Michel Pimbert International and national agricultural research is entrenched in a culture of top-down and often insensitive approaches to realities on the farm. This article by Dr. Michel Pimbert highlights the mismatch between the transfer of technology model of agricultural research and the needs and livelihood strategies of the poor. Michel is an agricultural ecologist and has conducted much research on biological pest control. He spent four years working at ICRISAT where his people-centred approach to research clashed against the internal norms of Green Revolution science. As Michel sees it, the professional challenge of the 1990s is to develop innovation systems and sustainable agricultures that support decentralisation, diversity and democracy rather than centralisation, uniformity and control. The transfer of technology model (TOT) of agricultural research is typical of both national and international agricultural research systems. In the TOT model, all the key research decisions are made by scientists who experiment on research stations or under controlled, simplified conditions in farmers ' fields. The resulting agricultural technology, such as pest resistant varieties and recommendations on fertilisation, is then handed over to the extension services for transfer to farmers. Industrial and green revolution agricultures have been well served by this model of agricultural research. Reductionist research, high input packages and top down extension have led to successes: in the uniform and controlled conditions of industrial and green revolution agriculture they have raised output per unit of land. The simplifying tendencies of reductionist science have meshed well with the ecological and social simplicity of standardised, specialised farming systems. The burden of blinders However, the TOT model of agricultural research has had limited successes in the context of complex, risk-prone, diverse environments where the majority of the world's rural people are dependent on this type of traditional agriculture which is mainly rainfed, on undulating lands and found in mountains, hills, wetlands and the semi-arid and people live today. The physical and economic conditions on research stations have, after all, been very different to those of resource poor environments. Some 1.4 billion humid tropics. For many agricultural technologies developed within the TOT framework, failure rates have been and remain high: the research priorities often turn out to be wrong, the packages are rejected, the technologies do not fit, are non-sustainable or inequitable because of an emphasis on purchased inputs in resource-poor contexts. Examples include: pest management research based on scientists ' perceptions of pest problems on research stations rather than on data derived from reliable pest surveys and farmers ' rankings of pests in order of importance; or farmers ' non adoption of improved high yielding, pest resistant crop varieties on account of their poor taste or cooking qualities. Agricultural scientists tend to perceive farming systems through the narrow window of their professional discipline. Their training has taught them to look at the aspect of farming systems on which they specialise. This is usually their main focus of attention when visiting a farm. However, there are many internal linkages that matter in farming systems, particularly in the complex farming systems that resource-poor farmers often want, but which professional disciplines neglect. For example, the link between crops and livestock is often described in terms of “left-overs”, as “crop residues”. But in many farming systems, the stover, used as fodder, is a vital part of the crop and of the farming system. This disciplinary specialisation often hides from professional view the risk minimising strategies built into traditional farming systems. Resource poor farmers often try to reduce risk by complicating and diversifying their farms and household enterprises. Furthermore, disciplinary specialists tend to adopt one or two single criteria to measure performance, e.g. yield, pest resistance. But farmers as managers of complex, risk-prone systems have many different criteria which they weigh up and combine in the choice of crop varieties or in the choice of farm or watershed management activities. When assessing improved pigeon pea varieties in India, women farmers had some ten different criteria for assessing varieties that had been advanced by ICRISAT scientists on the basis of two criteria: yields and pest resistance alone. This raises the central question: whose knowledge counts? Whose priorities and preferences count? Those of the scientist or those of the farmer? Much of the R&D programs on sustainable agriculture in the CGIAR and the NARS are attempts at systemic adjustments to the sustainability crisis rather than approaches that hinge on fundamental structural change. Agricultural science still operates on a narrow intellectual base emphasising single inputs for each purpose and ignoring the broader implications of recommended technologies. Research is still dominated by the search for marketable input commodities rather than by ecological and social knowledge geared to reducing the need for inputs. The professional challenge for the 1990s The professional challenge of both international and national public agricultural research is to: acknowledge the mismatch between the TOT model of agricultural research and the priorities and needs of the poorest sections of rural society; and recognise and build on the potential of complex, diverse and risk prone farming to meet the twin goals of sustainability and livelihood security. In practice this means that outside professionals (scientists, donors, development planners, policy makers...) should reject the arrogant dismissal of non-scientific or people's knowledge without adopting the naïve, uncritical, view that grassroot organisations and farmers always know best. There is now considerable evidence that experimentation is the norm rather than the exception among rural communities, particularly — but not exclusively — in developing countries. However, it is still heresy to many of today's agricultural scientists and economists to suggest that farmers and grassroot organisations have much to say in the process of technology generation, diffusion and adaptation. Facing the professional challenge also means that rural people should meet scientists on terms of equality. Outside professionals have to recognise that ordinary people have something to teach them and can become involved in key decisions relating to R&D priorities (from plant and animal breeding to the overall design of diverse farming systems and watershed management schemes). The crisis of the TOT model has already led some agricultural scientists to explore new approaches that hinge on farmer participation. These Farmer First approaches reverse parts of the TOT model. Rather than blame farmers ' ignorance or farm level constraints for the non-adoption of agricultural technology, a reversal of explanation points to deficiencies in the technology and the very processes that generated it. A reversal of learning has researchers and extension workers learning with and from farmers and rural people. Roles and locations are also reversed, with farmers and farms central instead of research stations, laboratories, scientists and abstract theories. Analysis, choice and experimentation are conducted by and with farmers themselves, with researchers and extensionists in a facilitating and support role. To combine effectively the theoretical insights and technical power of western science with indigenous knowledge, both Farmer First and TOT approaches are needed in agricultural research seeking sustainable agricultures. This more inclusive research paradigm is still largely in its formative stages. It recognises that both scientists and farmers have limitations and strengths, and so the challenge is to forge active complementarities between these social actors and fully express their comparative advantages in generating sustainable agricultures. Worldwide, there already exist examples of participatory research in which farmers and rural people play a greater role in shaping the directions taken by science and technology. The gene bank of Ethiopia involves farming communities in the conservation of genetic diversity. On-farm landrace conservation focuses on major food crops like sorghum, chickpea, teff, field peas and corn. The complementary knowledge and skills of scientists and farmers ensure that germplasm is conserved in a more dynamic and safer way than is the case in most other gene banks of the world. In Zambia, Colombia and India, participatory plant breeding and germplasm evaluation efforts that involve farmers, non governmental organisations and scientists have consistently better served the needs of farmers and rural people in complex, diverse and risk prone environments than conventional top down approaches. A rich repertoire of participatory methods is increasingly allowing outside professionals to learn with, by and from rural people and to create a working relationship in which people's priorities and values become more fully expressed in projects aimed at conserving and using agricultural biodiversity and other natural resources. Appropriate behaviour and attitudes allow outsiders to establish rapport, convene, catalyse, facilitate, adapt, “hand over the stick”, watch, listen, learn and respect. Meanwhile, rural peoples ' sense of empowerment grows as they map, model, diagram, interview, quantify, rank and score, inform and explain, show, discuss and analyse, plan, present and share their knowledge and experience with others. In these examples, scientists have clear advantages at two levels of organisation. At the micro level: accurate identification techniques for causal agents of diseases; taxonomic skills needed to identify pests and natural enemies (for biological control); instrumentation and expertise needed to understand cellular, physiological and behavioural processes. At the macro level: satellite remote sensing to spot biotic stresses; computer assisted geographic information systems and worldwide electronic communication networks and data banks. But the collective knowledge that farmers and rural people have of their watersheds and agroecosystems gives them distinctive advantages at the mesolevel — where the agricultural technologies are ultimately aimed at. This is, after all, the social and ecological context in which farmers experiment, adapt and innovate. Farmers Rights: democratising research The above are examples in which scientists, extensionists and farmers are more equal partners in agricultural research and development. To date, however, these initiatives remain marginal within the CGIAR and the NARS. Donor agencies as well as top and middle level managers of agriculture research institutes clearly have a role to play in encouraging the spread of participatory research model that seek to empower rural communities in the definition and implementation of their own development goals. Their challenge is to stimulate the creation of: 1. New learning environments for professionals and rural people to develop capacities. An interactive learning environment encourages participatory attitudes, commitment, and contributes to jointly negotiated courses of action. 2. New institutional environments. Institutional support is essential for participatory innovations to spread between and within institutions, and for innovators to gain the confidence and freedom to act and share. Without appropriate incentives and reward systems it is unlikely that participatory approaches that support local innovation and enhance local capacities will spread into mainstream agricultural R&D. The policies of donor agencies and institute managers will largely determine whether the CGIAR system promotes participatory approaches and methods as core professional activities, or whether these will remain isolated and marginalised within the IARCs. This a key challenge since the CGIAR has a professional influence out of all proportion to its size and budget. Through their training of national scientists and their prestige, the IARCs spread and reinforce the dominant concepts, values, methods and behaviour of agricultural science. The very concept of Farmer's Rights offers a unique opportunity to officially reestablish farming communities as the key players in the creation, the conservation and the sustainable use of genetic diversity (and, more generally, of agricultural biodiversity). Farmers Rights has come to describe the whole spectrum of requirements that enable farmers to fully benefit from that part of biodiversity that nurtures people. For example, in order for plant genetic materials to be a resource, farmers must have control over their own biomaterials and have access to the widest possible pool of genetic material. Farmers have a right to retain their own knowledge about genetic resources and to access all the information about their material that is available elsewhere. Also, in order for farmers to develop their resource, they need funds. Farmers must also be free to develop their own technologies and to take advantage of other technologies they find useful. Lastly, recognising that germplasm information, funds and technologies function within farming systems, cultural systems and also marketing systems, among others, the concept includes their right to choose and retain those systems that best meet their needs. In order to implement Farmers Rights, the key challenges for scientists are to act as searchers and suppliers of germplasm as well as removers of legal and administrative obstacles to the spread of farmers ' innovations. It calls for a radical restructuring of national agricultural research systems to include farmers and grassroot organisations as equal partners in innovation and give them more control over research priorities. The decentralised R&D approach described above is uniquely suited for generating diverse and knowledge rich sustainable agricultures. Moreover, its high level of participation also satisfies the equity criterion: it allows farmers to make their own demands on their national research organisations and introduces some measure of accountability and democratic control over agricultural research and extension. However, whilst these approaches support diversity, decentralisation and democracy, they do not, in and by themselves, guarantee public participation in science and the design of technologies. Democratising research obviously implies broader reforms within the scientific community itself and the social forces that largely determine today's public research agenda. For example, the social context of public plant and animal breeding work is such that the directions and uses of publicly funded research are increasingly specified by those who hold power in the food system. Official conservation programs (germplasm collections, gene bank policy, use of genetic materials...) often reflect and reinforce these “commercial imperatives”. The biotechnology revolution planned by and for corporate capital is transforming and further subordinating much of public research to its own ends. Universities and public research institutes, suffering funding cuts, increasingly do work contracted by the private sector — finding that they have to keep their research results secret while the corporations apply for patents on products partly or entirely developed with tax payers money. This greatly restricts the free flow of scientific information, to the detriment of learning and innovation. In this disturbing context, democratising research first means increasing public funding for research and, at the same time, restructuring agricultural R&D to allow for people's participation at all levels. Top down transfer of technology approaches need to be replaced by innovation systems that broaden the circle of social control over decisions on how biological resources are managed and used — and for whom. Budgetary allocations should clearly reflect and reinforce the goals of sustainable agricultural development. Appropriate changes in the training and reward systems of scientists and extension staff are required to encourage more equitable, participatory research modes in collecting, conserving, genetic and species diversity and in designing agroecosystems that rely more on nature's diversity and resilience than on capital intensive “solutions”. Farm tools, machines and food processing technology should be re-designed to cope with, and encourage, increasing biodiversity in agriculture. Public participation (by farmers, grassroot conservation groups and other social formations) is also essential in the various bodies (government, professional boardrooms) responsible for key decisions on overall national research priorities and R&D funding. This is probably the most important challenge facing agricultural research in the next decade. Sources: Chambers, R., Challenging the professions: Frontiers for rural development, Intermediate Technology Publications, London, 1993. MacRae, R.J. et al., “Agricultural science and sustainable agriculture: a review of existing scientific barriers to sustainable food production and potential solutions”, in Biological Agriculture and Horticulture, 6:173-219, 1989. Pimbert, M.P., Designing integrated pest management for sustainable and productive futures, International Institute for Environment and Development, Gatekeepers Series Nº 29, Sussex, 1992. Richards, P., “Farmers also experiment; a neglected intellectual resource in African science”, Discovery and Innovation, 1(1):19-25. Scoones, I., and Thompson, J., Beyond farmer first, Intermediate Technology Publications, London, 1994.