WS1.4 – Design methods, system approaches and co‐innovation 9 th European IFSA Symposium, 4‐7 July 2010, Vienna (Austria) 393 Co‐design as a distributed dialogical design Pascal Béguin ab , Marianne Cerf a , Lorène Prost a a INRA, UR SenS; pbeguin@grignon.inra.fr, cerf@agroparistech.fr b CNAM, Center for Research on Work and Development; prost@grignon.inra.fr Abstract: We propose the following four key assumptions, illustrated with three examples, to analyze and to organize and monitor design processes involving users. 1) Users redesign the designers’ technology by using it. Thus, the coupling of the technology with the users’ new activity is at the core of the process. 2) Design is a process distributed among various people whose interdependency has to be taken into account during the process. 3) Developing both the technology and the activities implies various levels of dialogue that we define, referring to Bakthin’s work. 4) Focusing on one of these levels, we argue that a key issue is to highlight the various actors' differing perspectives during the design process. Keywords: co‐design, learning, instrument, participatory design, innovative dialogue. Introduction To meet the challenges of sustainable development in agriculture, Meynard et al. (2006) argue the need for redesigning farming systems through an innovative design, i.e. a design that produces both new technical specifications and new knowledge ‐rather than using the available knowledge‐. We share this point of view, but believe that this technical dimension must be handled in close relation to the “political” dimension that contributes to identifying a desirable future (see also Godard and Hubert, 2002). In this paper we present four theoretical assumptions to anchor the monitoring of co‐ design processes and their analysis. We view co‐design as a process in which: (i) technical dimensions, on the one hand, and knowledge, practices and values, on the other, evolve jointly; and (ii) a desirable future is collectively discussed in order to define and implement acceptable solutions. In doing so, we hope to participate in debate on design in both the farming system research community and the design studies community. Our perspective emphasizes the need to foster cross learning processes amongst designers and users in order to achieve the joint building of a technology, of a desirable future, and of the activity or the collective action in which the technology will be used. Many authors have already suggested involving users in the design process. We can roughly distinguish three trends. The first one is well represented by Von Hippel’s work on lead users (see for example Von Hippel, 1986). In this trend, the main focus is on firms’ competitiveness. Enrolling lead users in viewed as a win‐win partnership: on the one hand, the firm that launches a new technology can anticipate some difficulties which might occur at the time of its marketing; on the other hand, the lead user firms can secure a competitive advantage due to quick access to the new technology and potential transfer of skills from the partner. The second trend is well represented by user‐ and use‐centred design methodologies. These were devised to develop a more effective technological design process while acknowledging the fact that use and users are always crystallized in a given technology. Their aim is to take on board use and user issues so that the technology will fit the requirements set for it (for more details see for example http://www.upassoc.org). But in this approach users document the design process, without necessarily being actors in it (Caroll, 1996). The third trend is known as participatory design (for example, see Kensing and Blomberg, 1998). Participatory Design (PD) takes on board some political issues while the former two trends did not. More specifically, PD researchers recognize that technologies can have strong impacts on workers, and therefore claim that workers can legitimately contribute to the design process. Our own perspective is related to PD, but we argue that there is a need to take users' inventiveness into account before crystallizing certain uses and users’ representations within the technology. We address this issue by focusing on the coupling that occurs