Know more

 “Innovation” has become central to the strategic orientation of many agricultural research institutes, particularly to address the challenges faced by agriculture as a result of new societal expectations and the multiple types of performance involved (agronomic, economic, social, and environmental). Nevertheless, the emphasis remains on upstream knowledge production, in keeping with the conventional understanding of “innovation”: a result that ensues linearly from the production and transfer of knowledge. We, on the other hand, propose to approach innovation as a process and to investigate the way in which it transforms the production of scientific knowledge. Thus, even though iterations between knowledge production and the definition of design problems are mentioned in the literature, little attention has been paid to the conditions and particularities of the production of scientific knowledge on the functioning of agroecosystems in the course of and through design processes.

We carried out a comparative case study of nine projects involving design processes that produced one or more “innovations” in the field of agroecology (e.g. low-input crop management sequences, variety association rules for multi-variety crops), and that were led by agricultural researchers (agronomists, geneticists, plant physiologists, ecologists). The case studies varied in duration (from 3 to 15 years) and spatial scale (from plot to landscape). The analysis combined semi-structured interviews with the researchers leading each design initiative and a literature review (articles, theses, technical publications, research projects or reports).

Our results show that in all the cases studied, the production of original and generic scientific knowledge on the functioning of agroecosystems is also an outcome of the design process. Timewise, such knowledge production takes place at different points in the design process. The originality of the knowledge produced partly stems from the inclusion of a targeted action (and not just the production of knowledge on the functioning of the system studied). It also relates to the evolution of the researchers’ representations over the course of the design process, be they representations of the processes of interest in the agroecosystem (e.g. they include objects that were previously ignored or neglected), or of ways to act upon it. These new representations, as we show, are gradually formed over the course of iterations between successive formulations of what the desirable object is and what the knowledge needs are. Finally, our results underscore the role of confrontations with real action situations in the evolution of these representations. These confrontations often take the form of diagnoses of the variability of situations (e.g. diagnosis of uses, agronomic diagnosis).
The results serve to discuss the scientific knowledge produced during and through the design processes undertaken by researchers, with different stakeholders in the agricultural profession. In particular, we show that in addition to knowledge that is situated, valid and usable mainly in the context of the innovation process, more generic knowledge is produced, and can even open up new fields of knowledge. Furthermore, our findings provide a basis for discussing approaches  to framing the problem to solve in design processes. While implementation is often understood as a repetition of loops between problem identification, the generation of solutions, and finally the implementation of solutions, we show that it is itself a driver of new knowledge production, and enables more than just refining a prototype.