effectively organize in order to identify community priorities and address local problems, and work in partnership with local governments and other institutions.

5.2.1.3 Agricultural extension and capacity-building opportunities
Although rural communities in SSA have a long history of self-help and community development, top-down approaches to the development and dissemination of AKST have traditionally been the norm. As such, rural communities typically have not been empowered with resources and decision-making authority, and the voices of socially excluded groups such as women and minorities are often not heard. Typically, extension organizations in the region have involved overlapping responsibilities and uncoordinated interventions between several public agencies and NGOs, with extension workers often lacking minimum means, such as vehicles, fuel and materials to fulfill their roles. In many SSA countries the linear approach of a centralized scientific organization transferring technologies down to extension agents and on to the farmers (reinforced by education systems that train scientists specifically to work in such institutions) has worked relatively well for major cash crops. However, this system has had little success for improving subsistence and food production (Hall and Nahdy, 1999). The typical linear approaches to extension that have been employed in SSA lack feedback loops from farmers to researchers and value “scientific” research and learning over more tacit forms of farmer learning and local and traditional knowledge (Ochieng, 2007).

 Participatory Demonstration and Training Extension System (PADETES) has been the national extension system of Ethiopia. Developed after a critical evaluation of the past extension approaches practiced in Ethiopia, this system accommodates present thinking in extension philosophy including research, education and extension as part of the knowledge system. PADETES puts equal emphasis on human resource development and the transfer of appropriate and proven technologies. Implementing extension services is entirely the responsibility of the Regional Agricultural Bureaus, while the Federal Ministry of Agriculture has the mandate to formulate and submit agricultural and related policies and, upon approval, coordinate and disseminate them through interregional development programs and/or projects and provide technical advice and training services
to the extension staff of the Regional Agricultural Bureaus (Ejigu, 1999).

 A number of approaches already exist to train farmers in research and extension. Farmer field schools (FFS) employ a pedagogical approach of “learning by doing” or “interactive learning” (Ochieng, 2007) that can improve
farmers’ knowledge, skills and sense of empowerment. Farmer field schools also allow local and traditional knowledge to be incorporated into the development of new approaches. Farmer field schools, combined with efforts to generate demand, have been successful in establishing producer and consumer markets for vitamin A enriched orange-fleshed sweet potato in east and southern Africa (Ochieng, 2007). Farmer field school shortcomings include relatively high investment costs; expensive to sustain and to

replicate; and they tend to exclude poorer farmers (Davis, 2006).

 Farmer field schools suffer from the same problem as other forms of public extension, namely they require sustained funding. In Kenya, extension-led farmer field schools can cost up to $600 per group of 25-30 farmers whereas farmer-led schools cost half of that (Onduru et al., 2002). Once grants from the International Fund for Agricultural Development Integrated Production and Pest Management (IFAD-IPPM) are depleted, these FFS are likely to cease unless local self-financing initiatives are identified (Onduru et al., 2002). Given the reported large increases in yields,
there may be potential for FFS to be self-financed by the farmer groups themselves, as has happened in other areas in Kenya.

 Lessons from FFS can be documented in relatively simple extension messages (Onduru et al., 2002). In Uganda, there has been a move to decentralize extension services and to encourage a plurality of providers and approaches. Particularly important is that extension services are being designed to be more directly responsive to farmers’ selfidentified needs.

 New approaches to extension that are more responsive to farmers, less top down and more integrated with research will require extension agents to have different skills from those they currently have and that are traditionally available. One option is to introduce mid-career training and diploma courses, as is being done in Uganda. Fee-based schemes are being introduced in part in response to a decline in public funding of extension services. This approach can expand the provision of extension services, but may exclude the poorest farmers. Increasingly, the private sector is becoming involved in the provision of extension services. Private extension services are typically linked to the provision of inputs such as seeds and fertilizer and the purchase of agricultural products.

NARS relevance to changing AKST paradigms. In many countries in SSA, most agricultural research is undertaken within the framework of the NARS and so is conditioned by these institutions (Hall and Nahdy, 1999). The adoption of participatory approaches within the NARS framework is hindered by issues of professional identity, lack of participatory research skills, and a professional reward system that makes it difficult to publish the findings from participatory research in the top academic journals (Hall and Nahdy, 1999). Extension tends to rely only on countries’ official languages as working languages. Though not yet proven, moving the use of selected SSA languages up the researchextension chain could have a significant impact on participation, relevance and results.

 There are a number of processes currently working to improve the relevance of the NARS. The Innovation Systems Framework and Integrated Agriculture Research for Development are highlighted below. An innovation system can be defined as networks of organizations or actors, and the institutions and policies that affect their innovative behavior and performance that bring about new products, new processes and new forms of organization into economic use (Hall et al., 2006). As an evolutionary model, the focus