Extracts: Most agricultural research and development (R&D) institutions have been shaped by traditional approaches to technology transfer, in which farmers, extension agencies and civil society organisations passively accept the technologies delivered by researchers.

In recent years a group of interdisciplinary researchers has been engaged in a process of critical thinking about impact assessment techniques, ‘farmer first’ approaches to agricultural research, and a systems view of innovation. One of these approaches has involved the creation of learning alliances.

The learning alliance approach emphasises the processes of innovation, and involves collective learning by research organisations, donor agencies, policy makers, civil society organisations and even private businesses. By improving the flows of information and knowledge, these multi-stakeholder platforms help to speed up the process of identifying and developing innovations, and ensuring their adoption by farmers.

The learning alliance also provided a platform for scaling up the use of the system as an alternative to the input-intensive practices that were likely to fail. The alliance has therefore helped not only to reduce environmental stress, but also to open up new markets for the region’s organic produce, both local and international, thus increasing the incomes of many small farmers.

The success of the learning alliance approach is based on the ability of the facilitating organisations to open up channels of communication between diverse partners. In particular, organisations with experience in designing and testing analytical tools and methods can facilitate collective learning within and between organisations.

Learning together often pushes existing institutional arrangements to become more open. The results can be surprising, sometimes leading to reversals of traditional roles – extension agencies and civil society organisations doing research, and research institutions repositioning themselves as knowledge brokers. They could lead to new learning laboratories and platforms where researchers can learn, reflect and report even if they do not have all the answers. This can lead to new knowledge emerging from healthier and more equal interactions among hitherto powerful scientific hierarchies. All of these changes are to be welcomed if we are serious about addressing the complex challenges facing the ‘bottom billion’ in the future.