«Abstract. The paper explores what exactly it is that users participate in when being involved in participatory design (PD), relating this discussion to ...»
634) The adults’ monitoring of the design workshops and the ways the children’s design moves were evaluated created a school-like situation, with the corresponding roles and power distribution. The adults had the role of ‘actors who know best’, acting in the interest of the children. Morrow and Richards contend that ‘the biggest ethical challenge for researchers working with children is the disparities in power and status between adults and children' (1996, p. 98). Thomas and O’Kane (1998) demonstrate how it is possible to be open to children’s agenda, have them voice their own concerns or questions, use participatory research techniques to give them control over the process, and also have them participate in the interpretation and analysis of research data. Sisom did not exploit all of these possibilities of sharing power with the participating children.
The Sisom project is also an example of power in the Foucault sense of ‘normalizing’ practices. Doing things in the right way was defined from the beginning as complying with medical evidence, as in the list of symptoms that the children were invited to translate. Sisom involved listening to how children talk about how they feel but with a standardized catalogue of symptoms in the background. In IPCity the possibility of ‘manipulating’ participants’ perception of a site, which can be powerful, is another example of Foucault’s ‘normalizing’.
Both, the designer and the urban team had the ‘power to’ select the representations of an urban site the workshop participants worked with, choosing and editing panoramic views of the site, and providing content.
‘Locking’ choices in an artifact, by making them in some sense material can be considered an act of power. Analysis of decision-making in IPCity revealed several instances illustrating that after a choice has been made, power comes to reside in the person who implements a decision, making it material. By being ‘chosen’ the artifact constrains in the sense of ruling out some previous alternative choices. An example is the power of editing the visual content that was shared by the visual artist and the urban planners but in the end the artist, as the person who performed the editing, took the decisions on what would become visible and how. In a similar way her editing of the panoramas that she produced according to the architects’ instructions became ‘critical’ as it shaped how participants would actually see the site. (Bratteteig and Wagner 2014, p. 83).
The four projects also give some additional insight into the ‘power of making’, which rests on ‘designerly skills’: the ability to develop a prototypal realization of some choices that has technical ‘functionalities’ to work with. In his studies of engineering work Bucciarelli has coined the term ‘object worlds’, which are ‘worlds of technical specializations, with their own dialects, systems of symbols, metaphors and models, instruments and craft sensitivities’ (1988, p. 162).
Technical expertise is a particularly strong form of power/knowledge as it is difficult to share with non-technical users. However, looking at decision-making in the four projects allows for a more nuanced analysis of the ‘power of making’.
We have identified several factors that may strengthen (but also constrain) it: the irreversibility of some choices when they have been made material; complexity resulting from the interdependencies of decisions (a point we will examine in 5.3);
the vulnerability of some technical choices; the multidisciplinarity of a project team and the designer’s dependence on users’ ‘seeing’.
Some design moves cannot be reversed or only at a great cost, once they have been concretized in a prototype. In IPCity the choice of color tracking offered the possibility to freely position tokens on a map of an urban site, which the project team thought essential. The project’s strong focus on materials and haptic engagement had blurred the visibility of other solutions, such as moving from a constellation of physical table and color tracking to a multi-touch version of the prototype. Once these choices had been made, too many steps would have been required to ‘undo’ them, starting from scratch with another solution.
The ‘power of making’ can also be enormously vulnerable: things may not work and a solution may not be ready at hand. In IPCity the incompleteness of the early prototypes and the vulnerability of some technical solutions (e.g. the tracking mechanism, the difficulties of placing objects in a photographic panorama) created conflicts with the urban team, and also tested the patience of the participants in the early workshops (Maquil 2010). The Florence project can be seen as almost a counter-example of the power of designers, with the nurses’ solution challenging the limits of the machinery not suited for implementing this particular solution creating a hard technical programming challenge to realize the solution anyway.
Considering real use, preparing for it and observing may be indispensable for design work to proceed. IPCity would never have reached the participatory results it aimed at without these experiences. The designer team needed the support of additional ‘observers’: ethnographers, architects, for interpreting the observations and coming to the ‘right decisions’ for redesign. Finken (2005) argues that it is in the power of designers to interpret what users voice as their needs and preferences and translate these into a design result, using their power/knowledge. This is true to some extent: the technical choices that are based on ‘seeing’ a prototype in use are taken by the designers who evaluate their options. However, also users enact their ‘power to’, not necessarily through their opinionating and arguing, but through material ways of using a design or refusing to use it in the way that had been envisioned. Resisting or complying in use may carry more weight than words, as the examples of Florence and IPCity demonstrate (see e.g., Bjerknes and Bratteteig 1987). Even in Sisom the testing, although carried out in a simulated setting with healthy children, settled some important choices.
The sharing of power in a PD project is a highly complex, multi-facetted issue.
There are numerous ‘forces’ at work: the expert knowledge of some participants, which may not be easy to share, but also the potential vulnerability of this knowledge in case it does not result in a successful design or is not acknowledged by the users. Still, the ability to materialize choices, ‘locking’ them in an IT artifact, is a strong act of power. Users’ main ‘power to’ in a PD project is in their creating choices that would not have been possible without their contribution; and in their ‘seeing’ and evaluating, using or refusing to use a design in a particular way. We can also see that ‘power to’ and ‘power over’ may be interrelated, as most evidently in the Sisom project where all decisions were taken without the children’s participation. Going back to the distinction between power and influence made by Zündorf (1996) may be helpful here. While ‘power represents an intervention in the action space of others’ (p. 38), influence requires listening to the voices of the other participants so as to be able to convince them and to get them on board. In some of the situations we described power as well as influence were involved.
5.3 Choices influence each other In design (as in other types of work) many of the choices that are created and selected to take a step further, concretizing them, are inextricably intertwined with other choices. CSCW research is based on a conception of cooperative work as ‘observable relations of interdependence that are formed in response to practical exigencies but which then in turn require the development of a family of equally observable coordinative practices’ (Schmidt 2011, p. viii). When looking at decision-making in design it is also important to account for interdependencies that result from the fact that decisions are intertwined and affect each other, and how some decisions spur others. Interdependencies between decisions can reveal how some decisions frame the whole design process – and other decisions – and therefore have much larger effects than what may be visible at first sight.
We go back to the proposition to look at decision-making as ‘a complex network of issues involving a whole host of linkages, more or less tightly coupled’ (Langley 1995, p. 275). Identifying these linkages in a design project is not only interesting, as it helps understand how the web of interdependencies evolves. In the context of an analysis of power relations it foregrounds a ‘logic’ that mediates these relationships. So we move from who has had a say in the design decisions we described to the interdependencies their choices created; and we come to the conclusion that participation in some design decisions is more important than in others.
Langley et al. (1995) provide a language for talking about how decisions (or issues as they prefer to say) are linked, having developed a typology of linkages.
All the four projects illustrate linkages that are sequential (linked over time) or precursive (framing later decisions). For the sake of our argument in this paper we just give a few examples that illustrate our point.
Sequential linkages characterize decisions where one decision leads to a series of others, either smaller ones (‘nesting’) or larger ones (‘snowballing’) or simply the same decision recurring, when a problem is difficult to resolve. These types of linkages are characteristic of the sequences of ‘moves’ that are directed towards a design result. Sequential linkages are the most frequent ones in a design project.
For example, in IPCity the original decision to build a haptic interface was followed by numerous smaller decisions concerning the design of the tokens and the interaction mechanism. Many of these sequential linkages are not strong in the sense that they do not necessarily narrow the design space: the decisions that ‘follow’ create their own dynamic, opening up new choices and interdependencies, such as (in the case of IPCity) deciding the size and material of tokens or the kind of visual feedback (as part of the interaction design). Langley et al. call this type of linkage ‘nesting’: big decisions generate a nested series of smaller ones. Also the choice of photographic panoramas as representations of an urban site illustrates how some technical choices may have unforeseen implications that enforce a series of additional steps. It turned out necessary to model the panoramas so as to support placing objects correctly with attention to height and volume and to handle occlusion properly: decisions that created an enormous amount of unanticipated work when preparing the ColorTable for testing in the context of a real urban project (Maquil 2010).
‘Snowballing’ captures the fact that a couple of smaller decisions ‘snowball’ into a major one. An example from the Sisom project is when the children tried out the prototypes while in bed, saying that the game prototype did not look ‘serious enough’. Nonetheless the decision was to keep the game-like presentation, while a ‘dinosaur’ version of the system, which was also tested by the children was dismissed. This was a major design decision that built on a series
of small decisions about which kinds of game-like features to keep in the design:
while the possibility to create your own figure (avatar) to explore the islands was included, the children’s racing and shooting suggestions were excluded.
A decision on one issue can critically affect the premises for subsequent decisions on a variety of other issues, and in different ways: linkages may be enabling, evoking, preempting, cascading, merging, or due to learning (precursive linkages). The decision in the Sisom project to have islands as the main navigation mechanism turned out to be ‘enabling’. It made it possible to include different navigation mechanisms, one for each island. This opened up for different ways of reporting symptoms, hence more choices. In Desarte, the modular organization of ‘worlds’ on an urban grid was also ‘enabling’, opening up for numerous design possibilities and, ultimately, for users to design their own worlds.
In the Florence project the participatory designers’ decision to build the system suggested by the nurses ‘evoked’ new problems that required a number of difficult decisions. It almost destroyed the nurses’ trust in the system. The decision was interpreted very differently by the nurses and the designers: the nurses had already seen a prototype and could not understand why the system was not ready, while the designers struggled with technical design decisions in order to make the available software tools enable the project to achieve its goal.
In a similar manner, IPCity included several design decisions, each of which influenced other technical as well as non-technical decisions. For example, moving mixed-reality technologies outdoors made it necessary to have a tent, which had to be transportable to different urban sites and easy to set up; it also added to the instability of color tracking.
Precursive linkages can also be seen as the effect of learning. ‘Mutual learning’ is a key concept in PD (Bratteteig et al. 2012), hence we would expect many such linkages. Numerous decisions concerning the tracking algorithm and interaction design in IPCity were due to a learning effect through iterations of ‘see-move-see’.
The notion of decision linkages points to the fact that decisions can affect future decisions on other related issues: new problems may be evoked, alternative solutions precluded or a promising new choice enabled. Considering and eventually anticipating such interdependencies, is a major challenge for designers.
The concept of ‘design moves’ captures this explorative way of thinking and working; as does the concept of ‘placeholder’ (Wagner 2004), which refers to the need to keep open the possibility to ‘undo’ decisions if further moves indicate that they have unwanted effects. Decision linkages may be difficult to see while being involved in a project, as a particular choice may have unforeseen consequences.
This has to do with the ‘wicked’ nature of design problems. It is even more complicated: some decisions are more important than others since they enable many other choices or constrain the design space rather early. When looking at the most important decisions in the four projects, we see that users did not participate in all of them. Often these are decisions that frame the whole project, such as the strong vision that guided Sisom and IPCity. ‘Big’ decisions in a design project may also concern the choice of a technical framework, as was the case in Florence and IPCity. Also the setting up of the context of a project is a ‘big’ decision that delimits the design space in many ways.
Power issues in a PD project interact with a web of dependencies: some of these have to do with the institutional context of a project, while others emerge from the choices that are made.
6 Assessing the depth of participation?
We find that the concepts outlined in this paper help becoming aware of the different ways participation in design can happen and also how these pave the way for design results that are participatory. The concepts can be used for planning for more or better participation, and for a more thorough evaluation of the degree of participation. This can be achieved by articulating in more explicit ways how the different design moves are accomplished, addressing questions such as: how can the space of possible choices be widened; which of these choices are selected, which are not – and why; were the choices participants created respected as valid choices in the decision-making? Figure 10 provides an overview of the questions we have explored in this paper.
In Bratteteig and Wagner (2014) we concluded:
We have shown that not all decisions in a PD project have to be made in a participatory way.