273 Script or improvisation? Institutional conditions and their local operation Barbara van Mierlo and Edmond Totin Knowledge, Technology, and Innovation Wageningen University, barbara.vanmierlo@wur.nl Abstract: In Benin a combination of governmental programmes effectively stimulated rice inten- sification by providing relevant institutional arrangements like subsidised seeds, credits and a market outlet. In this paper, we investigate the institutional character of the programmes, by un- packing the rules embedded in them as well as by showing how farmers mould, reject and change these rules or combine them with local rules; their practices of institutional bricolage. We show that the services provided by the programmes had great advantages for the rice farmers, but also an exclusive character. Because of local bricolage practices the programmes impacted not only rice production practices but also helped the rice farmers to deal with conflicts about inequitable land allocation and discriminatory participation in canal cleaning. These findings contribute to the discussion about the role of innovation platforms in the stimulation of institutional change and providing enabling conditions. Keywords: Innovation platforms, institutional conditions, institutional bricolage, rice production, water management , practices Introduction Innovation platforms are expected to bring key-actors together to exchange knowledge and stimu- late innovation of products and services and their adoption and diffusion (Consoli & Patrucco, 2008 ; Nederlof et al., 2011 ; Klerkx et al., 2013). The CoS-SIS (Convergence of Sciences – Strengthening Innovation Systems) programme implemented in three West African countries (Benin, Mali and Ghana) has facilitated the set-up of innovation platforms especially with the aim to reduce the institutional constraints that hinder small farmers taking advantage of technical op- portunities to increase food production, because they lack the power to change these constraints on their own (Hounkonnou et al., 2012 ; Röling et al., 2012). By bringing together farmers with powerful actors from a value chain and letting them conduct so-called institutional experiments a conducive learning process may be stimulated. At some point these Concertation and Innovation Groups (CIGs) might transfer into a true innovation champion able to generate systemic change. Notwithstanding the general agreement in the literature that systemic change cannot be managed or steered, and if at all possible, needs an adaptive, flexible approach, the huge rise of multi- stakeholder and innovation platforms set up in practice shows optimism about the possibilities to stimulate change deliberately and effectively. Processes of institutional and systemic change and the role of platforms therein need to be well understood for these platforms to be able to be effective. Institutional bricolage is a term devel- oped by Cleaver and others to understand why many interventions with designed institutions have unintended effects, in the sense of unexpected outcomes or a failure to reach the aim of improv- ing people’s livelihood (Cleaver, 2002 ; De Koning & Cleaver, 2012). This critical perspective asserts that institutions cannot be designed in a vacuum given the multiplicity and ambiguity of the institutions guiding people’s practices at a local level as well as their agency, their ability to mould institutions. A way to picture the idea of bricolage practices is as the performance of a