Document(s) 7 of 24

In Asia, hundreds of millions of rural men and women in uplands, semi-arid lands and infertile coastal areas are threatened by poverty and the degradation of the natural resource base on which they depend for their immediate livelihoods. These people are mostly far from urban market centres, far from capital cities and far from the minds and lives of the powerful. The ecological systems on which they depend are being depleted of nutrients, stripped of their biological diversity, and control of the valuable resources that remain is sometimes violently contested.

Yet in order to achieve global development agendas for poverty reduction, it is these people who must be reached. Current practices have failed to either significantly reduce poverty in these marginal areas, or to broadly stem the pace of environmental degradation. The problems are intensified as governments around the region decentralize the management of natural resources (Dupar and Badenoch, 2002). Innovations are needed, but what kind of innovations?

This was the challenge which IDRC faced in 1996, as we prepared a new approach for addressing research on poverty reduction and natural resource sustainability in marginal areas of rural Asia. In consultation with partners from national and international research centres throughout the region, the new research programme, CBNRM, was framed (IDRC, 2000). All the cases in this book arose from work sponsored under this programme.

This chapter reviews the rationale for the research approach adopted in the programme and the conceptual framework which has emerged from our partners applying this approach over the years. While each research project supported through the programme was unique, and there were few explicit links between them, the foundations of the CBNRM programme led to the adoption of a consistent set of tools and approaches. This starting point arose from the experiences of IDRC's professional staff and research partners in Asia, and from reviews of research approaches for rural poverty and environment issues in the mid-1990s. Some of the key intellectual threads of the CBNRM research framework will be introduced here (others are addressed in more depth in the synthesis chapters in Part IV of this volume). But the programme, as conceived by the research donor (IDRC), also evolved through its application by partners in the field into a richer and more complete conceptual framework. This framework for action research in CBNRM underlies most of the case narratives presented. This chapter provides the conceptual background to the cases and helps explain the methodological consistency among them.

Why has research failed to reach the poor?

There are close linkages between problems of rural poverty and natural resource degradation. However, each phenomenon is complex, and linkages are not simple in any causal sense. The ecological and geographic constraints of location are major contributors to the spatial concentration of rural poverty. Indeed, most of the rural poor worldwide are found in those least favoured areas where natural and human factors combine to constrain agricultural production and market access (Pender, Hazell and Garrett, 2001).

Classic approaches to rural development research invested in improved technologies to increase production in various sectors: forestry, agriculture, fisheries, livestock or irrigation. IDRC research support helped to build national scientific capacity in Asia in these fields throughout the 1970s and 1980s. Such research had the advantage of being able to connect readily to professionals, academics and government staff who were organized sectorally and trained in different applied science disciplines. This kind of work remains essential to broaden the range of responses available for farmers and to increase yields of staple grains. But it has not been successful in helping poor, marginal farmers to improve their conditions; nor has it been effective in addressing the fundamental causes of resource degradation (Chambers, 1997; Sayer and Campbell, 2004).

The green revolution packages of improved crop varieties and inputs that led to large increases in grain production in Asia did not reach such marginal areas. High-yielding varieties (HYVs) are designed to respond predictably to commercial inputs which standardize the production environment through irrigation, fertilizer and pesticide application. If managed carefully, these can be very effective in productive agro-ecosystems where conditions are broadly consistent and where access to input and product markets is assured.

However, in less favoured areas ecological conditions are heterogeneous. Soils, slopes, altitude and other microclimatic factors such as water availability, quality and accessibility are subject to wide variation over small areas. Farmers' practices are diverse and spread risk across a wide variety of livelihood strategies that are reliant on multiple resources. Standardized technical solutions do not work (Chambers, Pacey and Thrupp, 1989). Researchers and research donors alike could see that while scientific advances could be made in test plots, these were not benefiting the poorest farmers whose fields were much more diverse – if they had fields at all.

Innovations in integrated agriculture and forestry research and in watershed management provided models for joint investigation and management of agro­ecological systems, which spanned the boundaries of conventional, reductionist science in each of these related fields. Farming systems research, which extended the conventional commodity production approach of agricultural research, was one of these. This research approach recognized that poor smallholder farmers could afford neither the inputs nor the risks of HYV monocropping, and instead explored the productivity potential of linking intensive management of trees, staple crops, cover and conditioning crops, livestock and sometimes aquaculture. These joint systems benefited from careful agro-ecological studies to take advantage of nutrient chains, ecological complementarity and soil-replenishing features to strengthen resource productivity under diverse conditions. Low-input agriculture research helped to improve understanding of pest and soil management to enable farmers to improve resource quality and sustainability by reducing chemical use. Agroforestry research helped to validate and expand the range of livelihood options for farmers in forests and uplands by developing more productive tree crops, as well as fodder and tree-crop intercropping. Integrated watershed management demonstrated the links between land use practices and hydrology in conditioning both surface and groundwater supply and quality. The progress of these research efforts extended the ways that agricultural scientists and foresters viewed their roles in resource management. They required greater interaction between researchers and farmers to understand and test existing or novel management systems. The experiences from such work not only expanded the repertoire of applied science research methods, but provided a range of promising options for improving smallholders' livelihoods. These were helpful precedents for CBNRM researchers.

An even more integrated research focus on production systems, or on biophysical and ecological constraints to production, failed to address the problems of access by the poor to key natural resource assets. This is most obvious in the case of arable land, which has led to numerous efforts at land reform and land reallocation throughout Asia. However, it is also true in the case of access to pasture, forests, fisheries or water. These resources, essential to the livelihoods of poor farmers in less favoured areas, cannot easily be allocated to and managed by private households. Utilization decisions by one set of users affect the quantity or quality of resource available for other users of the same resource. Therefore, collective institutions are needed to manage these resources (Knox, Meinzen-Dick and Hazell, 1998). But this posed problems for resource management research. Little was known about such institutions and how to investigate them. Frequently they did not exist, or if they did exist, they were not recognized by governments as capable of management action and hence were not seen as targets for innovation.

The issue of resource tenure also arose in relation to the commercialization of resources. As private commercial interest in rural resources grew, for instance due to industrial plantation crops, large-scale logging, hydroelectric power development and agricultural colonization, the poor found that they could no longer control their local resource base. Both management by central government and private enclosure, which is mainly concerned with maximizing returns from the resource, neglected the role of common property resources on which the poor frequently depend. Commercialization of these common property resources led to changes in the de facto rights of poor local resource users. This excluded and further impoverished them (Dove, 1993; de Koninck, 1997; Beck and Nesmith, 2001).

More generally, natural scientists focusing on biophysical resources neglected institutional linkages to processes of impoverishment. Yet it is precisely the institutions of resource management which must be addressed if targeted poverty reduction measures are to succeed in resource-dependent communities (Béné, 2003). In the face of these persistent difficulties in how to apply research and innovation to the problems of poverty and environmental degradation, the role of scientific specialists in proposing solutions became increasingly discredited. They were proved to be frequently wrong, or had limited impact on actual practice (Chambers 1997; Sayer and Campbell, 2004).

Another part of the problem is that the sectoral organization of science, extension and policy does not match the perspective of the poor farmers themselves. They are obliged to adopt multiple livelihood and subsistence strategies, relying on agriculture, livestock, fish, forest collection, wage labour and a range of other resource utilization strategies, as well as increasingly off-farm labour and remittances. Therefore, technical improvements in one particular production technology may or may not fit within the diverse practices and adaptation strategies of farmers in marginal areas. An example of this relationship would be the seasonal migration of male members of the household, where the ensuing labour scarcity may constrain potential new farming strategies.

For some time, critical scholars of rural development and agricultural research have pointed to the need for a much greater emphasis on the perspectives of poor local farmers. They need to be considered as development actors themselves, with very different perceptions and motivations from external interveners, but who nonetheless have significant, albeit unrecognized, capacities (Chambers, 1983; Scott, 1985; Biggs and Farrington, 1991; Beck, 1994). There was also increasing recognition of the value of indigenous (or traditional) knowledge for informing scientific understanding of agriculture and natural resources and for guiding effective local interventions in environmental management (Berkes, 1993). There is now plenty of evidence in Asia of indigenous practices adapting to constrained environments to improve their productivity (Cairns, 2005).

These emerging lessons from research in applied sciences and rural development pointed towards the need for a new research approach for poverty reduction and natural resource management (see Figure 2.1). They also coincided with other factors that shaped the programming of many international development agencies in the 1990s. The UN Conference on Environment and Development in 1992 adopted the Agenda 21 action programme, responding to the call for sustainable development contained in the report of the World Commission on Environment and Development (Brundtland, 1987; UN Dept of Economic and Social Development, 1992). This high-profile international commitment strongly linked poverty and environment issues, placing them firmly on the development agenda. At the same time, the dominance of the 'Washington consensus' drove development policymaking to reduce the role of the state and emphasize the potential of smallholder farmers to generate economic growth by responding to liberalized markets. Rural development practice came to be strongly influenced by nongovernmental organizations (NGOs), rather than government or international organizations. The rhetoric of participatory development was widely adopted by most organizations engaged in rural development (Ellis and Biggs, 2001), yet rarely tested critically. All of these factors pointed to the need for a programme of applied research which could explore integrated approaches to natural resource management from the perspective of resource users.

Figure 2.1 Paths to CBNRM research

Defining CBNRM research

While it is important to have some clarity of basic concepts, the intention from the outset of this programme was that research partners should explore the meaning of CBNRM in practice for themselves. The programme resisted trying to define too specifically what ought to constitute community-based natural resource management. Throughout the 1990s this term came to be popularized in the rhetoric of donors, NGOs and development agencies, and was critically addressed by academics. It was applied to a very broad range of approaches and practices. IDRC's CBNRM research programme started from a set of principles which distinguished its work from that of other researchers and practitioners, and which responded to the particular concerns identified above.

The foundation of the research programme was its focus on poor people and on strengthening their livelihoods. Enquiry was oriented to natural resources, but from the outset the goal of the work was aimed at improving the conditions of poor men and women, where the quality of the resource base was a prime element in their well-being. In this respect, the approach of CBNRM departed from one of the antecedents to this research programme, that of community-based conservation.

In the late 1980s, large international conservation organizations began to work closely with local organizations and communities to support the creation of protected areas and strengthen the conservation of endangered ecosystems. These efforts led to many community-based integrated conservation and development programmes, which attempted to provide local benefits through wildlife and ecosystem protection programmes (Wells and Hannah, 1992). But recent critiques have emphasized the fundamentally co-optive nature of conservation initiatives with indigenous and local communities when goals are defined primarily by governments or external agents (Chapin, 2004). CBNRM took a different starting point, emphasizing that in principle, resource management objectives ought to be locally grounded.

A premise of CBNRM research was that the traditional and local knowledge of resource users deserved to be valued and treated with respect. The conventional professional training of natural and social scientists has led them to dismiss traditional knowledge, especially when it is not formally codified or documented. Yet traditional or local knowledge, in the context of its specific applications and relevant scientific validation, can make important contributions to agriculture and natural resource management (Chambers, Pacey and Thrupp, 1989; DeWalt, 1994). As a point of departure for research, the crucial considerations here were about the meaning of knowledge and the attitudes of researchers more than narrow methodological concerns. We recognized that knowledge has value both for its intrinsic meaning in a particular social context and also for what people can do with it. These different values and meanings of knowledge come from the social context in which knowledge is created and used. Therefore, meanings can be different in local and scientific contexts where knowledge is created and used under different criteria (Berger and Luckman, 1966; Kuhn, 1962). For problem-oriented researchers, it becomes important not just to catalogue and categorize different types of knowledge, but to build understanding and interaction between them. To begin with, this meant that researchers had to find ways to identify and assess knowledge in different forms and from various sources (Grenier, 1998).

In addition to local knowledge, the CBNRM research programme presumed that resource users also had rights to access resources essential to their livelihoods. These rights are typically complex, and can include overlapping customary and legal tenure rights, rights to different kinds of resources at different seasons, and rights which are recognized by different agents under different circumstances. These rights can be held by individuals, family groups or communities defined in various ways. They can be exclusive or shared, sometimes depending on context. Formal or informal rights might only be translated into practical resource access and use under certain conditions. As natural resources in marginal areas come under more pressure from competing users and from degradation, overlapping rights and tenures are increasingly contested (Vandergeest, 1997; Leach, Mearns and Scoones, 1999). Lack of formal rights, conflict over rights, or loss of longstanding resource rights all reduce the incentive for users to invest in managing the resource base and lead to degradation. Understanding these rights and the institutions through which they are contested was taken to be an important prerequisite for effective intervention and change.

The notion of introducing change was fundamental to CBNRM research. The research was intended to yield practical short-term benefits for resource users through improved natural resource management. This was not to be merely an exercise in analysis and theory- or model-building. The learning was to produce concrete, implementable and sustainable innovations to benefit local farmers and fishers. Depending on the problems and context, these were often linked to technical efforts to improve resource productivity, such as through introducing new agricultural, agroforestry, aquaculture, livestock or integrated techniques. But they could also include designing and introducing new institutions which might be required to resolve crucial conflicts or secure collective tenure. The research was premised on the notion that in these marginal agricultural contexts, where common property resources were essential to the livelihoods of the poor, both technological and institutional innovations would ultimately be needed. Hence, the question of where to start was largely a pragmatic one. In order for innovations to be adopted and adapted by resource users themselves, they had to be practical, sensible and understandable. They had to be tested and validated by the users, who were male and female farmers and fishers. In addition, they had to be endorsed by the best available knowledge of the researchers themselves.

CBNRM research was also intended to recognize from the outset that communities are ill-defined and heterogeneous. Interests diverge, wealth and power separate, social relations are complex and dynamic, and history matters. Differences of culture, ethnicity and language can make these characteristics opaque to outsiders, whether from the other side of the river or the other side of the planet. Researchers were encouraged not to romanticize the community and its potential for natural resource management, and to start from where local people were, rather than from any idealized notion of appropriate practices. Local situations are idiosyncratic and easily misinterpreted by outsiders, especially those with a predetermined agenda.

Applying innovations in the field meant recognizing that local change and development requires the agency of local men and women, who are capable of responsible and creative choices to improve their own circumstances. It was also crucial to obtain the support of local government, which can – and often must – endorse innovative resource management practices. For example, many of the research activities dealt with institutional innovations such as tenure and resource access, as well as individual and collective management interventions for resources which were not privately owned. Therefore, any research effort that expected to generate usable innovations had to address the concerns of local people and governments. The best way to do that was to ensure that these actors were fully engaged in the research at its outset, and that their voices were influential in directing the enquiry.

The research approach described above has several important implications, which are significant enough to be considered as underlying design principles for the research programme. First, and most important, the CBNRM research programme was premised on meaningful participation by local men and women in the research effort. The participatory nature of the research was intrinsically linked to recognition of poor farmers and fishers as crucial agents of change and development, and not merely targets of technological advice. Strengthening and empowering their actions required participatory approaches. The emphasis on indigenous knowledge, on understanding institutions and on practical and sustainable interventions all made it imperative for the researchers and the local people to be jointly engaged in learning. Lessons would have to be convincing and evidence credible for poor farmers or fishers to risk investing in innovations. The best evidence would come from the users themselves.

Second, an interdisciplinary approach to the research was fundamental. The intent was to develop practical and sustainable innovations to address the agro-ecological constraints within the dynamic social and institutional context of resource users' behaviour. Therefore, diagnosis and analysis had to cross disciplinary boundaries. From the outset, it was expected that a wide range of expertise would be called on in each research site. But more than that, it was apparent that the necessity of integrating different disciplinary approaches would require methods which used qualitative and quantitative data, analysis and interpretation. These stretched the conventional practices of any single discipline and required the development of new interdisciplinary tools and methods.

The nature of the research task meant that while a range of comparative and methodological research would be supportive, the fundamental test of this CBNRM approach would necessarily involve long-term, site-based fieldwork. The research programme was premised on shared learning from experience on the part of researchers and farmers. This was time-consuming: it involved joint diagnosis, analysis, exploration, intervention and evaluation. In addition, it necessarily focused on outcomes, not on analysis. Without specific reference to outcomes in a series of different contexts, little could be concluded about the framework as an innovative approach to the problem of rural poverty and environmental degradation.

These characteristics of the approach to CBNRM were well represented in the international academic discourse of the mid-1990s. At that time, they were beginning to appear in the work of leading rural development practitioners and on the curricula of graduate programmes in the industrialized world. They were spreading rapidly in international-language journals and international organizations. However, in many parts of Asia, these represented a huge departure from the conceptual frameworks, academic preparation and practices of local scientists and development specialists. In the poorest countries and subnational regions where the CBNRM programme chose to focus its research efforts, most of these concepts were completely new. And there were very few examples anywhere of how to integrate and then implement these concepts coherently in practice. Research teams had to build methods, tools and practical intervention strategies while they were digesting and evaluating the relevance of the concepts.

This would be a challenging research agenda anywhere. How reasonable was it to expect to implement this research approach under the difficult conditions of desperately under-resourced local research organizations in remote parts of the region?

It is important to keep in mind the fundamental capacity-building objectives of the programme. First, it was important to characterize the complex and dynamic problems of poor people's livelihoods and their relationship to local ecological degradation. The point of the programme was to build local expertise in addressing this complexity, among researchers as well other actors who were involved. The programme commenced by incorporating emerging perspectives in the literature related to agricultural research, environmental management and international development. IDRC sought to facilitate access to that literature for researchers in isolated and impoverished regions of Asia,¹ but our approach to CBNRM mainly sought to encourage learning by doing.

Therefore, the CBNRM research programme was based less on a preconceived model or formula for intervention than it was on a collaborative learning agenda. The concepts recognized and embraced the complexity of agro-ecological and social systems, while responding to the imperative of action on immediate problems. The emphasis was on building partial understandings and confidence for intervention, but then on iterative learning from those interventions and moving to longer-term issues. Having absorbed some of the lessons from critiques in the literature, the emphasis of the research programme was on practice more than theory. But given the natural science background of many of our partners, the programme particularly made an effort to draw on social science contributions to natural resource management so as to prepare research partners for this kind of work.

Building blocks of CBNRM research

Natural resource management at the local level involves interventions in the resource base. But there are other factors to consider. These include the embedded rights which different people and groups have to use or to manipulate the resource base, the social relations between them which condition the scope of actions which are possible, and recognition of who holds power and how it can be exercised. Therefore, research into this complex system must draw on a broad range of concepts and models from both social and natural sciences as they are applied to agriculture and natural resource management (Ashby, 2003; Sayer and Campbell, 2004). There is an enormous body of literature on which to draw, and we cannot make any systematic attempt to summarize it here. But most of the researchers who became involved in the CBNRM fieldwork approached it from a prior background in natural sciences, and so conceptual and analytical tools from the social sciences were those most frequently needing both introduction and support.

Participatory approaches are essential starting points for CBNRM research. While public participation is novel for some natural resource researchers in Asia, it has long been recognized as a term applied to diverse practices, ranging from the extraction of data to the full engagement of participants in decision-making (Arnstein, 1969). As a starting point when working with researchers who had little previous experience with participatory methods or tools, the programme emphasized collaborative learning and problem diagnosis with farmers through participatory rural appraisal (PRA) methods (Chambers, Pacey and Thrupp, 1989; IIRR, 1998). Participatory research methods take some getting used to on both sides. Practitioners and local people need to become familiar with these practices and to build expectations when they have little experience of this type. But the intention was that over time, using Arnstein's (1969) analogy of a ladder of participation, which sees token consultation at the bottom and full engagement in decision-making at the top, the role of farmers and local people in the research enterprise would ascend to more meaningful and profound levels of engagement. Research proposals were designed so that problem definition would be responsive to local priorities and knowledge, along with technical analysis. A key role of participatory research was to enhance the capacities of farmers themselves so that they became capable of assessing and articulating their own situations (Nelson and Wright, 1995). This meant that participation was not an exercise in merely generating data for researchers to analyse, but was an ongoing and iterative process of engagement, learning and empowerment.

For researchers, this meant learning that participatory approaches are not merely about technique. Indeed, the techniques themselves are easily learned. What participatory approaches are fundamentally about is agency, that is, a perspective on development which emphasizes the role of individuals and groups in applying knowledge, capacity and action to expand their choices in a particular cultural setting. This is a different perspective from one which sees development as being about structure. That type of development concerns itself with systems, organizations, capital or technology transfers, and the ways these interact to generate aggregate outcomes (Long and van der Ploeg, 1994).

Undertaking participatory research is therefore as much about the adoption of attitudes of humility, respect and shared learning in interaction as it is about specific concepts, methods and tools (Chambers 1995; Pretty and Scoones, 1995). These attitudes are best learned through experience, reflection and practice. In particular, participatory researchers require skills in communications, especially listening skills. Unfortunately, these skills are seldom taught in formal academic or professional programmes. When the objective of development research is social and organizational change, research needs to be structured so that all participants gain from the learning process. This demands recognition of inherent uncertainties in the process and commitment to empowering others, reflecting critically on the process and evaluating outcomes. These were all elements which were encouraged through CBNRM research (Vernooy, Sun Qiu and Xu Jianchu, 2003).

In recent years, there has been considerable critical assessment of participatory research in natural resource management (Leeuwis and Pyburn, 2002; Pound et al., 2003; Sayer and Campbell, 2004). Based on a comparative assessment of over 20 case studies, Vernooy and McDougall (2003) suggest five principles of best practice in participatory research. They must:

• reflect a clear and coherent common agenda among stakeholders and contribute to partnership building;
  
  
• address and integrate the complexities and dynamics of change in human and natural resource systems;
  
  
• apply the triangulation principle of using multiple sources and methods, and link together multiple knowledge worlds;
  
  
• contribute to concerted planning for the future and to social change; and
  
  
• be based on iterative learning and feedback loops where there is two-way sharing of information.

These principles capture well the ideals of engagement, joint learning, collective decision-making and scientific rigour, along with practical action and change, all of which are at the heart of the CBNRM participatory research framework. It is fair to suggest that none of the research projects started with best practices on any of these dimensions. Nonetheless, these principles came to guide the progress which all the projects have followed while conceptualizing and strengthening the participatory dimensions of their research.

In addition to meaningful participatory methods, another foundational building block for CBNRM was the study of resource tenure. The central issues of resource use, control and rights – and, therefore, power – are all encapsulated in the institutions associated with resource tenure. Their importance with respect to the reduction of poverty and natural resource degradation has been discussed above. The central tenet of common property resource theory is that it is not only possible, but empirically frequent, that collective institutions in rural societies are designed to manage resources which would otherwise be degraded by the pursuit of individual utility (Berkes, 1989; Ostrom, 1990).

However, while forests, irrigation systems, fisheries and pastures all appear to be amenable to community management, the devil is in the details. Rights to access and use resources are always linked to power relations. They are inherently bound to shared concepts which give meaning to resource use, as well as the social relations within which these are constituted. Hence, common property research inherently involves not only the study of resource rights and institutions, but also the ways in which they are moderated by culture, social organization, political power and resource values. All of these factors are dynamic and respond to both internal and external change. In the context of natural resource management diagnosis and intervention, it is possible to adopt a prescriptive approach, where you first analyse and then recommend conditions for successful community management of common property resources. Or you can take a historical approach, analysing the ways in which resource tenure, social relations and power have evolved in any given community to generate the contextual pattern in evidence (Johnson, 2004). For the purposes of CBNRM understanding and intervention, each of these research approaches offers value. Choices could depend on context and on strategic objectives.

Resource tenure studies quickly get to the issue of control, which is inherent in the whole notion of CBNRM. Practical measures to redress environmental degradation and strengthen livelihoods require secure common property tenures to respond to the decentralization of authority for resource management or the failure of state management systems. Although they were once typical of traditional rural societies, such tenure systems have been eroded by the state or by commercialization, or are no longer able to deal effectively with the new burdens of conflict and multiple demands on the resource base. The challenges of introducing new forms of collective tenure, or of managing the private enclosure of common property in an equitable fashion, were important issues on the research agenda of CBNRM projects.

These foundational elements of participatory methods and institutional analysis pointed to an area of particular challenge for research partners: the nature of social differentiation and exclusion. Both traditional and modern institutions for resource tenure and management are rooted in social and power relations, and hence in the nature of social and political difference. Such differentiation is structured and expressed differently from one society to another, and may involve a wide range of social characteristics, including wealth, ethnicity, religion and caste. However, a crucial consideration is the role of gender. It was impossible for researchers to structure their enquiry collaboratively with resource users, or to pursue questions of the social construction of rights and power, without confronting gender as a differentiating factor in participation and power with respect to resource management (Agarwal, 2001). In many cases, this was an issue which research partners had not previously considered. This was not only a matter of gender blindness, but often of avoidance. Researchers who had plenty of field experience or extension training usually were sensitive to the general issues of power, social relations and participation in learning among rural communities. But they found gender issues difficult to address. They often lacked tools for formulating concepts, collecting or analysing data, or even communicating sensitively with both men and women about gender differentiation in natural resource decision-making and utilization. This was an area which, although flagged early on for attention in the research programme, remains challenging (see Beck and Fajber, chapter 15).

None of these building block elements were new. But there were several novel features in a research programme based on such foundations. First was the emphasis on introducing and building meaningful participation of local men and women in all components of research, from problem definition to interpretation and assessment of outcomes. This was not a matter of sacrificing rigour in the research approach, but of ensuring it. The rigour demanded of this research programme was that of effective practice. Results had to stand the test of farmers' fields and practical village life. There is great value in research that expands the frontiers of conceptual knowledge, builds theories for wider exploration and challenges conventional disciplinary wisdom through critical analysis. However, none of these goals were primary objectives of this research programme, which focused on practical outcomes and local change.

By adopting an action-oriented approach, the research programme intended to cross the divide between applied research and practice. Its emphasis on learning by doing as a way to build research capacity using challenging new concepts allowed researchers as practitioners to test theory against local outcomes. By also engaging local resource users as learners, the research was meant to develop models of practice for researchers and professionals, practice which would simultaneously build capacities for local organization, action and continued learning among poor farmers and fishers.

This was also an explicit attempt to combine perspectives on social and natural science problems through an integrated mode of enquiry. It was not merely a matter of building research teams of diverse specializations, but of structuring the programme of research to address issues which lie between and across the disciplinary specializations (see Figure 2.2).

Finally, the research programme was meant to engage governments as well as local resource users. Research projects were structured to involve local government officials in diagnosis of problems as well as in interpreting results. From the outset, policy innovations were identified as valuable targets for the research effort. Taken together, these were the characteristics which set the stage for the cases which are presented in subsequent chapters. But these programming ideals also evolved as research projects were implemented by teams in a dozen different countries in Asia.

Figure 2.2 Building blocks of CBNRM

A practical framework for action research

A research framework arose not only from these objectives, from the adoption of concepts and models in several disciplines, and from guidelines for implementation which emphasized the characteristics described above. It also matured through the fieldwork undertaken by research partners, including those not represented in this book. Their struggles with practical and conceptual issues helped to refine the initial model, sometimes by simplification and sometimes by elaboration.

One of the first lessons from employing participatory methods in the field was that, not unreasonably, it generated expectations of assistance among local people. Any approach to poor communities by outside expertise is structured in a context of differential power and wealth which itself creates initial expectations, most of which are unhelpful to communication and shared understanding. From their extensive experience, poor communities are generally suspicious of the motives of external interventions. They may have little time, patience or shared vocabulary for explaining to outsiders complex resource, agricultural and social systems which to them are self-evident. Therefore, building positive relationships with local men and women in poor communities meant the research teams had to develop better communications skills. They learned that to build trust they had to set aside preconceived research agendas and start with practical responses to locally perceived problems.

Sometimes this led in directions where the researchers could do little of value. For example, in several of the field sites profiled in this book, villagers were concerned about inadequate local facilities for primary education. The CBNRM research teams were not equipped to address this problem, but could mediate between local communities and NGOs, or government agencies with resources that could be leveraged for this purpose. The research team identified local needs to external agencies, helped local leaders to apply or to meet external criteria, and helped communities to mobilize their own resources. The researchers adopted facilitation roles for local development inputs, which also built trust and shared commitment to their CBNRM research efforts.

This process worked both ways. As the community and its leaders gained confidence in the researchers and their potential to contribute to local goals, the researchers themselves gained greater respect for the strategic capabilities of local people and their leaders. Nonetheless, these processes take time. Expectations on both sides have to be moderate. Building trust and participatory skills is not easily accomplished to a predetermined deadline or a fixed work plan. Project planning had to be flexible and responsive. In the course of implementing the CBNRM research programme, IDRC learned it also had to accommodate reasonable adjustments to project timing and costs.

Another factor which became evident as the research gained momentum in the field was conflict over resource use. Natural resource conflicts are inherent to human valuation and exploitation of resources, and are found everywhere. But these conflicts have become more widespread and more intense as resource demands multiply (Buckles, 1999). Research teams almost always found unanticipated and sometimes complex conflict situations when they diagnosed resource access rights and use in the field. In many cases, they were not well-prepared to address questions of resource conflict. Having established the inequities and abuses of power which frequently lay beneath these conflicts through participatory social and historical research, researchers were no longer able to rely on narrowly legalistic or officially sanctioned solutions. They typically sought to introduce new processes for conflict management, or create new win-win options which would fulfil the mandates of state control yet increase benefits for poor local resource users. This was an important new area of research and practice for the researchers. It required training and the elaboration of new tools and methods. It also required special skills in communication and negotiation, or the ability to foster such skills in others. Developing tools to respond to resource conflicts became an important part of the CBNRM framework, as the research teams repeatedly recognized the frequency of such problems in the field, coupled with the inadequacy of existing institutional mechanisms to address them.

As the researchers gained confidence and experience, and as they were able to demonstrate local successes in addressing natural resource degradation, strengthening local resource rights and improving livelihoods, it became evident to them that these successes needed to be extended to other situations. In many cases, the researchers' first attempts to extend this learning were in response to requests from communities near their field sites, who had seen the improvement in their neighbours' practices and conditions, and wanted to adopt similar measures themselves. This was sometimes a difficult request to address. Researchers are often wary of early positive results and (in good scientific fashion) prefer to wait for validation and repetition before systematic extension. Usually, the teams had limited resources to respond to these requests. But as they became more confident with the methods, and as results were seen to be robust, the researchers recognized that they had to develop better ways to share and extend their lessons. This led to expanding the scale of their innovations.

The research teams approached scaling in two ways: they scaled out especially using farmer-to-farmer and community-to-community extension approaches to reach neighbouring user groups and local government units adjacent to their field sites. In addition, they scaled up by engaging senior government agencies in the elaboration of the methods, and in their replication and adaptation to new conditions. Both scaling out and scaling up research innovations became part of the CBNRM lexicon as the programme evolved, and the reader will find examples of these approaches in the cases.

After gaining experience in the field, and through interaction with local government units, research teams came into greater contact with the policy constraints to local CBNRM action. They learned that local resource degradation was an outcome of complex failures of rights, of conflicts, of power disparities and of inappropriate production systems. But when they tried to design interventions to address the problems they found, they commonly ran into constraints imposed by existing natural resource tenure or management policies. Perhaps there were no opportunities for collective local management of common pool resources. Sometimes there were official systems for community resource management, but these were not implementable due to onerous administrative and technical requirements. Occasionally, the government was one of the parties exacerbating resource conflicts through inept or inadequate implementation of well-intentioned resource policies.

These experiences led most of the CBNRM research teams to explore avenues for influencing natural resource management policies in order to ensure that field-level CBNRM innovations could be put in place. For many of the researchers, venturing into the realm of policy decision-making was another new experience. Policymaking in these countries is far from transparent, and in some cases the mechanisms of policy development were neither obvious nor accessible to the researchers. However, most of the research teams were able to find ways of influencing policy after devoting effort to networking and learning (see Tyler and Mallee, Chapter 17).

CBNRM researchers have come to interpret this general conceptual framework as one of participatory action research (PAR). In this context, the action is not necessarily political action (in association with a particular social movement, for example), but action in the sense of local innovation and change. The main lessons from the research are those which local people can take away and apply themselves to improve their economic welfare, their political power, their social status and the quality of the resource base on which they depend. Learning in this conceptual framework becomes a social process, engaging multiple actors with different interests.

The researchers themselves are only one of the learning groups in CBNRM research. The research process is an iterative one, where scientific and local knowledge are both applied critically to diagnose problems and design interventions. The crucial element of a formalized PAR process is the assessment of interventions by the various actors involved. Were they effective in addressing the problems? Did they generate unexpected consequences? Who benefited? Has resource quality improved? In order to reach some conclusions about what has been learned, actors need to reflect critically on what was being attempted, on the data they have to assess it, and on the surprises which arose along the way. The context is always dynamic: resources, demographics, markets, power relations and policies change with time and require adjustments and adaptations at the local level. Research teams have developed participatory monitoring and evaluation tools to strengthen this learning process and ensure that various actors formalize their learning (Vernooy, Sun Qiu and Xu Jianchu, 2003). Table 2.1 summarizes the main ways that the CBNRM research framework was adapted and elaborated by research practice.

Table 2.1 Building a practical framework in the field

Applying the framework: case studies from Asia

The CBNRM research framework applied in the cases in this book emerged from a concern with the failures of applied research to reach the rural poor in developing countries. It was based on principles which tried to keep poor men and women central to the framing of research problems, the implementation of projects and the learning which resulted. The framework evolved through application, and was flexible enough to look a bit different in the various site-based studies, in response to the political, agro-ecological, and social contexts, as well as the capacities of the various players.

The collection of cases which follows is representative of the long-term site-based research projects which were the core of the CBNRM research programme over the period 1997–2004. Their methodological commonalities can be attributed to the framework described above, with principles and concepts articulated jointly by the researchers and IDRC, and then elaborated by the researchers themselves during implementation.

The poverty and natural resource contexts in many of these locales were not well-documented in the mid-1990s, in terms of political, institutional, governance and ecological characteristics. Therefore, there was a very limited amount of international literature which might provide relevant background or precedents for any of the sites. In most cases, local studies and background data existed, but these were either of dubious accuracy, or were premised on more traditional agricultural research which was of limited value for CBNRM.

In this context, at the outset most of the projects evolved in unexpected ways. Everyone was in learning mode. Objectives shifted, sites changed, practical difficulties were either overcome or else derailed the efforts. Local researchers also had to deal with the normal challenges of working in any remote part of a developing country. An array of challenges presented themselves: flooding of project offices and computers, unreliable logistics, changing organizational structures and intervening opportunities, and generally low administrative predictability.

At every step, research teams created their own precedents in a field which was new to them, using concepts and methods which were not widely recognized by their peers. This required courage. There was no practical option to learning by doing, and the research teams tried to integrate new processes of collaboration, reflection and integration as well as analysis. The learning process was not smooth, and rough spots remain. But the experiences of implementing and working with this conceptual framework for CBNRM research have generated many insights into practice, as represented in the cases which follow.

Document(s) 7 of 24