to be used before meeting with a medical doctor. The Sisom System was developed with health care professionals and healthy children representing the main stakeholder group: children with cancer. Two agedifferentiated groups of children from a nearby school participated in a series of workshops aiming to design the interaction of the system in ways that children understand and like, following Alison Druin’s method for participatory design with children (Druin, 2002). In parallel to designing the interface and interaction mechanisms the project team collected and evaluated symptoms reported in the medical literature. These evidence-based symptoms were checked with oncology experts and children. The system was first tested with the healthy children and only at a later stage with two children at the cancer ward. The Sisom System is still in use in the hospital (Ruland et al. 2008).
In the Desarte project (1999 - 2001) IT designers, together with architects and landscape architects, developed several prototypes in support of professional design practice. One of these design results, the 3D Wunderkammer, is based on fieldwork revealing the relevance of and need for inspirational objects throughout the design process. The 3D Wunderkammer is an inhabitable multi-media archive, collection support, and view generator. Users can place inspirational objects – images, sketches, 2D or 3D scans of samples and objects, sound, and video – in a metaphorical space of cities and landscapes. They can navigate in this space and explore it, search and collect material, and generate different modes of viewing it (Büscher et al. 1999). The participating architectural office used the 3D Wunderkammer for a little while but found it difficult to maintain due to lack of support. It inspired numerous publications but at some point became technically obsolete and too much work to revive.
The IPCity project (2006-10) aimed at designing a collaborative mixed reality application for supporting mixed teams of urban planners, politicians and citizens in using participatory technologies to create and manipulate design alternatives for
real urban planning projects. Two groups of users were involved in this project: