Wandering off the paths of planned innovation: The role of formal and informal intermediaries in a large-scale irrigation scheme in Morocco J. Poncet a,c, * , M. Kuper b,c , J. Chiche d a UMR G-EAU, International Centre of Agronomic Research for Development (CIRAD), Aix-en-Provence, France b Umr G-eau, CIRAD, Rabat, Morocco c Cirad, UMR G-EAU, Montpellier, F-34398, France d Department of Social Sciences, Hassan II Institute for Agronomy and Veterinary Medicine, Rabat 10101, Morocco article info Article history: Received 16 February 2009 Received in revised form 11 November 2009 Accepted 1 December 2009 Available online 15 January 2010 Keywords: Innovation system Innovation network Innovation intermediary Extension Large-scale irrigation schemes Morocco abstract For many years, large-scale irrigation around the world was based on state-driven water management and on a planned innovation process and diffusionist extension services. The purpose of this study is to show that formal extension services are not the only intermediaries of innovation and that local inno- vations take place at the initiative of farmers even in state-driven irrigation schemes. The study addressed changes in farming systems in a large-scale irrigation scheme in Morocco from two angles: a review of planned innovations and analysis of the actual innovation process at the village level. We show that the implementation of the large-scale irrigation scheme contributed to agricultural develop- ment, but often indirectly, and that it was not the only source of innovation. We also show how these results support recent thinking on innovation systems. Today, informal labour, neighbour and marketing networks are the main innovation intermediaries. New practices in agricultural extension are required to facilitate local innovations and to link farmers to more global innovation networks. Ó 2009 Elsevier Ltd. All rights reserved. 1. Introduction Large-scale irrigation worldwide has long been based on state- driven development and management, and on a planned innova- tion process and ‘diffusionist’ extension services. This centralized and linear model for the transfer of technology has been highly criticized, and the results obtained are often judged disappointing. Wisner (1979) stressed the quasi-systematic failure of technology transfers in developing countries and proposed a new field to adapt technology to a given population using knowledge from the social sciences. In this study, innovation is not restricted to the invention, diffusion and adoption of technology, but is defined as the interac- tive process of developing and implementing a new idea, tech- nique, know-how or institution (Olivier de Sardan, 1995; Van de Ven et al., 2000). In the linear view of innovation, extension services transfer re- sults of agricultural R&D to farmers. This was certainly the princi- ple on which the implementation of large-scale irrigation was based in Morocco. However, the linear view of innovation in devel- opment projects has been vigorously contested and consequently replaced by innovation systems approaches (Edquist, 2006; Klerkx and Leeuwis, 2008a). In these approaches, innovation is considered as a complex interactive and learning process (Klerkx and Leeuwis, 2008b; Douthwaite et al., 2003) and an innovation system is the combination of different factors – economic, social, political, organisational, institutional – that influence the development, dif- fusion and use of innovations (Edquist, 2006). In this way, research is no longer seen as the dominant provider of knowledge and infor- mation but as one of the many stakeholders within an innovation network (Sumberg, 2005). These innovation networks have be- come more and more complex (Klerkx and Leeuwis, 2008c) due to the redefinition of the role of government extension services, and to the development of agriculture (diversification of products, specialization of producers). Many authors have stressed the role of multiple actors in inno- vation processes (e.g. Akrich et al., 1988; Chauveau et al., 1999). Some authors specifically highlight the role played by farmers and their networks in agricultural innovation processes (Olivier de Sardan, 1995; Chiffoleau, 2005). Deugd et al. (1998) and Darré (2006) argue that socio-technical networks represent a space where actors (such as farmers) produce knowledge for innovation. Given the heterogeneity of actors and networks, the emphasis is on improving the connectivity of the different groups (Spielman et al., 2008; Klerkx and Leeuwis, 2008c). Over the last 20 years, interme- diaries in innovation processes have been studied in a range of 0308-521X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.agsy.2009.12.004 * Corresponding author. Address: Umr G-EAU, CIRAD, Aix-en-Provence, France. Tel.: +33 633424381. E-mail address: julie.poncet@cirad.fr (J. Poncet). Agricultural Systems 103 (2010) 171–179 Contents lists available at ScienceDirect Agricultural Systems journal homepage: www.elsevier.com/locate/agsy