pluralistic extension systems, such as that provided by pri­vate input suppliers, NGOs, farmer associations, agropro-cessing companies and private consultants.

NGO-led approach. NGOs have long articulated the needs of small farmers and other socioeconomically vulnerable groups. They have advocated more equitable and sustain­able economic development and poverty alleviation pro­grams in Bangladesh, India, Nepal, the Philippines, Thai­land and Viet Nam. The range of their activities has varied: agroforestry in Nepal, tea production and vaccine research on cattle disease in India, soil and water conservation tech­niques in the Philippines. The extent of NGO inputs in re­search and extension, especially in technology adaptation and dissemination, has been quite large, as has been the magnitude of their organizational network.
     In most countries, the relationship between NGOs and the government has often been adversarial rather than co­operative. But, whenever cooperation is possible, results can be extremely fruitful. In India, the government has taken concrete steps to establish close ties with NGOs. The Indian Council for Agricultural Research (ICAR) set up farm sci­ence centers to serve as centers for demonstration and train­ing in "scientific farming" to open NGO access to the public research system (Sulaiman and Hall, 2005).
     The central role of NGOs and farmer organizations in reducing poverty has chiefly focused on building social capital, catalyzing entrepreneurship and disseminating pub­lic information. NGOs rely on the concept of participatory research, where all stakeholders play a role in setting and implementing the research agenda. Initial attempts at con­ducting participatory research gave greater priority to involv­ing poor people in evaluating new technology, rather than in setting priorities for the research (Hazell and Haddad, 2001). Their involvement in diagnosing problems and field testing technologies has provided national researchers with useful information, resulting in more useful products for farmers (Hazell and Haddad, 2001).
     Participatory  research  promotes   organizational   and skill-building capacity to communities to help solve col­lective problems and resolve conflicts. However, it is con­strained by the need for multidisciplinary teams willing to work together, respect and value each other's knowledge and appreciate the high initial cost of many personal interactions among team members. In a project to develop pest control measures in Ghana, farmer participation costs increased by 66% and accounted for 80% of researchers' time, although this might lead to higher returns in reducing time needed to identify promising technology (Hazell and Haddad, 2001). Also, developing the farmers' own capabilities in developing improved pest management systems, conducting field trials or breeding could be cost effective in adapting technology to diverse local needs. Thus there are no cost tradeoffs where participation might mean the difference between success and failure in technology development.

Private sector-led approach. Extension services can be com­pletely privatized (Siamwalla, 2001). New Zealand is the only ESAP country with fully privatized extension. Farmers and extension agents sign profit- and risk-sharing contracts. The extension agent serves as a consultant, selling services

to farmers for a fee. Thus, consultants are important sources of information and advice to large commercial growers and are valued for customizing advice to individual farms. They also provide expert counsel to international development agencies because they have a collective memory of what worked in formulating new initiatives. There is, however, the risk that employing the same consultants and advisers will lead to adopting old recommendations, some of which have failed in the past.
     Technology transfer by the private sector through the system of contract farming is popular in Thailand and the Philippines. The Thai case, depicting the soybean trader sit­ting astride commodity input and credit markets, is an ex­ample of informal contract farming. More formal contract farming exists. A well-known example is in poultry farming pioneered in Thailand by the Charoen Pokphand Company, a firm that later became a large conglomerate, with agribusi­ness interests in other Asian countries. Charoen Pokphand forged contract growing arrangements with small poultry growers. The arrangements varied from a guaranteed wage contract to a guaranteed price. Charoen Pokphand brought in a hybrid breed from the Arbor Acres Company in the US and set up large automated feed mills, which remained the core of their operations. In Australia, the Grain Grow­ers Association supports the grains industry through direct research and development funding, largely in plant breeding and grain-quality testing. In recent years, the Grain Growers Association has supported research on developing commer­cially viable biological control agents and development of best management practices.
     While these approaches have their strengths, innova­tions in providing extension services should be viewed not only from an institutional perspective but also from a func­tional one. Extension can still occur even without organiza­tions, since imparting knowledge has always been between individuals who trust each other, rather than from an exter­nal agent. Extension was once understood as "extending the knowledge imparted in class to those who cannot attend the class," suggesting that extension is not just the mechanical transfer of technology or information, but also instructive.

Regional and international research and development insti­tutions. Regional and international research organizations have been set up to meet regional and global demands in agricultural research. In 1960, CGIAR set up 15 interna­tional commodity research institutes, a third of which are based in Asia. These international agricultural research cen­ters (IARCs) and many other international programs based elsewhere have done considerable work within ESAP. Aside from making headway in global research on frontier and cutting-edge science, the greatest achievements of these cen­ters have been to encourage open germplasm exchange, sup­port human resource development and training, and create links with the national agricultural research systems.

Traditional, local and indigenous knowledge systems. Tradi­tional knowledge, indigenous knowledge and local knowl­edge are often used interchangeably to refer to the matured and long-standing traditions and practices of regional, in­digenous or local communities, which encompass their wis­dom, knowledge and teachings accumulated through gener-