was industrialized and has been used in India for several millennia (Sheridan, 2005).

Access to and reform of AKST education

A broader set of issues concerns the formal training of scientists and related workforce. As the MDG Task Force 10 has emphasized, higher education is increasingly being recognized as a critical aspect of the development process; at the same time, however, most universities are ill-equipped to meet the challenge. Outdated curricula, under-motivated faculties, poor management and a continuous struggle for funds have undermined the capacity of universities to play their roles as engines of community or regional development (UN Millennium Project, 2005).

     A report by the InterAcademy Council (IAC, 2004) recently underlined the relative decline of the agricultural research and education system in Africa in the past decades. Among the reasons discussed in the report are the relative weakness of science education in African schools, low investment in research in general, and the growth of student numbers (by 8% per year), with funding falling short of this increase and funding decline accentuated by structural adjustment. The report also notes an unexpected renewal phase initiated by a half dozen African universities in the recent past.

     Some MSc and PhD programs in industrialized countries do not always suit the needs of less industrialized countries. The implications both for curriculum revision and access are therefore considerable from an AKST standpoint and will be covered at various points in this report. A positive example is the higher education system in Costa Rica, which is making significant efforts to focus agricultural development on knowledge and technological innovation. It is also important to take into account the gender disparity in training as well as the lack of focus on gender analysis in the curricula of agricultural universities in developing and-most often also in industrialized-countries.

     Besides overcoming shortcomings with regard to quantitative aspects of human and financial resources, it will also be of paramount importance to combine an increase in resource allocation and further capacity development of actors involved in research and extension aimed at a qualitative shift towards more societal modes of knowledge production emphasizing inter- and transdisciplinary approaches (Hurni and Wiesmann, 2004).

     Capacity development is broadly defined here and includes developing (1) common understandings of problems, solutions and ways to approach them, using a variety of interpersonal and intra-social processes; (2) social and cultural resources, not just human resources; (3) multiple, strategic skills across a range of areas to intervene and advocate, not just passive receipt of programs and policies, and (4) institutional and organizational bases of power. If policies for organizational reforms are introduced, medium- to high-level scientific resources are made available for formal higher and tertiary education systems, and organizational change is initiated in the structure of relevant governance procedures, such as those concerned with the management of extension services, funding of R&D, mobilizing of informal inputs from NGO and related bodies, optimizing the use of forchapter

eign technology, and providing procedures for a balanced use of the private sector, deployment of AKST will become far more effective. Indeed, such changes will enable more adequate analysis of agroecosystem services, which is usually not included in production-oriented AKST, and the finding of strategies to mitigate negative impacts ("damages") caused by agricultural practices to such services. Further improvements can be achieved by promoting knowledge of interventions that are environmentally and socially sustainable, including measures to empower women to a much greater degree than has been the case in the past.

Measurement of "knowledge" categories

There is a large gap in research intensity (measured as public R&D investment as a ratio of agricultural GDP) between developing and industrialized countries. In 2000, the intensity ratio for the developing countries as a total averaged 0.53%, compared to 2.36% for the developed countries as a group (Pardey et al., 2006). This intensity gap has increased over the past decades as a result of a much higher growth in agricultural output in developing countries as group than in the developed countries.

     One of the problems in dealing with AKST policy (indeed, KST of all types) is that of measurement-both for "inputs", i.e., investment in AKST, and "outputs", i.e., indicators of resultant knowledge impacts. In the case of the former, a range of proxies are used, the most common being agricultural R&D expenditures in the public sector. Another is the number of persons with PhDs currently working in agricultural R&D organizations. Both are unsatisfactory for the obvious reason that they probably give a distorted picture of knowledge investment. For example, they do not account for external inputs from overseas, which may be higher than the internal inputs. A similar problem exists on the output side since outputs can also take a variety of forms, for instance number of patents, number of new plant varieties registered or number of relevant scientific papers published in refereed journals. Again, all kinds of problems involved in the interpretation of these data are due to paucity of information, lack of disaggregation, variations in national practices, and of course the fact that they often do not pick up on several types of tacit knowledge. It is therefore worth noting that attempts to be quantitative in this area need to be treated with great care.

     Giving local knowledge due recognition means to specifically monitor its integration into the processes of knowledge production at the interface of research and practice. The above indicators must be differentiated more accurately, taking into account the share of research and development expenditures per sector, number of PhDs, and scientific publications, explicitly in relation to the search for new modes of knowledge production that focus on the integration of local forms of knowledge. Indicators must not only allow quantification of resources allocated to local and traditional components of AKST systems. They must also make visible to what degree the resources allocated to these components of an AKST system reflect the overall relationship that local or traditional knowledge and external knowledge actually have in ensuring the livelihood systems of rural people in general and of poor and marginalized people in particular.