Dissonance and diplomacy: coordination of conflicting values in urban co-design

ABSTRACT This article seeks to contribute to the ongoing discussion on values in co-design processes, by introducing concepts from the emerging field of valuation studies. Using the work of David Stark and Ignacio Farías as an entry point to this perspective, it shows how co-design can be understood as a collective process of finding negotiated settlements among conflicting accounts of value, through practices of coordination. This idea is illustrated by a case in which co-design is mobilised as a tool for developing and governing ‘active frontages’ in a regenerating district in Gothenburg, Sweden. The article shows how the valuation studies perspective relates to, and in part differs from, other approaches to collaborative and participatory design. While sharing some of the intuitions of both agonism- and actor-network theory-informed approaches, its front-staging of practices and principles of valuation does nevertheless provide an alternative perspective on co-design. The valuation approach depicts co-design processes as a negotiation-based search for settlements, which suspends rather than solves value conflicts. Thus, co-design may be construed as a form of diplomacy, which operates within certain political limits of designerly peacemaking.


Introduction
In recent years, the notion of values has been given more attention in the discussion on co-design. In this very journal, several interventions have highlighted different aspects of value in co-design processes: The values of users in co-design (Halloran et al. 2009); how values can act as virtues that guide the co-design process (Sejer Iversen, Halskov, and Leong 2012); collaborative value-creation (Mose Biskjaer et al. 2019); conflicting values among participants in co-design (Le Bail, Baker, and Détienne 2020; Agid and Chin 2019). Moreover, a special issue has been published on understanding, capturing, and assessing value in co-design (Whitham et al. 2019), mainly focusing on conflicting accounts of the value of social design (Kimbell and Julier 2019). Along with this intraco-design discussion, there also seems to be a broader interest in values in design, for instance in relation to Value Sensitive Design (Friedman and Henry 2019).
This article will extrapolate from this emerging literature and construe co-design as a process in which 'conflicting accounts' of value are exposed to each other, yielding a form of 'dissonance' (Kimbell and Julier 2019, 13), from which a common design may emerge. divergent objectives of the municipality, development companies, and future user groups. As such, it serves as an example of co-design among a heterogenous set of participants, ripe with dissonances and frictions, stemming from different criteria and practices of valuation. It also shows how co-design may be mobilised in the shift from government to governance in urban development (Swyngedouw 2005). As such, the relevance of the case stretches beyond the specifics of the active frontage design, as well as beyond the Gothenburg context.
The argument is based upon 2 years of ethnographically inspired fieldwork, conducted between early 2017 and late 2018, encompassing approximately 50 hours of participant observation of urban design practitioners engaged in the regeneration of Masthuggskajen. Thus, the research is based on observations from formal meetings, workshops and seminars, and informal discussions in different settings. The argument also draws upon 23 thematically open interviews with public officials, politicians, property developers, architects and urban design consultants conducted during the same period. Furthermore, the article also draws on official documents, internal documents, meeting notes and PowerPoint presentations dating from different periods of the regeneration process.
The data gathered was coded in an empirically driven fashion, inspired by Aspers (2004). The first order constructs were coded using NVivo, while the second order constructs were elaborated upon through printed tables from the software. In the case of the latter, particular attention was placed on values and valorisations expressed in the data. The researcher obtained access to the site by being invited by the municipality to study the dynamics of the process. This implies that the researcher was primarily observing the actions of the participants, rather than participating in (or facilitating) the co-design process. Upon the completion of the research, the findings and an early version of the analysis was presented in an internal report. This has been vetted and validated by two practitioners -one person centrally positioned within the process, and one more peripherally located person with good insight into it.

Co-design as the coordination of dissonance
The proposition that co-design is a process of reaching settlements through negotiationsagain, as suggested by recent work on conflicting values in co-design (Le Bail, Baker, and Détienne 2020; Agid and Chin 2019) -chimes with some of the other approaches listed in the previous section. Indeed, it is useful to place this proposition in a historical context, which it shares with ANT. The 'realist' idea that design can be understood as a stabilisation of a conflicted situation can be traced back to Machiavelli (von Busch and Palmås 2016a, 286). John Pocock (1975) suggests that Machiavelli's The Prince -often read as a manual for attaining and maintaining power -actually centres on the problem of innovation. For Machiavelli, innovation is that which the ruler needs to deploy as a means to restore order to a turbulent world. This is especially the case in specific moments when 'working customs, traditions, and routines' (Marres 2005, 142) are dissolving. In those moments, the political act of creating stabilising innovations is called for.
This approach can be traced in Actor-Network Theory (ANT). Indeed, Latour (1988) explicitly referenced Machiavelli as a starting point, framing the ANT project as one of 'rewriting The Prince' for a technological age, in which social order is maintained in part by technological artefacts (Latour 1988, 275). So, ANT can be understood as the Machiavellian approach to innovation, in which the successful solution 'stabilises an acceptable arrangement between the human actors (users, negotiators, repairers) and the non-human actors (electrons, tubes, batteries) at the same time' (Akrich, Callon, and Latour 2002, 210, italics added).
In ANT, innovation and design is described by the notion of translation. As applied in the context of urban co-design by Palmås and von Busch (2015), this concept captures how actors with disparate aims align their projects so as to fit into a common, composite goal. Action comes from this stabilisation of networks of actors. However, in this 'translation' of goals, there is always 'a drift, a slippage, a displacement' (Latour 1999, 88) -all the actors in the networks see their original aims and interests being modified. Indeed, in ANT, interests are merely 'temporarily stabilised outcomes of previous processes of enrolment' (Callon and Law 1982, 662). In sum, the translation model highlights how co-design is a matter of translation and stabilisation of interests, played out in sites where Machiavellian micropower games are rife (Binder and Brandt 2008;Andersen et al. 2015;Palmås and von Busch 2015;Stuedahl and Smørdal 2015;Pedersen 2016;Rice 2018;Rørbaek Olesen, Holdgaard, and Laursen 2020).
In valuation studies, we find an alternative model of the emergence of the new -be it an innovation or a design. There are similarities with ANT -both are accounts of innovation as an accomplishment of negotiation, and both are interested in 'the Machiavellian moment' in which the new is to be stabilised and consolidated. However, there are also differences. Stark's (2009) argument starts from an account of his 1983 fieldwork in a Hungarian factory, which operated in the planned economy during standard hours, while producing for individual firms during off-hours and weekends. Here, Stark's attention was drawn to 'the clash of contending principles of evaluation' (xiii) that came to the fore in this situation. The things valued highly in the planned economy, did not carry the same weight in the deregulated 'intra-enterprise partnerships', and vice versa. Given the confused dissonance of competing principles of valuation, how does coordination of work tasks emerge?
Valuation studies scholars such as Stark are thus interested in how actors negotiate situations in which there is an uncertainty regarding 'what counts, what matters, what is of true relevance' (Stark 2009;Marres and Stark 2020). Dissonance, then, is what emerges when different, out-of-sync valuation practices meet. On the one hand, this phenomenon involves friction and conflict, on the other, it creates the impetus for the search for the new. Farías (2015) has applied these concepts to architectural design processes, in which dissonances are omnipresent. Farías describes design projects in terms of dissonant valuations and knowledge claims, brought forward by architects, specialists (such as light designers and engineers), and clients. Dissonances may also emerge from material artefacts, such as conflicting models, renderings, sketches and plans. This dissonant reality -the uncertainty regarding what constitutes a 'good' or 'bad' design -does not only lead to 'unpleasantness'. Indeed, such dissonance is 'necessary and even desirable for the proliferation of design alternatives' (Farías 2015).
Thus, dissonance is the harbinger of 'newness' (Hutter and Farías 2017), in the form of new design alternatives. However, with reference to the previous discussion on the Machiavellian moment" in design and innovation, the proliferation of design alternatives is not an end in itself. Dissonance and indeterminacy might as well result in the 'collapse' of a project (Stark 2009). Therefore, actors continuously need to find ways of pragmatically 'coordinating' their work. When this is achieved successfully, the new emerges as a 're-combination' -that is, the settlement of compromises creatively arrived at through new combinations of ideas, tasks, actors and materials. The innovation process is a 'search' for viable re-combinations, which have emerged as potentials given the dissonance of different principles of valuation. (Here, there is some overlap with ANT's understanding of innovation as the emergence of an 'acceptable arrangement' among participating actors.) Stark borrows the idea of coordination through negotiated compromise from two other central scholars in valuation studies, Boltanski and Thévenot (2006). Through a number of empirical studies, they have shown how conflicting principles of value (e.g. market, efficiency, loyalty, fame and the civic) in people's day-to-day work are resolved through a number of 'strategies'. These include agreeing on the legitimacy of a specific valuation, and pragmatically settling compromises between incommensurable valuations. Another strategy is that of letting 'an outsider' decide on how the process is to proceed, something referred to as relativisation. Girard and Stark (2002) has also, in an ethnographic study on web-design firms, shown how coordination can also be attained through ongoing processes of realignment and persuasion, as well as to through the existence of misunderstandings. The next section will show how these strategies have been pursued in the coordination of dissonance in the empirical case of co-design of active frontages.

Dissonance and coordination in the co-design of active frontages
This section will describe the co-design of 'active frontages' in the redevelopment of the Masthuggskajen district in Gothenburg, Sweden. It will first provide a brief background to this co-design project, and its participants. It will then -in line with the valuation studies approach outlined above -analyse the process in terms of dissonances (which conflicts of valuation emerged, and how?), and in terms of coordination (which strategies of coordination were employed?). Thus, the aim of the section is to demonstrate which aspects of a co-design process that the valuation studies approach puts centre stage.

Background: the Masthuggskajen process
The object of design at stake in this case is the 'active frontage'; an entity typically understood as the ground floor of a building that features non-domestic uses and has plenty of human activity visible through frequent doors and windows (Heffernan, Heffernan, and Pan 2014;Kickert 2016). Thus, this object of design is a complex and composite one, involving physical design, as well as the design of policies, business models, organisations, and so on (Tonkiss 2013). This is not least due to the fact that an active frontage tends to reduce the amount of rent that can be extracted from ground floor of the building, while at the same time potentially making the neighbourhood as a whole more attractive. An indepth rendering of the process can be found in Molnar (Forthcoming), but the short story runs as follows: The redevelopment of Masthuggskajen began in 2008, when the municipality of Gothenburg started producing a vision of how the area, at that time mainly consisting of parking lots and a handful of office buildings, could be turned into a dense, high-rise and mixed inner-city district. This would be achieved by creating outdoor public spaces and an orthogonal streetscape featuring a number of new mixed-use buildings. In 2012, the municipality entered a cooperation with a consortium of five property development companies to realise this vision through a public-private partnership. A key initiative within this partnership was the creation of a working group called 'the Active Frontages Team', which would guarantee that the property developers really did commit to the construction of active frontages.
The team consisted of at least one representative from each of the six consortium companies -a municipally owned urban development company, four private real estate companies that already owned real estate in the area, and a Swedish multinational property development company. The overarching remit of this team was to find ways of creating active ground floors in the existing buildings, as well as in those awaiting construction. More specifically, this implied exploring business models for the ground floor spaces, not least for less commercially lucrative tenants such as artists, non-profits, municipal social services and sharing economy actors that cannot afford market rents. Each of the members of the team put in working time, as well as additional funding used for hiring consultancy services tied to the team's remit. The team had formal regular meetings together, often at least one to two times a month. Typically, each meeting would host 10-20 participants. Aside from the regular meetings, team members, expert consultants, and representatives from different present or future user groups (artists, social and cultural entrepreneurs, and non-profits) worked together in different constellations, in order to interrogate certain issues in more depth, or jointly co-produce materials such as texts, visuals, and calculations. As such, the Active Frontages Team operated in the intersection of the divergent views of different stakeholders: Not only did they have to manage the dissonant voices from within the consortium; they also had to contend with the voices of experts and future user groups.
While Masthugget was initially referred to as a 'test bed', the language of design was increasingly mobilised within the process, through open explorations of user needs, as well as concept development through prototyping and testing. This approach to the process came in two guises. From 2015 and onward, the process was supported by a Gothenburg-based urban design consultancy, which framed it as a design thinking process, following a double diamond framework. In the consultancy's own description of the project, the key challenge of this endeavour was to 'create a collaborative climate', in order to realise 'the common vision of active frontages' (FOG Innovation 2021). Furthermore, the notions of concept creation and prototyping were also introduced through 'Innovation lab' workshops led by the municipality. Here too, the objective was to use prototyping to 'involve citizens and businesses in urban development' (Fernström 2015, 7), though based on the 'entrepreneurial learning' appropriation of workshop-based practices of pitching, user-testing, iterating, and pivoting. As will be shown below, the process also sought to receive a 'Citylab' certification from the Sweden Green Building Council. This implied that the efforts to develop active frontages were also aligned with the Citylab Guide, which is a co-creation-based (Sweden Green Building Council 2019, 49) process tool for sustainable urban development.
Over the upcoming years the municipality and the consortium, led by the Active Frontages Team, engaged in a range of initiatives leading to a gradual reframing and revaluation of Masthuggskajen´s future 'active frontages': from a space aimed mainly at commercial tenants such restaurants, bars, shops and cafés, to also offering facilities and rental levels which could cater for a greater variety of tenants such as artists, NGOs, social services, and sharing economy initiatives. In March 2019, a land-use plan was adopted, which meant that the goal of creating active frontages had been legally enforced. However, the organisations continued their search for organisational structures and business models that could enable a mix of tenants and rent levels. Construction work for the first building commenced in September 2019, with further construction projects due to be set in motion through to 2024. The district is to be completed by 2028.

Dissonances in the co-design of active frontages
In analysing conflicts among participants in a co-design process, one is tempted to start from the key interests of each actor. Indeed, an ANT-inspired account of a translation process may well start from charting the chief interests of the participating actors. Thus, the municipality has the interest of regenerating the old harbour area into an attractive, mixed and growth-generating, inner-city like district. The five property developers, on their end, enter the process with an interest in developing its own plot of land with a reasonable return on investment. There is also a set of tenant associations and user groups that represent their particular interests, and a variety of consultants representing different forms of expertise.
Still, however important these interests may be, a valuation studies approach shifts the analyst's attention to the different forms dissonances that emerge in this process. Such dissonances are not necessarily in congruence with a mapping of overarching interests. So, for example, the Masthuggskajen process was continually influenced by a dissonance regarding the kind of physical characteristics that signify an 'active frontage'. The dominant view throughout the process was that the active frontage indeed, in congruence with the international discourse on the subject (Kickert 2016), should characterised by facilities at ground floor level with large transparent windows and doors that open towards the sidewalks. However, some participants (within all of the key participating organisations) questioned the dominant discourse, asking why not a bike room, a loading dock or music studio could be referred to as an 'active frontage', as long as it features 'human activity'. 1 Another major source of dissonance concerned what kinds of tenants and uses that should give life to the ground floors. During the early phase of the project, the participants involved, including municipal representatives, tended to frame Masthuggskajen's active frontages as commercial spaces aimed at restaurants, cafés, bars and shops. Even during a later stage, when the notion of 'tenant mix' gained prominence, the dissonance between 'commercial-oriented design' and 'non-commercial-oriented design' remained.
In the final Strategic Masterplan (finished in 2019), 'commercial services' occupied over 70% of the proposed ground space in future Masthuggskajen. This caused members of the Active Frontages Team to question whether property developers really were prepared to go beyond 'business as usual' and orient the ground floor facilities towards alternative uses. Similar dissonances also emerged in relation to the composition of the 'non-commercial' ground floor tenants. Some participants suggested that the process tended to place an undue emphasis on tenants from the cultural sector, whilst municipal social services and NGOs had received less attention.
Yet another form of dissonance concerned that between economic criteria and noneconomic criteria, such as social mix, social inclusion and energy efficiency. The economic criteria were not -as one might suspect -only championed by property developers. For instance, this dissonance was evident within the municipal organisation, which had to balance social and environmental concerns against economic returns on investment. Thus, one local politician lamented the fact that since the municipality is in part driven by an 'economic logic', their investment in social services such as 'kindergartens and schools' is curtailed. This holds particularly true in an attractive area such as Masthuggskajen, where 'more money is at stake'.
Lastly, the design process involved dissonances between the will to preserve Masthuggskajen´s existing cultural-historical values, along with values related to the economy and energy efficiency. This dissonance became prominent during the latter half of the process, when the participating organisations started putting increasing design emphasis on the fact that the new Masthuggskajen district ought to be experienced as an extension of the already existing surrounding neighbourhoods. With a few exceptions, these dissonances did not revolve around the issue of architectural style -around whether certain buildings or characteristics of the built environment should be preserved or not -as very little demolition work was to be carried out. Rather, the dissonances involved whether specific tenants of a certain cultural pedigree should be allowed to persist in the area. They also involved the issue whether diverse 'ways of life' -which currently give the district a specific charactershould be allowed to persist in the future. Taken together, these conflicting values shaped the co-design process, and the subsequent expression of the active frontages concept.

Coordination practices in the co-design of active frontages
So how is it that conflicting principles and practices of valuation, such as the ones outlined above, can be resolved in the form of common designs? A valuation studies approach leads the analyst to focus on a series of seemingly mundane practices that nevertheless allow resolutions to emerge.
One such mundane, everyday practice is talk. Indeed, existing research in urban planning (Beauregard 2015;Flyvbjerg 2002) highlights the role that 'talk' plays in coordination, through meetings, corridor encounters, conferences, as well as over phone and email. This holds true in the case of Masthuggskajen, in which different modes of talk functioned as a key coordination practice. Firstly, talk enabled coordination through providing a means for actors to enrol others. During the early phases of the process, this practice played a significant role within the Active Frontages Team. Over time, the format for talk shifted, turning it more into a device for joint learning, finding agreements or settling compromises.
In the co-design process, such new formats for talk were constantly and pragmatically created, with the aim of solving specific problems or finding joint agreements. For example, in 2016, the Active Frontages Team had formed an agreement around 'tenant mix' as an important criterion of worth which would characterise Masthuggskajen's future ground floors. Since there were still dissonant views on what this mix could imply, several cross-stakeholder working groups were formed. The remit was to -quite simply -talk their way through the question of what the tenant mix should look like. During consortium meetings, it was evident how much hope some of the participants placed in 'talk' as a way to learn, find agreements and solve problems. As one participating property developer put it: 'We have a lot of expertise in the group sitting here, we don't really need consultants, we will probably manage quite well ourselves, through reasoning with each other'.
A second, and related coordination practice involved introducing or inventing words and concepts. For example, terms like 'identity', 'active frontage', 'mix' and 'horizontal house of culture' 2 were instrumental in the Masthugget process. They were jointly developed, agreed upon, and used by the consortium actors and their consultants over several years. 3 Through agreeing that a certain word or concept should be used, the participants felt that they were making progress and working towards the same goal. At times, such terms had the effect of obfuscating dissonances, making the participants believe that they were in agreement when in fact they were not. This coordination through misunderstanding has the effect of suspending frictions, though in an indeliberate manner. As such, it differs from more formal ways of deliberately postponing the resolution of a thorny issue.
As with the role of talk, the introduction of terms and concepts sometimes also functioned as means of enrolment and persuasion providing scripts for what should be done and how. For example, the introduction of the term 'active frontages' provided a way for the municipality to get the property developers on board, as this term seemingly denoted business development and commercial opportunities, compared with for example a term such as 'sustainability'. There are also several examples of how situations of dissonance forced the actors to invent new terms, by drawing on different perspectives and evaluative criteria as a means to get the process going. One example is how the notion of 'shared spaces' was emphasised during the later phase of the process, as an attempt to square the difficult equation of achieving 'tenant mix' while accepting the principle of 'market rent'. As it were, shared spaces such as co-working spaces were construed as a way in which low paying tenants could find space in Masthuggskajen, whilst providing the property owners with market rents.
A third coordination practice has involved creating or introducing documents and tools. In Masthuggskajen, such devices at many times enabled the spread and legitimisation of certain scripts or frames among involved participants. There are several examples of how participants explicitly created specific tools or evidence bases as a means to strategically enrol people into them in their own plans. A few years into the scheme, for example, the municipality took help from a team of researchers from University of Gothenburg in order to illustrate the qualities and values of a local flea market which was threatened by demolition. Even though the flea market building is still, to this date, awaiting demolition, this valorisation practice (Vatin 2013) succeeded in the sense that it resulted in the joint agreement of the idea that the activities of the flea market would move into the ground floors of the new buildings. At other times, documents and tools were co-produced among the involved parties through the creative combining of the disparate beliefs, ideas and objectives into objects built on joint compromising. One example of such an object is the Strategic Masterplan, produced during the final phase of the process, as a way to solidify the exact meaning of the notion of ground floor mix. As the consortium parties could not come to an agreement on what types of tenants to cater to, the Masterplan spatially divided the future ground floors into a number of sectorial clusters, each cluster catering for a specific sector, such as culture, food and beverages or sharing and re-use.
A fourth coordination practice displayed in Masthuggskajen was the creation of new organisations. As suggested by Thévenot (2014), organisations can sometimes function as coordination devices by assembling disparate actors around certain shared and more or less compromised evaluative criteria and formats. One of the clearest examples from Masthuggskajen is the establishing of the Active Frontage Team, something resembling an organisation within the larger project organisation with its own objectives, budget, permanent roles and members, chain of command, and routines. The group was created so that the consortium organisations and the City of Gothenburg together and over time would experiment with different models for how the goals of mixing and rent variation could be achieved in the long term. Through the organisation, a stable structure and common working method was created, which allowed coordination of the members' actions despite basically different assignments and interests. In a similar vein, the consortium plans to establish an organisation that will safeguard the continued maintenance of the 'active frontages' idea. This organisation is meant to transcend individual interests, and seek to preserve the vision, coordinating matters like 'jointly share' responsibility for securing lower rents and the 'mix' objectives. This would not only enable coordination through the agreement of certain values and routines, but also through providing a means of sanctioning organisations that did not live up the organisation's standards.
A further, fifth coordination practice that was enacted in Masthuggskajen was involved seeking help from outside the project organisation. For instance, such 'relativization' practices (Boltanski and Thévenot 2006) involved the invoking of outside standards, such as the certification program Citylab. At a particular time in the process -when it was ridden with conflicts and low levels of trust among participants -the Citylab program acted as a neutral forum to generate new conditions for collaboration across organisational boundaries. Moreover, the never-ending flow of outside experts and consultants is another example of coordination through relativisation. So too is the influence of outside legal expertise, for instance on the issue of the legal conditions for the creation of rental mix.
One final, sixth practice of coordination that was evident in Masthuggskajen was the deliberate decision to push unpleasant subjects into the future, thus suspending any talk about that issue. This practice allowed the project to move forward on other fronts. This was evident in the case with the objective of ground floor 'rental mix'. During interviews and meetings, participants would regularly suggest that the practical modes of realising the idea of 'mix' were to be discussed at a later stage. In fact, plenty of dissonances still existed on the issue of mix, throughout the process. Moreover, the 'rental mix' question was sensitive in legal terms, influencing how 'subsidized rents' or losses of income should be distributed among the parties. However, despite such obstacles, the shared commitment to abstract principles of 'mix', 'new business models' and 'experimentation', enabled participants to muddle through the co-design process.

Concluding discussion: dissonance and diplomacy
The case study above illustrates how co-design processes can be described as that which happens when dissonant valuation practices are resolved into a shared design. This section will first discuss what this focus on value dissonances and coordination practices brings to the analysis of co-design processes, comparing the approach to ANT and other approaches. It will then turn to consider potential shortcomings of this perspective, exploring what the analyst may lose sight of when studying co-design as a process of settling value conflicts.
In congruence with the structure of the previous section, a comparison with other approaches may usefully be structured around the two terms dissonance and coordination. The notion of dissonance in valuation practices is a good starting point to explicate how this approach differs from that of agonistic design (Björgvinsson, Ehn, and Hillgren 2012;Di Salvo 2012). The latter approach suggests that design may facilitate the reemergence of politics by explicating conflicts among actors with particular interests. Following the theoretical presuppositions of this perspective, these interests are more or less fixed, and determined by socio-political forces (Palmås and von Busch 2015). In Stark's valuation approach, the conflict is not between actors with fixed interests, but between principles and practices of valuation, which are not necessarily tied to either actors or particular interests. Indeed, the dissonance of clashing values may emerge within any single actor or participant in a co-design process.
So, for instance, in the process described above, the municipality did not merely represent the public interest against the private profit motive of the property developersrather, the municipality was within itself struggling with the cognitive dissonance of balancing social values, environmental values and economic return on investment. Similarly, a developer may promote social values of 'social mixing' in the context of one meeting, and the values of return on investment in another. Further, in valuation studies, there is a focus on collectively shared valuation principles and practices -say, the value of the environment, or the value of economic growth -not private motivations or beliefs.
On this account, Stark's approach also differs from that of Schön, and that of ANT. For Schön (1984), the conflict to be resolved is not one of interests as such, but of frames, which are nevertheless tied to individual actors. In ANT, there is no such thing as a-priori individual interests, as these are configured by the network around the actor, and thus become subject to translation and drift. Nevertheless, there is such a thing as an interesteven though it is mutating and detached from individual actors. This also means that an ANT-inspired study of co-design has another focus than a Stark-inspired one. Thus, Palmås and von Busch (2015) studies the process of translation through the chain of material inscriptions -be it post-it notes, PowerPoint presentations, posters, newspaper articles, or municipal planning documents. A Stark-inspired study will instead focus on traces of valuation in the co-design process -such as verbal or written communication of valuation principles enacted in negotiations, or socio-material apparatuses that support the practicing of valuation. Nevertheless, the two approaches are similar, inasmuch as they highlight how a co-design process is a process of negotiation and stabilisation, transforming 'a plethora of dissonant voices' (Palmås and von Busch 2015) into a shared settlement.
As regards to coordination, the approach introduced in this article shows the different ways in which designs may evolve out of value conflicts. Le Bail, Baker, and Détienne's (2020) laboratory study of value conflicts in a design process concludes that 'value conflicts cannot be genuinely resolved, in the sense of achieving sincere agreement' (18). This proposition does not, in itself, contradict Stark's approach, which suggests that agreements do sometimes emerge among actors -even in situations when the dissonance persists. Bluntly put, the shared outcomes of co-design processes do not necessarily solve value conflicts among participants; they only suspend them.
Again, this point goes back to the fact that for Stark, any one actor may be torn between several different modes of valuation. In contrast, the Le Bail, Baker, and Détienne (2020) laboratory study was premised on letting each participant act as a representative for one particular mode of valuation. Here, it is also worth noting that Stark's account of coordination rests on empirics of a more macro-sociological nature, namely Boltanski and Thévenot's (2006) work on how value disputes in the public sphere are generally resolved.
In summary, a valuation studies perspective does shine new light on specific parts of co-design processes. Nevertheless, in doing so, it may also obscure other aspects of such processes. This problematic may be approached by examining the question of the values espoused by the analyst. For instance, Friedman and Henry (2019) seek to devise prescriptive design methods which moderate the values of participants, while also actively instilling human values into a design process. This position is unabashedly idealist (von Busch and Palmås 2016b, 18-19), positing that design, when construed and executed properly, can represent 'the good'.
The alternative approach, which is more in line with the value agnosticism of valuation studies, is to aspire to the descriptive and diplomatic. Here, the task of the analyst is to explicate how design emerges in the context of such disputes -studying which modes of valuation that become dominant in the design process, and which modes of valuation that become side-lined. This approach heeds Pedersen's (2016, 182) call for 'a more pluralistic and descriptive understanding of the "co" in codesign', based on studies of 'how designers and relevant stakeholders in actual practice work together' (italics added). Further, Pedersen suggests that the task of co-design practitioners is to act as modest peacemakers who refrain from taking sides. This proposition is also in alignment with the valuation approach's focus on innovations and designs as outcomes of negotiations and settlements.
Here, there is an overlap with some of the aims of the agonistic design, which posits that the ambition of co-design should be to reinvigorate democratic dissensus through the cultivating of peaceful agonism (Björgvinsson, Ehn, and Hillgren 2012). However, this very affinity with the agonistic design approach also signals a potential pitfall of the valuation studies approach to co-design. According to Kiem (2013) and Di Salvo's (2012) version of the agonistic design approach extols the democratic virtues of 'the political' over managerialist-administrative 'politics', celebrating open-ended dissensus -yet fails to explicate the implicit meta-values and taken-for-granted assumptions that nevertheless underpin the account. For instance, Kiem suggests, this brand of agonistic design presupposes a liberal democratic order. Moreover, Di Salvo fails to define what he means with 'democracy', leading the term to be mobilised 'as an unexamined signifier of "the good"' (Kiem 2013, 35). Thus, the proposition that design should articulate democratic dissensus is deceptive. It hides meta-values that are not accounted for by the analyst, and obscures the limits of 'the political' in a particular design process.
Arguably, the same charge can be levelled against the valuation perspective, as explicated in this article. In the case of the Masthuggskajen process, there were, after all, clear limits of the political. Indeed, the value conflicts played out in the case were among a set of actors who had accepted the general framework of co-design organised within a public-private partnership. Just like notions of agonism and dissensus, the idea of co-design as modest diplomacy may obscure the view: The negotiating parties in Masthuggskajen may have been torn between different modes of valuation, but they all accepted the overall framing of the process as a creative search for mixed active frontages, within pre-established political limits. Indeed, the wider stakes of the development of the area had already been settled by other modes of doing urban politics -that is, in closed negotiations between the municipality and development companies. Indeed, this diagnosis chimes with extant critiques of new urban governance arrangements (Swyngedouw 2005). This is not to say that co-design invariably plays a depoliticising role -just that the valuation perspective may fail to spot the instances when it does.
Such shortcomings aside, the valuation perspective does nevertheless reframe the study of co-design processes. In deviating from traditional modes of social inquiry, focusing on conflicting values rather than conflicting actors or interests, it prompts us to see cocreations as mundane settlements arising from uneasy states of dissonance.

Notes
1. The quotes in this and the following subsection are sourced from the participant observations. 2. This is a reference to the high-profile and high-rise, 'vertical', house of culture in Stockholm. 3. The terms occur frequently in the participant observations, meeting notes, plans, design briefs and strategy documents.