NUPI Forum for Development Studies No.2

Dilemmas in Norwegian DevelopmentForum for Assistance Development to IndigenousStudies | No.2-2000 Peoples 205

Dilemmas in Norwegian Development Assistance to Indigenous Peoples A Case-Study from Botswana

Sidsel Saugestad

1. Introduction In September 1999 the Norwegian Ministry of Foreign Affairs presented a document entitled ‘Plan for the follow-up to work with indigenous peoples in development assistance’.1 It is a concise 12- page document that sums up the main areas of Norwegian engagement in international organisations and bilateral assistance, and concludes by setting out a plan for follow-up. The choice of the term ‘follow-up’ may be significant. It indicates that this is not an area where Norway is developing a new policy, but rather a question of consolidating and further development of ongoing commitments. But does Norway have a policy for development assistance to indigenous peoples? If so, where can it be found, and what does it say? A possible answer, implicit in the matter-of-fact nature of the follow-up document, is that a policy can be deduced not only from statements but also from actions. This position will be my point of departure in an effort to unravel some aspects of the policy that now is going to be followed up. The discussion will be on two levels. More generally I will outline Norwegian positions on indigenous issues as stated in international/multilateral arenas (the UN system and elsewhere). These positions, however, do not address development assistance specifically, although clearly of relevance. More specifically I will look at the challenges in targeting indigenous peoples as part of Norwegian development assistance, using Norwe- gian assistance to the Remote Area Development Programme in Botswana as a case in point. This programme was set up to assist

1 ‘Oppfølgingsplan for arbeidet med urfolk i bistanden.’ Author’s translation.

NUPI OCTOBER 00 206 Sidsel Saugestad

poor people living outside established village structures, and the majority of people meeting this description in Botswana belong to the indigenous minority known variously as Bushmen, San, Basarwa or more recently N/oakwe.2 To my knowledge this is the most comprehensive programme undertaken on a state-to-state level for assistance to indigenous populations, although it was not specifically defined as an indigenous peoples programme. I will discuss later some implications of the choice of terminology. In the new context of a ‘follow-up’, it may be useful to examine more closely the achievements and the shortcomings, in short the experiences, of the engagement in Botswana. To anticipate some of my conclusions: If we analyse Norwegian involvement on three levels – multilateral, bilateral and through NGO structures – Nor- wegian engagements appear to have been most successful either in international/multilateral fora, or in the support extended to organisations within the framework of assistance to NGO/voluntary organisations. Support to indigenous peoples on a state-to-state level has turned out to be the most difficult, and this article will discuss possible reasons for this. A recognition of indigenous status implies a recodification and rearrangement of existing political structures. Almost inevitably, a call for such changes is resented by existing power structures (as was also the case in Norway for a long period of time). In the context of development cooperation, backing a movement that calls for fundamental political changes easily comes into conflict with an overarching principle in Norwegian development assistance, namely recipient orientation and respect for the political priorities of the collaborating countries. A recognition of this latent dilemma may help to avoid some of the complications that arose in the Botswana case. As a background to the case study of Botswana, it may be useful to summarise some core aspects of Norwegian international engagement for indigenous peoples.

2. Indigenous Rights and the Saami Background Norwegian development policy can be seen as a combination of development ideologies, i.e., normative expressions of values and desired objectives, and development strategies, i.e., instrumental,

2 The empirical material is taken from Saugestad (1998a). Adaptation for this article is done with a grant from the Norwegian Research Council under the programme Globalisation and Marginalisation.

NUPI OCTOBER 00 Dilemmas in Norwegian Development Assistance to Indigenous Peoples 207 often pragmatic, ideas about how to achieve desired development objectives (Hettne, 1990:135). Two sets of ideologies may be said to have prominence. The first is the consistent emphasis on poverty alleviation, both in terms of assisting the poorest countries, and in trying to reach the poorest sections within these countries. The other ideological basis, which has increased in importance, is a concern for human rights issues. This is reflected in the kinds of causes that Norway takes up or supports in international fora. The priority assigned to democracy and human rights issues reflects a long-term commitment to a humanitarian tradition in international affairs, from Fridtjof Nansen’s humanitarian work to contemporary peace negotiating efforts. A fresh report to Parliament (No. 21, 1999–2000) presents an Action Programme for Human Rights. This is an uncontroversial priority in Norwegian politics, and an important part of the preferred image presented abroad, as part of a Norwegian master-story. Indigenous issues have become part of this wider concern for civil, political and cultural rights, democracy and good governance. An early highlight in this process was Norwegian participation in the UN Sub-commission on Human Rights, widely credited for the increasing acceptance of a ‘maximalist’ interpretation of Article 27 of the UN Covenant on Civil and Political Rights, i.e., an interpretation that opens up for positive measures to protect the culture of minorities, and recognises the need to protect the material base for minority cultures (Eide, 1985). The same thinking is reflected in the ILO Convention No. 169 concerning indigenous and tribal peoples in independent countries (1989). Norway played an active role in the preparation of this convention, and was the first country to ratify it (Proposition to Parliament No.102 (1989–90). Likewise, in the process within the UN system towards the adoption of a Declaration on the Rights of Indigenous Peoples (Wille, 1995), and the establishment of a Per- manent Forum for Indigenous Peoples, Norway has contributed with large delegations and strongly worded commitments. The agenda in this international process has not only focused on the atrocities committed against indigenous peoples in many countries, but also, more constructively, on the rights that ought to be granted in order to strengthen their position. The prevailing emphasis has been both on the need for indigenous peoples to maintain a cultural identity during an inevitable process of modernisation, and on the value to the international community at large that cultural diversity is being preserved. Recently these issues were reiterated

NUPI OCTOBER 00 208 Sidsel Saugestad

by the World Commission on Culture and Development in Our Creative Diversity (1995). In this process, the Saami presence has become gradually more institutionalised. Saami organisations have been a driving force, natio- nally, towards engagements in international debates about and among indigenous peoples, and Saami participation has become part of the image Norway has wanted to present abroad. Saami have been part of official delegations (from 1987 as delegates from the Saami Parliament), and as NGO delegates (from the Saami Council or national organisations) (Brantenberg et al., 1995; Thuen, 1995). The ‘follow-up’ document from the Ministry of Foreign Affairs confirms this priority, by stating that ‘The international engagement of the Saami Council is an important contribution to the work to promote human rights globally’ (paragraph 3.2.2, author’s translation). This international concern has not always been reflected by a similar liberal policy on national issues. A less-than-honourable past history is part of the experience that has shaped the contemporary position. The Saami were a marginalised group both geographical- ly and socially, being subjected to an active assimilation policy up to the 1950s. Even when special policy-measures were formulated to benefit the Saami, they were not recognised as a distinct group, but were identified according to criteria such as language (Saami speaking), domicile (inner Finnmark), or by economic adaptation (rein- deer herding) (Mathiesen, 1987). It took the greatest civil conflict in postwar Norway (The Alta–Kautokeino case in 1981) to establish a political climate where the Saami claims were taken seriously and dealt with through the established system (Thuen, 1995). The two government commissions that subsequently were appointed introduced some significant changes. The 1984 report on the Legal Position of the Saami introduced a change in the Constitution, to recognise that the Kingdom of Norway is inhabited by two peoples, Norwegians and Saami. It also recommended the establishment of a Saami Parliament (NOU 1984, No.18). The Report on Saami Culture and Education proposed a number of steps to promote language development and mother tongue education (NOU 1985, No.14). The strides made by the Saami people were strongly influenced by the international indigenous movement, and also contributed to the same movement. Saami organisations were active in the establishment of the World Council for Indigenous Peoples in 1975 and in subsequent lobbying in ILO and UN fora (Minde, 1996). Nor- way has been involved from the start in the broad international dis-

NUPI OCTOBER 00 Dilemmas in Norwegian Development Assistance to Indigenous Peoples 209 course on indigenous issues, as well as in a debate about its national implications. The emerging Saami organisations have engaged in two directions simultaneously: towards setting an agenda in international fora, and using these achievements as a lever for further changes in national affairs (Eidheim, 1992). A good illustration of the changes that have taken place in prevailing attitudes is the difference between the handling of the two ILO conventions on Indigenous and Tribal Peoples. While Norway never ratified the ILO Convention No. 107 on Indigenous and Trib- al Populations (the forerunner for Convention No. 169), the reasons for non-ratification underwent dramatic changes. When the Nordic states at the end of the 1950s considered whether or not the Saami were covered by the ILO Convention No. 107 from 1957, the decision was negative on the grounds that the Saami were well integrated into society generally, and therefore could not be considered as being ‘semi-tribal’ (NOU 1980, No. 53:20). However, at the meeting in Geneva in June 1989, when the ILO Convention No. 169 was adopted, the justification given by the Norwegian delegate Arne G. Arnesen, who spoke on behalf of the governments of Denmark, Finland, Norway and Sweden, was as follows:

The reason for our non-ratification of Convention No.107 is to be found in its integrationist approach and paternalistic form, which is acceptable neither to our indigenous peoples nor to our Governments. In our countries, we seek to establish a situation of co-operation and mutual respect between governments and indigenous peoples, with self-identification and cultural freedom as keywords (Proposition to Parliament No.102 (1989–90):38).

And he added, as further justification for a revision of Convention No. 107:

More so, because our indigenous peoples themselves have for a number of years been very active in the support of indigenous peoples around the world and in helping to develop standards for indigenous rights in international law. One might say that in this case governments are now following in the footsteps of indigenous peoples in recognising and addressing the needs for such standards (ibid.).

Such was the prevailing official thinking about indigenous peoples by the second half of the 1980s, that is to say at the time when an

NUPI OCTOBER 00 210 Sidsel Saugestad

agreement with Botswana for support to the Remote Area Devel- opment Programme was negotiated and signed. The strong moral, sometimes missionary, commitment to human rights in Norwegian development assistance has been reflected, for instance, in steadfast support to liberation movements in Africa (Mozambique, Zimbabwe, Namibia), and to the anti-apartheid struggle in South Africa. Violation of human rights was the reason for the withdrawal from Uganda and the break with Kenya. The Re- port to Parliament on foreign aid being prepared around that time (No. 51, 1991–92), promised ‘increased emphasis on the rights of ethnic minorities and indigenous people as part of the efforts towards democracy and human rights’ (p.219, author’s translation). Thus, included in NORAD’s overall strategy for rural development and poverty alleviation, the concern for especially disadvantaged sections of a developing country’s population can be said to have followed logically. It seems reasonable, then, to conclude that at the time that Nor- wegian involvement in the Remote Area Development Programme in Botswana was being negotiated, the Norwegian position on what would be the ‘right’ policy towards indigenous peoples had been stated repeatedly, clearly, and publicly in international fora. To unravel the precise linkages, however, between a general concern for indigenous peoples and the specific formulation of priorities with regard to the Remote Area Development Programme, is to some extent a case of interpretation and reconstruction. Out of deference for Botswana’s non-racial policy, the concern for indigenous peoples was communicated in a somewhat ambiguous manner. The official justification of the support to Botswana’s Remote Area De- velopment Programme was to support a programme providing benefits to poor and marginalised people. In other words the programme was categorised as a case of ‘poverty alleviation’. To understand the preference for this label we need to look closer into the particular situation in Botswana.

3. Challenges to Nation-Building and the ‘Bushman’ Problem When Botswana became independent in 1966 it declared itself a non-racial, not a multiracial state. The difference is significant. The Constitution adopted guaranteed the ‘protection and fundamental rights and freedoms of the individual ... whatever his race, place of origin, political opinions, colour, creed or sex’. At that time, with

NUPI OCTOBER 00 Dilemmas in Norwegian Development Assistance to Indigenous Peoples 211 apartheid South Africa and Rhodesia as hostile neighbours, this was a courageous and visionary statement, stressing a policy of equality and equal opportunity. Included in the policy of the new state was an explicit renunciation of tribalism. In the geopolitical climate of the 1960s, tribalism was generally seen as one of the products of colonialism and its divide-and-rule policy, and the newly independent African states considered any reference to sub-national ethnic or cultural differentiation as one of the gravest dan- gers to national unity.3 In Botswana, this rejection of any expression of cultural diversification became a cornerstone in policy formulation, owing much of its strong moral overtones to the wish to demonstrate a contrast to its apartheid neighbours. National unity, however, can hardly be construed without the use of some common cultural emblems and unifying symbols of state (Anderson, 1983; Gellner, 1983). Thus, in order to establish a new political entity and invest it with shared meaning, the new states may find it expedient to elevate those cultural symbols and traditions that are most readily available, and can claim recognition and allegiance from the greatest possible number of people. In this process, national identities are typically made to appear as a matter of doxa rather than of opinion, in Bourdieu’s terminology, i.e., as part of a ‘self-evident and natural order which is taken for granted, which goes without saying and therefore goes unquestioned’ (1977:166). In Botswana it was possible to construct such a ‘self-evident and natural’ order of national unity by elevat- ing the culture, language and traditions of the numerically dominant Tswana tribe, and from the very beginning a de facto hegemonic status of majority Tswana culture was assured. The presence of the Bushmen disrupts this homogenous picture of Botswana. As far back as archaeological records go, they have made up a distinct population in Southern Africa. Their subsistence economy was based on foraging (hunting and gathering); they speak a number of phonetically highly complex Khoesan click-languages; they are somewhat smaller in physical stature and light skinned, and their social organisation is based on small bands united in flexible egalitarian structures, with a leadership ethos based on consensus.

3 Davidson (1992:99) describes how the new nationalists embraced the concept of the nation-state: ‘Striving to transform colonial territories, they would find Africa’s wealth of ethnic cultures both distracting and hard to absorb into their schemes. They would fall back into the colonial mentality of regarding it as “tribalism”, and as such, retrogressive. This diversity, it seems, had to be just another hangover from an unregenerate past.’

NUPI OCTOBER 00 212 Sidsel Saugestad

All this in contrast to the darker Bantu-speaking agro-pastoralists that migrated into the region between 1800 and 800 years ago. Over the years, many of these distinctive features have been blended through annexation and assimilation, but an ethnic boundary persists. Tswana tribes with a stratified system of hereditary chiefs and headmen migrated into present-day Botswana as late as the nineteenth and twentieth century. From an initial period of contact based on exchange, relations gradually took the form of subjugation. Tswana tribal structure allowed for integration of ‘foreigners’ at the bottom of the hierarchy, in a more or less serf-like position. The Bushmen basically had the option to accept this, or to withdraw. For a long time, exit could be practised in dry bush savannah with little attraction for agro-pastoralists. Extensive conflicts of interests regarding land use came to a head as late as the mid-twentieth century, when new borehole technology opened up areas previously inaccessible for cattle ranching. Presently the greater part of the San eke out a living as unpaid or poorly paid labourers on cattle posts, or as squatters near established settlements. British colonialism changed the traditional social structure in Bots- wana less than elsewhere in Africa. The protectorate position meant a large degree of indirect rule, and at Independence in 1966 Tswana tribal dominance, epitomised in ‘eight main tribes’ mentioned in the Constitution, was smoothly transformed into a new political order. Such was the situation when Botswana gained independence. On the one hand, a social stratification where ethnic and economic mechanisms combined to place Bushmen at the bottom of the social ladder according to almost any criteria (Wilmsen, 1989; Good, 1993). On the other hand a national ideology of equality, morally endorsed by the contrast to neighbouring apartheid countries, that negated the de facto cultural differentiation. The new leadership was not blind to the particularly disadvantaged situation of the Bush- men. The problem was that it was not politically acceptable to single them out for special measures on ethnic grounds. The solution was to define the problem as mainly one of poverty, and after some years of probing the Remote Area Development Programme was born.

4. The ‘Rise and Fall’ of Norwegian Involvement In the 1970s, when Botswana became a main recipient of Norwe- gian development assistance, the basic-needs approach and the principle of recipient responsibility were prominent aspects of the policy. The recipient country should be responsible for the planning

NUPI OCTOBER 00 Dilemmas in Norwegian Development Assistance to Indigenous Peoples 213 and implementation of all development activities, while development assistance should strengthen the country’s own ability to promote development and to solve poverty problems. We find all these ideas reflected in the bilateral agreement between Norway and Botswana. The main reasons for making Bots- wana a new partner in 1972 were: (a) that it was among the 25 nations listed by the UN as the least developed countries; and (b) its difficult geopolitical situation as a land-locked country surround- ed by hostile South Africa, Rhodesia and South West Africa (Gran- berg and Parkinson, 1988:vi). It was further assumed that the government was able to create the conditions necessary for Nor- wegian assistance to contribute effectively to the combat of widespread poverty. Sectoral programmes emphasised primary health care, rural welfare, labour-intensive programmes and district development, and included measures for institution-building, in particular technical assistance and training. On all counts, support to the Remote Area Development Pro- gramme fitted the Norwegian paradigm of basic-needs orientation, poverty alleviation and recipient responsibility. In making the decision to support the Remote Area Development (RAD) Programme, Norway formally accepted the definition of the target group as laid down by Botswana. From the very beginning, however, there has been ambiguity in this definition.4 The official Norwegian designation of the programme, BOT 022: Minoritetsgrupper i utkantstrøk (Minority groups in remote areas) is not a direct translation of the title in English – a fact which in itself is rather unusual. It does, however, reflect the dual justifications for Norwegian involvement: that the RAD programme addressed a general development objective of poverty alleviation, but in addition it was seen as a means to reach an especially disadvantaged minority group.

4 The history of the naming of the programme illustrates the problems of conceptualising the type of problem that it was set up to address. It started up as a Bushman Development Programme in 1974, it was renamed the Basarwa De- velopment Programme in 1975 to reflect the official substitute of Basarwa for the term Bushman, which was seen as demeaning. For a brief period in 1976 it was called the Extra-Rural Development Programme to distinguish it from the regular rural development programmes in the more densely populated eastern Botswana, and finally in 1977 it was officially named the Remote Area Devel- opment Programme. In 1988, with the influx of Norwegian donor money, it was again renamed the Accelerated Remote Area Development Programme. After some years, however, the acceleration lost its momentum, and by 1991 it reverted to its former name of the Remote Area Development Programme (RADP). After Norwegian withdrawal it has continued as a conventional low intensity rural development programme.

NUPI OCTOBER 00 214 Sidsel Saugestad

Clearly, the poverty alleviation was more congruent with Bot- swana’s policy than the minority concern. The agreement for programme cooperation was signed between the governments of Botswana and Norway in 1988. The objectives of the programme are briefly presented in the main text, and are expanded upon in an annex to be:

to contribute to the improvement of the living conditions of the Remote Area Dwellers (RADs) of Botswana. This should be done by ensuring that public services are extended to RADs, that their economic opportunities are promoted and that they are ensured political and legal rights. RADs are under this Agreement understood to be people living permanently outside established villages. They will mainly be descendants of ethnic minority groups living under poor conditions in remote rural areas (GOB/GON, 1988).

The main components of the programme have been the provision of clinics or mobile health posts in remote areas, the building of small schools and hostels for primary education, the provision of water supply for new settlements, roadbuilding, and various efforts in the areas of employment promotion and vocational training. These are conventional development measures, of obvious practical benefit. The provision of such services, however, requires a minimum population-size. A pressure towards sedentisation presented new problems for people who have depended on foraging over large tracts of land for their subsistence, and the need to secure an adequate land-base to cater for productive employment has remained an unsolved issue. Moreover, the initial agreement also stated among its objectives ‘the promotion and formal recognition of local institutions and leadership to represent target groups’, and ‘the retainment of RADs cultural integrity’ (GOB/GON, 1988). If we look at these statements as a hierarchy of objectives (using a concept from Logical Framework Analysis), we see that the immediate, tangible output in the form of provision of public services or infrastructure, is meant to contribute towards the more general, intangible development objectives such as ‘recognition of leadership’ and ‘cultural integrity’. Such objectives are often subsumed under the concept ‘empowerment’. Behind these wordings in the initial agreement, the Norwegian concern can be noted. Many Norwegian programme documents

NUPI OCTOBER 00 Dilemmas in Norwegian Development Assistance to Indigenous Peoples 215 express a belief that experience from Norwegian–Saami relations provide some relevant background for assisting the Botswana government in addressing similar challenges. The parallels in the problems encountered are hinted at regularly, and serve as an added justification for involvement. But respect for Botswana’s non-racial policy has prevented explicit statements from the Norwegian side on what the policy implications of a recognition of the Basarwa as a distinct cultural minority might mean. The expectation, which can be read both in and between the lines, was that over time it would be easier to take up this dimension more explicitly. It is a rather sad paradox, then, that at the same time as this minority dimension was first recognised as a topic in Botswana’s national discourse (I would put this point to be the San Conference in 1993, to be mentioned below), the first steps towards Norwegian withdrawal were taken. Overall, the history of Norwegian involvement in the Remote Area Development Programme, its ‘rise and fall’, can be divided into four phases. t The first phase, from late 1988 up to 1991, emphasised the programme’s character as a welfare programme. Towards the end of this period, an evaluation entitled Let Them Talk (Kann et al., 1990) brought attention to significant achievements, but also highlighted the programme’s failure to reach an empowerment objective. t In the second phase, around 1992, some of the inherent ambiguities in the bilateral cooperation came into the open in a series of confrontations that brought Norwegian–Botswana relationships to an all-time low. The controversy was brought out by a new group of actors who started to voice demands: a fledgling indigenous organisation named The First People of the Kalahari. t A third phase brought diplomatic relationships ‘back to normal’, leading up to the time of the Second Regional Conference on Africa’s San Population in October 1993. This event was to a considerable extent initiated and funded by NORAD and SIDA and so far represents the peak of official (Norwegian as well as Botswana) recognition of an indigenous status for San people. t During the final stage, the relationship changed again. From 1994 to 1996, Norwegian bilateral assistance to Botswana was gradually scaled down, and between 1997 and 1998 Norway pulled out altogether. The RAD Programme has reverted to a low intensity continuation of the infrastructure components. Controversy over the Central Kalahari Game Reserve has taken over in terms of attracting national and international attention.

NUPI OCTOBER 00 216 Sidsel Saugestad

In retrospect it may be relevant to ask to what extent the arguments put forward for Norwegian participation, and the measures Norway supported, truly added up to a policy for ‘work with indigenous peoples in development assistance’. To what extent did NORAD involvement actually reflect the concerns that had been voiced so clearly in international fora? While the ideology was clear, we find many ambiguities on the level of implementation. The ambiguity affected three areas particularly: (a) the achievements of development objectives, (b) the learning potential of the ‘Saami parallel’, and (c) the format for withdrawal from the programme.

5. Implications of the Client Model The achievements of the programme have been evaluated a number of times (Gulbrandsen et al., 1986; Hitchcock, 1988; Kann et al., 1990; GOB, n.d; CMI, 1996). From these reports, two main trends emerge:

(1) There are considerable achievements in terms of provision of basic services. The output of brick-and-mortar components such as water points provided, schools and health posts built, are relatively simple to measure, and there is a general consensus on their usefulness. As far as implementation of infrastructure components goes, the programme also has a decent record of efficiency and fiscal accountability (Karlsen, 1992; Economic Consultancies,1994). Other components of the programme, originally designed to stimu- late income-generating activities (an Economic Promotion Fund was established) have had less success. This, however, is not surprising if one considers the extreme marginality of the target group in terms of resources and marketable skills. The lack of economic diversification and growth in the Botswana countryside is a general problem, increasing along with the urban–rural divide.

(2) However, as all evaluations point out, the programme has not to any significant extent achieved such overarching objectives as can be subsumed under the label ‘empowerment’. A comprehensive review in 1990 concluded:

In its sixteen years’ existence the Remote Area Development Programme ... [has] already achieved much, and the services and opportunities now available to many RADs are far better than they were when RADP started in 1974 …

NUPI OCTOBER 00 Dilemmas in Norwegian Development Assistance to Indigenous Peoples 217

However, although there have also been some achievements in less tangible aspects of development, e.g. economic development and employment, the provision of land rights, and education and training beyond a few years at primary school, these achievements in such critically important areas are far less than can be observed in infrastructure development. Much has been achieved – but even more remains to be done ... The five main issues faced by the RADs themselves, and therefore by RADP, are Poverty, Insecurity, Inadequate Edu- cation and Training, Weak Institutions and Leadership, and Nega- tive Public Attitudes (Kann et al., 1990:ix).

The National Development Plan 7 (1991–1997), which was being prepared at that time, also picked up on this concern: t The priority will be shifted from basic infrastructure development (primary education, health and water) to land rights issues, employment, and education opportunities, institution building and leadership training. t A campaign will be launched aimed at changing negative public attitudes towards the target groups, through information on language, culture, lifestyle and traditions (NDP7:390).

Given the considerable resources that went into the programme, how can this limited achievement be explained? Some critics will point to a lack of genuine commitment by the government (there was no institution building or leadership training following the NDP7, and the information campaign promised was never launched).5 My point is a somewhat different one: Even assuming a commitment to the laudable development objectives stated in many programme documents, the crucial question is whether the model set up to achieve these objectives was appropriate. Much has to do with the way the target group was defined. Over the years, the definition of the target group has changed in detail, but has been consistent in being descriptive. RADs are basi-

5 A story that circulated in the Norwegian expatriate community was that a Norwegian Planning Officer seconded by NORAD and given the task of prepar- ing the input from the Ministry of Local Government to the National Develop- ment Plan 7 was frustrated because no policy statement concerning the RAD Programme was forthcoming. Eventually he took some sections from Kann et al. (1990). I have not been able to verify the story, but it is ‘true’ in the sense that it reflects the low priority of the programme within the Ministry of Local Government.

NUPI OCTOBER 00 218 Sidsel Saugestad

cally defined as all citizens who live in small communities outside the traditional village structure, and who tend to be poor tend to rely heavily on hunting and gathering tend to lack livestock tend to have no, or inadequate access to land tend to have no or inadequate access to water tend to be marginalised ecologically tend to be culturally and linguistically distinct, with another language than Setswana as their mother tongue tend to have low level of literacy and little access to formal education tend to have egalitarian political structures tend to be a ‘silent’ sector politically (MLGL, 1989:6–7).

This reads like a catalogue of social problems, which amply justi- fies a social welfare programme and makes it well suited for NORAD assistance. In line with the government’s non-racial policy, the target group is not defined in cultural terms, but according to socio-economic characteristics. The few cultural characteristics that are included appear as part of the problem description: it is a problem not to speak Setswana as a mother tongue, and a problem to have an egalitarian political structure. The emphasis is on what the target group is lacking, in terms of important resources and (efficient) organisation and leadership. Thus, the programme does not only define a situation of scarcity, it also portrays a target group in terms of their perceived shortcomings and defects. For anyone familiar with the general discourse in Botswana, the Basarwa-ness of the programme is meta-communicated by the use of communicative conventions which, by their very nature, need not be stated openly. Basarwa are regularly defined by negations in Botswana discourse: they belong to a primitive past, in contrast to the prosperous economy and culture of the majority; and to the margins of society, in contrast to the rapidly growing centres. The definition provides a detailed description that fits the San/ Basarwa more than any other group. At the same time it is emphati- cally denied that this is a programme targeting the Basarwa particularly. Inevitably, the programme has come to mean different things to different people. To illustrate: on occasions where ‘community representatives’ are invited to discuss the Remote Area Development policy (following the Botswana tradition of consultations), an impoverished

NUPI OCTOBER 00 Dilemmas in Norwegian Development Assistance to Indigenous Peoples 219 man from a remote settlement may stand up and speak about ‘the problems that we Basarwa encounter’. Invariably he will be repri- manded by one of the organisers who emphasise that the meeting is called to discuss ‘poor RADs, not the Basarwa’. This situation is confusing for everyone, but is frustrating and humiliating for the Basarwa whose very real concern about discrimination is consist- ently denied legitimacy. It should be added, however, that the ambiguity created by combining (implicit) cultural and (explicit) socio-economic criteria to form one bureaucratically defined category is not always perceived as a problem, because there is a considerable empirical overlap of the two sets of criteria. Not all the poor people in marginal areas are Basarwa, but most of them are. Conversely, not all Basarwa qualify for assistance according to the RAD concept, but most of them do. Empirically, the programme deals mostly with clients who are both Bushmen and RADs, and for many practical purposes it is not always necessary to make a distinction. Thus, the welfare objective can be addressed irrespective of the terminology used. Absolute or relative deprivation may be good justifications for setting up a welfare programme. However, as noted in the 1988 agreement, a basic prerequisite for Basarwa development is the ‘promotion and formal recognition of local institutions and leadership’. To achieve this, a client model is not adequate. Any standard manual in community development and a whole school of grass root community developers (from Chambers, 1983, onwards) argue convincingly that in order to achieve development one has to take as the point of departure the resources people possess. This does not primarily mean material resources, as poverty will always be a defining feature of a development programme. But development depends on local mobilisation and local skills. It is only by recognising the resources that people possess in terms of skills, aspirations and organisations that a programme may adequately mobilise those resources in a development effort. The designation Remote Area Dweller (RAD) is presumed to be a neutral, non-ethnic bureaucratic term. It is, however, deeply resented by those who are so called. To borrow a set of concepts from gender studies, it seems that the RAD programme, in its effort to be culture neutral has become culture blind. In the Botswana context, the carefully worded neutrality of the programme, trying to be all things to all people, has in effect deprived the target group of a cultural identity which could have been a mobilising factor. It has also deprived the target group of their dignity, as they are re-

NUPI OCTOBER 00 220 Sidsel Saugestad

duced to passive welfare recipients. Moreover, by disregarding cultural characteristics which distinguish the San/Basarwa (and other minority groups) qua ethnic groups, the accumulated professional knowledge about the culture of such groups becomes, by definition, irrelevant. In its place strategies of implementation are formulated according to prevailing superficial stereotypes about the Basarwa: as people of the past who, because of their nomadic disposition, have yet to learn to value village life. In any evaluation of the RAD programme, and the achievements of NORAD support, it is important to make this analytical distinction between a welfare approach to the target group, and an approach that recognises their minority status as part of the problem. A development programme needs to go beyond the symptoms (the mani- fest poverty) and look for the generative processes that create these symptoms. The terminology used to describe a given situation also indicates what can be done to change it. The poverty angle has contributed to a clientisation of the target group. This emphasis may also explain the lack of acknowledge- ment of Norwegian achievements within the RAD programme when development assistance to Botswana was phased out. The long-term perspective needed to benefit from a ‘Saami parallel’ came to a rather abrupt stop. What can this tell us about Norwegian strategy for ‘work with indigenous peoples in development assistance’?

6. A Hidden Agenda? In communication about the Remote Area Development Pro- gramme, a wider indigenous perspective and the relevance of the Saami parallel are both indicated by the use of euphemisms. Care was taken not to force a specific terminology into the official project documents. Instead the more rounded formulations referring to groups who ‘tend to rely on hunting and gathering’ and ‘tend to be culturally and linguistically distinct’ have been used. The title of a pamphlet produced by NORAD in 1989, as Nor- way became the main foreign donor to the RAD programme, illustrates the Norwegian thinking behind. It is called From Bushmen to RADs6 (Eidhammer, 1989), and was produced to inform the Norwegian public. The pamphlet gives a balanced discussion of the problems the Bushmen are facing in contemporary Botswana society, describing their structural position as an indigenous people. A

6 ‘Frå buskmenn til RADs’. Author’s translation.

NUPI OCTOBER 00 Dilemmas in Norwegian Development Assistance to Indigenous Peoples 221 historical sketch outlines stereotypes in attitudes and expressions that the Bushmen have suffered over the centuries, the realities of the contemporary precarious adaptations, and the need for innovative measures in order for them to survive in a country undergoing rapid socio-economic transformation. The pamphlet acknowledg- es that international ideas about indigenous self-government do not go down well with the Botswana government, hence it accepts Bot- swana’s position: that Bushmen have to become RADs (Remote Area Dwellers). The reader is reminded that a major feature of the cordial relations between Norway and Botswana is their opposition to apartheid policies, and Botswana’s objection to separate development is to be understood on that background. The author worked at the NORAD office in Gaborone at the time when Norwegian support to the RAD programme was negotiated, so one may assume that the text reflects quite closely the official view. Permeating the entire discussion is a belief that the Norwegian presence may make a contribution through the further development of the most progressive components of the programme, and that this contribution is a question of influencing attitudes, not only to allocate some money. From this and other documents one may infer the outline of a strategy: The best way to show a concern for the Bushmen is through cooperation with the authorities in Botswana, and the way to achieve cooperation is through the formal agreement. It was tacitly assumed that over time it would be possible to steer the implementation of RADP in the ‘right’ direction, i.e., to contribute to a process whereby internationally recognised standards would gradually be taken into consideration, and that this would be possible whichever terminology was used. If this meant ‘swallowing some camels’, as one programme officer expressed it, this was seen as a small price to pay for the opportunities provided for presence and dialogue. Assuming my analysis is correct, does this strategy make for a hidden Norwegian agenda? And do the attitudes outlined above add up to a Botswana–Norwegian disagreement underneath the surface of cooperation? The question cannot easily be answered by a yes or a no. I would suggest that to the extent that there are some differences in the agendas, these have in no way been hidden. How- ever, the presentation of the agenda has clearly been modified by some rhetorical conventions that are common in political communication. It is important to remember that the communication about poli-

NUPI OCTOBER 00 222 Sidsel Saugestad

cies and programmes for remote area development has taken place in a public arena, between different stakeholders, and with different audiences in mind. Communication is not only between the two governments, but also between the governments and their respective constituencies. It is a strategy with a proven effect, in Botswana as in Norway, to formulate policies in a terminology that is sufficiently vague to command the broadest possible support, and to eliminate possible criticism. It can be claimed with considerable justification that the ostensibly neutral RAD terminology has been essential for getting the programme under way. But with equal justification it can be claimed that the ambiguous terminology used in the programme agreements has invited the misunderstandings that came up at later stages, and certainly has made it difficult to make realistic ad- justments as the programme progressed. The complex layers of communication and meta-communication about Bushmen that char- acterise Botswana society, gave little reason to expect the terminology of the RAD programme to be straightforward, but should perhaps have prepared better for the problems in communication that arose at a later stage.

7. Oppositions Revealed From long before NORAD became the main donor to the Remote Area Development Programme, there has been a certain annoy- ance in Botswana about foreign interest in the Bushmen. An abid- ing allegation has been that foreigners want to keep the Bushmen apart in reservations, preventing them from reaping the benefits of modern life. While anthropologist, explorers and film-makers have revelled in the exotic way of life, the national policy has been, as noted earlier, to integrate them in what is called ‘the mainstream’, in other words to make them as like the Tswana majority as possible. This, however, has not prevented the acceptance of foreign donor money, which always has been more easily solicited for programmes affecting the Bushmen. As it turned out the controversy that came up between the Norwegian and Botswana sides did not originate from within the programme, but was due to a very different process: the emergence of a San-based interest-organisation, called The First People of the Kalahari. It all started at a (NORAD-supported) workshop on Sustainable Rural Development in April 1992. A fragile elderly man looked down on the chequered carpet of the conference hall, and compared it to the land of his native Ghanzi, which was partitioned

NUPI OCTOBER 00 Dilemmas in Norwegian Development Assistance to Indigenous Peoples 223 into squares by fences separating people from the land they used to live on. His sense of injustice and grief was eloquently translated from Naro to English by his companion, John Hardbattle. While the sentiments expressed were well known to those familiar with the situation, it was the first time they were expressed publicly, even in a Khoesan language, with prominent civil servants and politicians present. As a follow-up to the statements made, the Permanent Secretary of the Ministry of Local Government and Lands invited those who had spoken up to a meeting with the Ministry, to discuss their grievances. For a number of reasons the meeting went off rather badly. The delegation had expected an opportunity to present some of their needs, and to propose new channels for communication and dialogue between the San and the government. The government did not want any new structures. Their main concern was to instruct the delegation to follow proper procedures: that is to say to take their grievances to the Chiefs, Councillors and Members of Parliament who were duly elected to be their spokespersons. The meeting led to an avalanche of accusations in the press, claiming that ‘Basarwa demand self-rule’ (Mmegi 22.05.92). The Permanent Secretary, who had recently been on a study tour to Norway, was quoted as saying ‘she believed that the Basarwa’s demand for “self-rule” was being instigated by some nongovern- mental organisations and some donor agencies from outside the country [who] want to see them lead a nomadic life’ (Mmegi 22.05.92), and she further ‘volunteered that the idea of a national council for the Basarwa probably originated with the Norwegians, as it seemed to conform to what had been organised for the Saami people in Norway, which she had just visited’ (Gazette 03.06.92). The allegations of untimely outside interference and instigation of the Basarwa went on in the press for some time, with a representative from SIDA defending the comparison with the Saami (‘SIDA sides with “Voiceless” Basarwa’, Gazette 27.05.92). All through 1992 the relationship between Norway and Botswana was distinc- tively cool, and the annual programme meeting postponed twice. As it turned out the diplomatic tiff was an interlude, and by 1993 the relationship between the two countries had resumed its normal cordiality. The incidence, however, had revealed two things: First, it showed that while there was no document spelling out a specific Norwegian stand on Bushmen development, a Norwegian attitude was certainly inferred, but perhaps not the one intended. Moreover, it showed that the efforts made on the Norwegian side to hold up the ‘Saami parallel’ as an example to others, had not had the de-

NUPI OCTOBER 00 224 Sidsel Saugestad

sired effect. The second point deserves some closer examination. The quotes above express the conclusions drawn by a permanent secretary who had recently visited Oslo, Tromsø, Karasjok and Kautokeino together with other Ministry staff in charge of the Re- mote Area Development Programme, to study the situation of the Saami. She made it clear that she was not very impressed. The same conclusions had been expressed even more clearly by the first Prin- cipal Administration Officer assigned to the programme. As part of the ‘acceleration’ following NORAD involvement from 1988, the NORAD office in Gaborone set up a comprehensive training programme to look into minority issues, including a study trip for two persons to Norway and the Saami areas. The officer in question returned to Botswana with the distinct impression that in Norway he had learned what not to do. If the Saami were given their own parliament they would have no influence in the national parliament; they had their own homeland, hence were not integrated with the Norwegians; and at school they had to learn their minority language which reduced their mastery of the majority language (Benedicte Ingstad, personal communication). Almost a decade later, in November 1996, the then Minister of Local Government and Lands met with the President of the Saami Parliament in Tromsø. A terse notice in the government paper after this visit announced that ‘Mr. Balopi who is on a European tour to discuss the Basarwa issue ... met with Mr. Ole Magga, the leader of the Sami which is a tribe indigenous to Norway. Mr. Magga told Mr. Balopi that the Sami do not have the right to own land ... For his part Mr. Balopi said the Basarwa have the right to own land’ (Daily News 17.10.96). It is hard to tell why these trips produced such amazingly slanted conclusions. Partly they were too brief. They were all defined as study trips, but did not leave much time for discussion and clarifica- tions, as the organisers tried to include a many elements as possible showing off the achievements of Saami emancipation.7 This could have worked if the visitors had an understanding in the first place of why the Saami wanted separate organisations, and hence were interested to learn how they had achieved this. Apparently this was not the case. The official visitors from Botswana brought with them old stereotypes about Europeans wanting isolation and separate development for the Bushmen, and this was not a policy they wanted

7 This is noted with the benefit of hindsight, as the present writer was involved in the planning of two of these visits.

NUPI OCTOBER 00 Dilemmas in Norwegian Development Assistance to Indigenous Peoples 225 to adopt. In the spirit of cordial exchange, these different premises were never made explicit. In my own job experience, I encountered the same ambiguity in expectations. I was recruited by NORAD to a position as research facilitator to RADP (based at the University of Botswana 1992– 93), with a background from research on relationships between nation-states and indigenous minorities in other contexts (Norway, USA, New Zealand). I understood my brief to be among other things to introduce a comparative perspective on the situation in Botswana (Saugestad, 1993). Predictably, some Botswana government officials did not like this angle. Moreover, and perhaps equally predictably in view of the diplomatic considerations outlined earlier, some NORAD officials may easily have thought that this position was stated to clearly. (Basically, I believe that representative indigenous organisations are necessary in order to arrive at a constructive dialogue with governments.) In view of the types of misunderstandings cited above, however, it is hard to see how a fruitful development dialogue can be reached without more straightforward analyses of the situation. The most successful example of donor involvement has been the two regional seminars on the development of San people, which took place in Windhoek in 1992 and Gaborone in 1993. These came about as the result of a long process, with SIDA and NORAD jointly suggesting that such conferences be held and offering to fund the expenses. By any reasonable standard these conferences were quite successful. The Botswana conference, under the subheading Common Access to Development was preceded by a lengthy process of consultations in the districts that brought forward some 40 San representatives, duly elected from their communities, and a similar delegation form Namibia. Three days of lectures, debates and group work were attended by governments officials, diplomats, NGOs and academics. The last night 11 resolutions were unanimously passed at a session chaired by two permanent secretaries (GOB, n.d). I have counted the open-minded proceedings of the Second Regional Conference on Development Programmes for Africa’s San/Basarwa Population as the first major achievement of the Norwegian effort to bring San (meaning indigenous) issues on to the public agenda. It was also to be the last. There may have been a foreboding of this in some of the statements. In the greetings presented at the opening of the conference, Swedish Ambassador Sten Rylander expressed his concern:

NUPI OCTOBER 00 226 Sidsel Saugestad

If you go away from this conference without a commitment to further the process of consultation between people and Gov- ernment, and on that basis, formulate, modify and implement policies with respect to the special needs of disadvantaged minorities, no matter how impressive the resolutions passed at this gathering are, the whole exercise is of little use (GOB, n.d.:9).

The meagre follow-up to any grand statement of support is a common experience for the San, as demonstrated by one of the spokes- men, Aron Johannes, who concluded his opening statement with the following remarks:

We have come to this conference, we have been received well and we have been treated nicely. It should not be that after the conference we go back to our tears and to the dust. We left the hunger behind us, here we are well fed. We should not go back to the way it was before (from fieldnotes).

The day after, preparation for Norwegian phasing-out started.

8. Norwegian Withdrawal Norwegian involvement in the Remote Area Development Pro- gramme was an ambitious and in many respects pioneer effort of development assistance. It deserves an analysis in its own right. While recognising the value of the RADP as a rural development and welfare programme, two factors have reduced the potential benefit of a substantial Norwegian involvement:

Bad timing Withdrawal from the RAD programme has been a case of exceptionally bad timing. We have seen how a carefully formulated long-term strategy towards dialogue was linked to the programme for finan- cial assistance. It was made clear, also to Botswana counterparts, that Norwegian acceptance of the non-cultural, supposedly neutral, character of the RAD programme was to some extent a matter of expediency – once established as a legitimate partner in the development process it was envisaged that it would be possible to convince the GOB gradually to change their position in line with the international trends. Even simple development programmes take time, but in this case the objective had not only been to induce changes among the

NUPI OCTOBER 00 Dilemmas in Norwegian Development Assistance to Indigenous Peoples 227 target group, but also to influence, through a diplomatic dialogue, some basic value premises of the counterpart government. By any reasonable standard this should call for a very long-term involvement indeed. Instead, participation in this extremely complex and controversial programme only ran fully for five years (1989 to 1993). The decision by the end of 1993 to terminate assistance was made by the Ministry of Foreign Affairs, on a state-to-state level, and with good geopolitical justifications. It was the success of Bots- wana’s fiscal policy (aided by a considerable profit from diamond mining) and the rapid democratic transformation of South Africa that resulted in Norwegian withdrawal from bilateral cooperation, on the most cordial of terms. There is no indication that the nature of the RAD programme was given specific consideration. However, the division of labour in the foreign service means that the policy decision to withdraw, made in Oslo, did not necessarily reflect the assessment of needs for further involvement, as seen by NORAD Gaborone where the day-to-day implementation was carried out. Even given the overall policy decision, the withdrawal could have taken different forms. The resolutions passed at the Regional San Conference in October 1993 had cleared the ground for an intensified involvement in the priority areas identified. Norway could also more actively have sought out new alternative avenues for collaboration, following the emphasis on human resource development, NGO and institution-to-institution contact after withdrawal (GOB/ GON, 1993).8 The procedure actually followed was to start a lengthy process of evaluation in 1993 which effectively blocked any new initiatives during the entire phasing-out period. By the time the recommendations from the Chr. Michelsen evaluation had been through the political apparatus (early 1996) the winding up of the Gaborone office had already started.9

8 An extensive NORAD-funded health programme is still ongoing at a ministry- to-ministry level, but then the Ministry of Health has always been more appre- ciative of Norwegian development cooperation than has the Ministry of Local Government. 9 Two activities have continued at the NGO level: a collaborative programme for San/Basarwa research between the University of Tromsø and the University of Botswana (with funding from the Norwegian Ministry of Foreign Affairs through NUFU) and support to the Naro-based Kuru Development Trust through the Norwegian Church Aid. These are among the recommendations made by the CMI evaluation, while the many other recommendations made have not been followed up.

NUPI OCTOBER 00 228 Sidsel Saugestad

Problematic implications of non-interference There are constraints of a diplomatic nature inherent in bilateral cooperation that contrast sharply with the freedom to voice progressive views in international fora. Should NORAD have come out and stated its priorities more clearly in Botswana? NORAD would probably answer no to such a question, because NORAD’s stated policy has been to support Botswana’s policy. Norway is certainly propagating new measures in international fora, stressing the need for representative structures that can voice indigenous claims and engage in constructive dialogues with their governments. Norway has given, and may continue to give, direct assistance to indigenous organisations in Botswana, and to support structures that may assist the San in gaining political concessions. Although such support, and indeed all contact with the emerging organisations, have always been open and above board, it has not been easy to avoid accusations of interference. Moreover, the expression of support, and support given, have also opened the way for charges that San initiatives are influenced or even dominated by foreign interests. However, such problems are structural problems, predictable within the parameters of indigenous organisational development over the last decades. All acts of international solidarity render participating nations vulnerable to charges of im- proper meddling in internal affairs. In Botswana’s political context, the RAD programme could only be expected to contribute to development in a more restricted sense, addressing practical development needs. The potential contribution of the programme towards overarching development objectives such as the mobilisation, organisation and self-determination of San people, was not realised, and, given the design of the programme, this was not to be expected. The impact of bilateral aid has a limited positive effect if some measures require political concessions that come into conflict with the recipient government’s other priorities. Thus, the principle of recipient orientation left little room for a policy that could bring about more fundamental changes. In a country programme the recipient is by definition the state, on behalf of the section of the population that the programme in question is targeting. The contemporary debate on indigenous rights, however, has exposed the limitations to a majority’s ability to adequately represent the basic interests of an indigenous minority. By being loyal in the support to the Botswana government, the end result has been NORAD support to a rural development policy which basically has served to maintain a status quo in the distribu-

NUPI OCTOBER 00 Dilemmas in Norwegian Development Assistance to Indigenous Peoples 229 tion of power and property. However, interests and attitudes are not uniform in Botswana either, and a dialogue resonates differently with different individuals. While bureaucratic systems strive to present an outwardly unified front of stated policy and codified rules, they include internal differences in backgrounds, allegiances and commitments. There are liberal elements both within the government and civil society that would make natural partners for further involvement. Events in Botswana since 1993 have indeed shown that the momentum has changed, from originating from within a state-to- state relationship to a situation where organisations, or civil society, have become the most important driving force. The two organisations present at the 1993 meeting, the First People of the Kalahari and the Kuru Development Trust, have continued to develop, and strive to define their position vis-à-vis the ‘electorate’ that they represent and vis-à-vis government structures that they try to engage in dialogue. New organisations have been formed, forging alliances with San people across the border to Namibia and South Africa. Community Based Organisations (CBOs) have become a popular new model for community development, incorporating more directly a concept of empowerment. These emerging organisations are part of and are being supported by the wider indigenous movement, having ethno-political objectives on their agenda (Saugestad, 1998a, 1998b). This process is taking place outside the confines of the Remote Area Development Programme, and NORAD withdrawal has not significantly halted this new development. It means, however, that an opportunity to support such new and vulnerable structures (a main objective noted in the ‘follow-up’ plan) has been passed over.

9. To Follow Up In view of the renewed wish to include indigenous peoples in development assistance, a few more general conclusions may also be useful.

Indigenous is a cumbersome bureaucratic concept First and foremost one should keep in mind that even under the best of circumstances, ‘indigenous’ is a cumbersome concept that disrupts administrative preferences for clear-cut and unambiguous target groups. While recognising that Africa and Asia represent particular conceptual challenges, the notion of indigenous peoples

NUPI OCTOBER 00 230 Sidsel Saugestad

is perceived by bureaucrats all over the world as a concept that is inconvenient, diffuse and difficult to handle. Any procedure for singling out one group for special treatment and/or affirmative action goes against standard democratic guidelines for equal treatment. The implications of ‘fairness, not favours’ that Norway has put forward so well as a general policy, need careful justification in each national context.

State-to-state collaboration is the most difficult The ‘follow-up’ plan notes that presently there is little collaboration on this level, but goes on to say that in ‘future efforts to improve the situation for indigenous peoples it will be important to engage the authorities in approaches related to indigenous peoples’ (4.3). This will certainly require diplomatic balancing acts also in the future. However, the shift towards more emphasis on human rights in development assistance will probably make it easier also to take up issues of indigenous rights. Moreover, the concept of conditionality has become a new buzzword (Report to Parliament No. 51 (1991– 92); Stokke, 1995). This feature may well modify the emphasis on recipient responsibility that characterised Norwegian involvement in the RAD programme. Norway may have been hesitant in accepting conditionality in the economic sphere, but in the diffuse, but also more pervasive field of democracy, human rights and good governance, conditionality matches the Norwegian long-standing commitment to these values. This should make it more easy to incorporate Norwegian assistance to indigenous peoples in strategies for development assistance, as already noted in Report to Par- liament No. 21 (1999–2000) on the promotion of human rights. At the present point in time, Guatemala and Nicaragua are probably the two countries where links between human rights and indigenous rights are being explored most explicitly. Perhaps the best testimony to the constraints of state-to-state collaboration is the exponential growth in assistance channelled through voluntary organisations (Visnes, 2000:102). It has opened up for a considerable increase in support to projects directed towards indigenous peoples, and their organisations, through NGOs.10 An obvious advantage of working through voluntary associations is that the (often self-imposed) restrictions of a diplomatic nature that

10 According to the follow-up plan 52 million NOK were distributed through NGOs in 1999 to indigenous peoples and/or their organisations.

NUPI OCTOBER 00 Dilemmas in Norwegian Development Assistance to Indigenous Peoples 231 characterised the state-to-state cooperation with Botswana, are reduced. As long as activities do not directly contravene national law, support can be given to projects that explore alternative development avenues. It is a common observation that social innova- tions can more easily be introduced outside of government structures, and the type of changes associated with the ‘indigenous movement’ makes no exception.

Poverty alleviation implies a danger of ‘clientisation’ The plan notes that in ‘the development of country strategies the situation of indigenous peoples should be considered both as part of strategies to relieve poverty, … and as part of the work for ad- vancement of human rights’ (4.3). Humanitarian assistance directed to the poorest sections probably remains among the least controversial forms of assistance to accept for most governments. There will certainly be situations also in the future which justify the introduction of measures targeting indigenous minorities through ‘the back door’ in order to relieve oppressive poverty. But keeping in mind the pervasive social stigma that attaches to indigenous peoples, particularly in the Third World, the negative effects of being identified as welfare recipients should not be overlooked. There is no such thing as a free gift, writes Mary Douglas (1990:vii) in the foreword to Mauss: The Gift. What is wrong with the so-called free gift is the rejection of mutuality between the one who gives and the one who receives. The gift that cannot be returned becomes a humiliation.

Challenges in the use of the ‘Saami experience’ Last, but not least, the case-study from Botswana has shown us that Norway’s considerable experience in the area of indigenous relations, accumulated over years of sometimes turbulent relations between the Norwegians and the Saami, has been applied somewhat inconsistently to the situation of the San in Botswana, Na- mibia and South Africa. The close links between experiences from Norwegian–Saami relations and Norwegian international advocacy for indigenous peoples’ rights are persistent themes in Norwegian official rhetoric. However, there has been little or no overlap between people engaged in this field of international discourse and the personnel that deal with indigenous issues as part of foreign development assistance.

NUPI OCTOBER 00 232 Sidsel Saugestad

To extract the universal lessons from the complex history of Saami–Norwegian relations would require other avenues for communication than the traditional brief visit by an official delegation. Even in the extended Norwegian–Botswana relationship more stereotypes were confirmed than modified. I do believe, however, that there is a great potential for learning, if this sort of exchange is handled professionally. There is an important role here for Saami organisations. The ‘follow-up plan’ emphasises this, by calling for more involvement by the Saami Council and the Saami Assembly (7.2, 7.9). This is also in line with international trends stressing indigenous peoples-to-peoples contacts as important supplements to state collaboration. Saami organisations have a high profile internationally and there is a tremendous demand from indigenous organisations worldwide for collaboration and support from their ‘Saami brothers and sisters’. This demand, however, is presently far exceeding the capacity for Saami organisations to address all deserving cases. To judge from the involvement with San in Southern Africa, the involvement by Saami organisation so far (mainly the Saami Council and Saami Council of Churches) has brought forward some positive encounters, but little continuity in exchange of experiences. It is perhaps necessary to recognise that familiarity with Saami (or indigenous) organisational development does not automatically bring about a similar interest or competence in issues of development assistance. The constructive use of the ‘Saami experience’ should therefore be seen as a professional challenge, to involve policy makers and researchers with experience from both sides of the Norwegian– Saami relationship. The presentation of a plan for a renewed focus on Norwegian development assistance to indigenous peoples should be welcomed. It is setting out a path ahead that calls for new initiatives. It is also a cautious document, calling for consolidation of ongoing commitments. Norwegian assistance to indigenous peoples has been sin- cere, but erratic in the past. The present article has brought out some of the implicit assumptions during a sustained engagement in Bot- swana. It is argued that some of the problems encountered might have been predicted and more easily handled if the latent contra- dictions in the programme had been recognised. Such analyses of achievements, shortcomings and challenges should be an integral part of the planning ahead. Hopefully some of the points raised above can be of use in this process.

NUPI OCTOBER 00 Dilemmas in Norwegian Development Assistance to Indigenous Peoples 233

References Anderson, B., 1983, Imagined Communities, London: Verso. Brantenberg, T., J. Hansen and H. Minde, eds., 1995, Becoming Visible. Indige- nous Politics and Self-Government, Centre for Sámi Studies, University of Tromsø. Bourdieu, P., 1977, Outline of a Theory of Practice, Cambridge: Cambridge Uni- versity Press. Chambers, R., 1983, Rural Development: Putting the Last First, Harlow: Long- man Scientific & Technical. CMI (Chr. Michelsen Institute), 1996, NORAD’s Support of the Remote Area Development Programme (RADP) in Botswana, UD Evaluation Report 1.96, Oslo: Norwegian Ministry of Foreign Affairs. Davidson, B., 1992, The Black Man’s Burden. Africa and the Curse of the Nation- State, London: James Currey. Douglas, M., 1990, ‘Foreword’, in M. Mauss, The Gift, London: Routledge. Economic Consultancies, 1994, Remote Area Development Pre-Evaluation Study. Assessment of District Level Expenditure Record, Gaborone. Eide, A.,1985, ‘Indigenous Populations and Human Rights’, in Brøsted et al., Native Power, The Quest for Autonomy and Nationhood of Indigenous Peo- ples, Oslo: University Press. Eidhammer, A., 1989, Frå buskmenn til RADs, Oslo: NORAD. Eidheim, H., 1992, Stages in the Development of Sami Selfhood, Working Paper No.7, Department of Social Anthropology, University of Oslo. Gellner, E., 1983, Nations and Nationalism, Oxford: Blackwell Publishers. GOB (Government of Botswana), n.d., Regional Conference on Development Programmes for Africa’s San/Basarwa Populations: Common Access to De- velopment, Gaborone. GOB/GON, 1988, Agreement between the Government of the Kingdom of Nor- way and the Government of the Republic of Botswana on support to the Accelerated Remote Area Development Programme, Gaborone. GOB/GON, 1993, Memorandum on the Strategy for Development Cooperation Between Botswana and Norway 1994–1996 and Beyond. Good, K., 1993, ‘At the Ends of the Ladder: Radical Inequalities in Botswana’, Journal of Modern African Studies, Vol. 31, No.2. Granberg, P. and J.R. Parkinson, 1988, Botswana. Country Study and Norwegian Aid Review, Bergen: Chr. Michelsen Institute. Gulbrandsen, Ø., M. Karlsen and J. Lexow, 1986, Botswana. Remote Area De- velopment Programme. Report to the Norwegian Ministry of Development Cooperation. Hettne, B., 1990, Development Theory and the Three Worlds, Harlow, Essex: Longman Development Studies. Hitchcock, R., 1988, Monitoring, Research and Development in the Remote Areas of Botswana. Report to the Ministry of Local Government and Lands, Bot- swana and the Ministry of Development Cooperation, Norway. Kann, U., R. Hitchcock and N. Mbere,1990, Let Them Talk. A Review of the Accelerated Remote Area Development Programme. Report to MLGL and NORAD.

NUPI OCTOBER 00 234 Sidsel Saugestad

Karlsen, M., 1992, BOT 022/LG127. Remote Area Development Programme. Focus on NORAD’s Expenditure, Gaborone. Mathiesen, P. 1987, ‘Bureaucratic Categories and Ethnic Ascription’, Ethnos, No. 3–4. MLGL (Ministry of Local Government and Lands), 1989, Guidelines for the Economic Promotion Fund (EPF). Minde, H. 1996, ‘The Making of an International Movement of Indigenous Peo- ples’, Scandinavian Journal of History, Vol.3, No.21. NDP7, National Development Plan 7 1991–1997, Gaborone, Ministry of Finance and Development Planning, Government of Botswana. NOU (Norwegian Official Report) No.18, 1984, Samenes rettsstilling, Oslo: Ministry of Justice. NOU (Norwegian Official Report) No.14, 1985, Samisk kultur og utdanning, Oslo: Ministry of Culture and Science. NOU (Norwegian Official Report) No.5, 1997, Urfolks landrettigheter etter folke- rett og utenlandsk rett. Bakgrunnsmateriale for Samerettsutvalget, Oslo: Ministry of Justice. Saugestad, S., 1993, Indigenous Peoples, National Models, and Recent Interna- tional Trends, Working Paper No. 64, National Institute of Development Research, University of Botswana. Saugestad, S., 1998a, The Inconvenient Indigenous. Remote Area Development in Botswana, Donor Assistance and the First People of the Kalahari, Faculty of Social Science, University of Tromsø. Saugestad, S.,1998b,‘“When I Say Land I Talk about my Mother”: Contempo- rary Perspectives on Indigenous Organisations and Encounters in Southern Africa’, in A. Banks, ed., Proceedings of the Khoisan Identities and Cultural Heritage Conference, Cape Town, 1997, Institute for Historical Research: University of Western Cape. Stokke, O., ed., 1995, Aid and Political Conditionality, London: Frank Cass. Thuen, T., 1995, The Quest for Equity. Norway and the Saami Challenge, Insti- tute of Social and Economic Research, Memorial University of Newfound- land. Visnes, K., 2000, ‘Skaff folket et talerør’, in G. Dale et al., Grenser for alt. Kri- tiske perspektiver på norsk utenrikspolitikk, Oslo: Spartacus. WCCD (World Commission on Culture and Development), 1995, Our Creative Diversity, Paris: UNESCO. Wille, P., 1995, ‘Norges engasjement i menneskerettighets- og urfolksspørsmål’, in Brantenberg et al. Wilmsen, Edwin, 1989, Land Filled with Flies. A Political Economy of the Kala- hari, Chicago: University of Chicago Press.

Botswana newspapers: Daily News, 17.10.96 Gazette, 27.05.92, 03.06.92 Mmegi, 22.05.92

NUPI OCTOBER 00 Forum for DevelopmentThe InnocentStudies | DeveloperNo.2-2000 235

The Innocent Developer: Ethics and Ox-Cultivation A Case-Study from the Southern Sudan

Arne Olav Øyhus

1. Introduction Since the commencement of the industrial era more than 250 years ago, development – as generally conceptualised from a European perspective – has been more or less synonymous with modernisation. Modernisation was regularly interpreted as the constant generation and application of more and more efficient technology in the societal production process. This view has had a significant consequence for the moulding of development aid during the second half of the twentieth century. Transfer of technology has been considered by a great community of development specialists (‘developmentalists’) to be the quintessence of development aid. 1 When defining technology I believe it is important to recognise that it has a two-sided character: it has a practical side, i.e., it is a tool (a ‘hardware’); and it has a cognitive side, i.e., it is a corpus of knowledge (a ‘software’). As a corpus of knowledge, technology is part of a larger symbolic system, a culture (Ingold, 1986). In my opinion, it has been a problem that the great majority of development agencies and experts have concentrated their attention and efforts so much on the practical side of technology, more or less overlooking the fact that it is knowledge and skills (theory) which make it possible to use a tool in practice.

1 From 1983 to 1986 the author of this article was employed as a development worker for an NGO in the Eastern Equatoria Province of the Southern Sudan. From 1985 to 1986 he headed the extension activities in the eastern regions of the province, i.e., within the ‘ethnic’ areas of the Didinga, the Toposa and the Boya. During this assignment he undertook a community survey. The data for this article build on this survey.

NUPI OCTOBER 00 236 Arne Olav Øyhus

If we accept that technology is the core issue in development aid, and that a technology is embedded in a larger symbolic system, a couple of relevant questions to ask are: is it possible to transfer technology from one setting to another without causing ethical problems? (see, e.g., Goulet, 1989), and to what extent do development professionals take such problems into consideration when they plan and implement development activities? In the rather brief history of development aid the general picture so far has been that technology has moved unilaterally in one direction, from the modern, industrialised countries in the North to the ‘traditional’, non-industrialised countries in the South. In a North- ern context (the giving end) the technology that was transferred quite regularly belonged to a holistically integrated sociocultural system; in a Southern context (the ‘receiving’ end) the same technology was quite regularly unknown and non-integrated, i.e., it was alien to the sociocultural system. This does not, by itself, represent an ethical dilemma: what may represent a dilemma are the methods by which the technology was transferred, and the consequences of its transfer. There are many examples showing that technological innova- tions have brought unforeseen consequences into the receiving system, even to the degree that the viability of the system has been threatened (see, e.g., Toulmin, 1991; Franke and Chasin, 1981; Hall, 1990; Perrings, 1989; Repetto and Holmes, 1983; Swift, 1976). In a broad historical and comparative perspective we may quite commonly perceive a calibration between three aspects of a society, its technology, the physical resources it utilises, and its social and cultural structures and institutions. We may say that it has been a common feature for societies to apply a technology that is adapted to its social and natural environment. The adoption of a new technology will therefore, by necessity, cause changes in the operation of societal structures and institutions, and it will disrupt the ‘adapted’ relationship between society and nature. Consequently, a change in technology will imply a change in a society’s mode of adaptation. From this, we may further logically deduct that a change in a society’s mode of adaptation will simultaneously lead to changes in the structures and strategies that various social units (for instance, households, villages, and tribes) apply to secure their viability, defined as their material and spiritual survival and continuity. Some pertinent ethical questions to ask are thus: is it possible for those who advocate the adoption of a new technology to anticipate all the relevant consequences of the adoption? Is it possible that one consequence might be the demolition of core social structures and insti-

NUPI OCTOBER 00 The Innocent Developer 237 tutions? Is it possible that another consequence might be a more fragile relationship between a society and its physical environment? In case of failure, for instance that the people who adopt the new technology will suffer spiritually and/or materially, who will then take the responsibility? If the advocates of a new technology cannot guarantee its ‘positive’ results, should they then desist from advocating it all together? The aim of this article is to shed light on ethical issues connected to aid. I will do so by presenting a concrete case-study, namely the introduction of ox-cultivation to the pastoral Toposa in the mid 1980s. At first glance this seems to be a rather straightforward case, both from an ethical and a technical point of view. Technically, the case concerns the introduction of a simple mechanical item – the ox-plough. Ethically, the case concerns the considerations taken by development experts when they were advocating for the adoption of the plough. Before introducing the concrete case, I will briefly discuss a couple of theoretical questions connected to the ethical component of ‘developmentalism’.

2. The Ethics of Development That ethical issues were not in the forefront in the first decades of international development aid does not imply that aid did not have an ethical foundation. Although we may find that development ethics diverges from colonial ethics, we can also find some common ground between what we may call the ethics of imperialism and the ethics of developmentalism.

Imperial ethics The ethical standards underlying imperial politics expressed, more or less explicitly, the right of the civilised nations to superimpose their own ethical standards on foreign peoples – if necessary – by force. In character, imperial ethics was not global or universal, but purely Eurocentric. It was the ethics of the ‘civilised’ world. An important foundation for imperial ethics was the evolutionary perspective, firmly claiming that the ethical standards of Chi- nese, Indian, Egyptian, Arabian, Persian and African peoples, according to a human evolutionary ladder, were at a lower level than the European. Consequently, European nations, standing at a higher evolutionary level, had the right to suppress and subordinate other nations since this was part of a civilising mission.

NUPI OCTOBER 00 238 Arne Olav Øyhus

Development ethics: Universal standards versus relativistic standards Both for the capitalist and the communist block development aid became an important instrument for controlling global structures in the 1950s. This shift from imperialism to ‘developmentalism’ also brought about a seeming shift in ethics. While the ethics of imperial politics was explicitly based on European standards, the ethics of development aid was perceived to be universal or global. The United Nations International Conventions on Human Rights represent a profound example on these new ethical standards. It is possible to argue that this modern, global ethics is composed of two basic universal principles: universality in values, and universality in obligations. The principle of universal values implies that human values are, or should preferably be the same among all peoples and nations. Universality in obligations implies that fortunate peoples, wherever they live, have an obligation to assist less fortunate peoples, wherever they live. What I feel to be the common ground between imperial ethics and development ethics is that neither of them have any doubts regarding the conviction that one party has the right to superimpose its own standards on another party if only the cause is right and just enough. An alternative to universal ethics, especially to that part which concerns the universality of values, is ‘ethical relativism’. The pro- ponents of this view maintain that values are not universal but relative, i.e., different peoples and nations have different systems of values, and hence, different ethics (Goulet, 1989). Not only do the relativists argue that ethical standards actually vary a great deal among different peoples and nations, they also argue that this fact should be reflected in international aid (see, e.g., Barth 1972).

3. The Case: Introducing Ox-Cultivation to the Pas- toral Toposa

Toposa ethnography

Geographical setting The Toposa is the biggest ethnic group on the East Bank (of the Nile) of the Eastern Equatoria Province, the Southern Sudan. It is not easy to delimit the exact border of the Toposa area, but if we consider the whole area of the East Bank east of the Boya and the Didinga ethnic groups as belonging to the Toposa, the area covers

NUPI OCTOBER 00 The Innocent Developer 239 about 40–50,000 km2 (almost 50 per cent of the whole East Bank). But within this area there are Murle groups, Dinkas and others using the far northern and north-western parts as pastures. In the south-east the Turkana graze their livestock within the Ilemi triangle, and at the Ethiopian border in the east the Nyangatom herd their cattle. On the other hand, we will also find groups of Toposa herders bringing their cattle to camps both within Kenya and Ethiopia. For analytical purposes we can divide the land of the Toposa into three concentric circles, each with a different land use pattern. The inner circle contains the central zone where we find the permanent settlements, and where all farming activities take place. The next circle contains the semi-central zone where we find the wet season pastures and cattle camps. The extension of this area is more or less delimited by the access to the seasonal streams coming from the Didinga Mountains. The outer circle comprises the peripheral area which borders the territories of the neighbouring ethnic groups. This is the area where we find the dry season pastures and cattle camps. Parts of this area are highly contested, and most of the conflicts and warfare between the Toposa and their neighbours have been about the ownership to and the utilisation of this area. The two first circles comprise the area that is more or less exclusively utilised by the Toposa, and the size of it is between 20–30,000 km2.2 Whereas it can be said that the central zone represents the ‘feminine’ sphere, the semi-central and the peripheral zone represent the ‘masculine’ sphere. Farming, house-building and domestic work, all taking place in the central zone, are dominantly female activities. In the two outer zones animal husbandry mainly based on mobile herding performed by men is the dominant activity. We should add, however, that many men do take part in farming activities to some extent, for instance fencing fields, and cutting trees and bushes to clear a piece of land for cultivation. The Toposa land is the driest area on the East Bank. According to records from Kapoeta3 (Bjørtuft, 1984) the average annual rainfall in the period 1980–84 was 695 mm, ranging from 372 mm in 1984 to 818 in 1982.4 The rainfall is extremely unevenly distributed both in amounts and space. For a particular area the variation from year to year may be more than 100 per cent, and even within short

2 K.M. Barbour (1961) The Republic of the Sudan, University of London Press, states that the Toposa area is 36,000 km2 (p.68). 3 Kapoeta is the administrative centre and the only town within the Toposa area. 4 According to S.D.I.T. (1955), the average annual rainfall in the period 1921– 50 was 777 mm.

NUPI OCTOBER 00 240 Arne Olav Øyhus

distances there is a great variation within the same year. Except for the riverine forests, short grasses and scattered thorny trees and shrubs vegetationally dominate the landscape. Topographically, the area is a flat savannah descending slightly from south to north and from west to east. The central parts are situated at an altitude between 600 and 700 m with Kapoeta town at 670 m. There are no high mountains, but some lower hills (‘insel- bergs’) scattered throughout the area. Some bigger, but shallow, seasonal streams are cross-cutting the land from south to north. The streams have their origin in the Didinga Mountains, and are only flooding during the peak of the rainy season. All permanent settlements are situated at short distances from the streams. With some exceptions, all domestic water, besides the water for the animals, is collected from the seasonal streams. Since they are flooding only a couple of months per year people are compelled to dig wells within the sandy riverbeds for the rest of the year. Each herding group has its own wells for cattle watering. Since both the settlement pattern and the mobility pattern of the Toposa are so much determined by the streams, they in many respects function as a nervous system for the Toposa mode of ecological adaptation. It is claimed by many pastoral researchers that pastoral environments are inherently unstable because of climatic factors (see, e.g., Ellis and Swift, 1988; and Fratkin, 1997). Michael A. Little claims that savanna ecosystems impose a number of environmental stresses on people, for instance seasonal aridity, including peri- odic droughts and a scattered and patchy resource base; high ambient temperatures; and serious human (and livestock) diseases (Little, 1980). These stresses are also typical for Toposa. The un- even and erratic rainfall pattern in the area makes crop husbandry extremely vulnerable. What frequently happens is that after the first heavy rains, when people have dug and planted their fields, they have to wait for up to one month for the next rains. Even though the soils have a high water-holding capacity, the seedlings generally wither after 10–14 days. It is hence quite common to seed two to three times before crops are established. This does not represent a dramatic situation when seeds are in plenty, but in periods when one dry year with low yields follows another similar year, the stock of seeds will diminish, even to the extent that the livelihood of many Toposa is threatened. During the famine in 1984–85 we witnessed how Toposa women mixed seeds with sand, and deposited the mixture into holes dug in the ground to protect it from their hungry husbands and children.

NUPI OCTOBER 00 The Innocent Developer 241

Farming Generally, we find two major classes of soils in Toposa. The predomi- nant is the ‘black cotton soils’ (Vertisol), called naro in the Toposa language. This is the soil of the plains. Along the streams and pied- mont areas we find the sandy to sandy-clay soils. Traditionally, these latter soils are not utilised by the Toposa for cultivation. The naro soils have some advantageous qualities making them the most important areas for agricultural production. First of all, they can retain moisture for plant growth for a couple of weeks after the last rains. Secondly, they have a high content of plant nutritional matter. Due to these factors the fields can be cultivated for more than 20 years consecutively. Sorghum (‘durra’) dominates completely the cultivation of these soils. These soils do, however, also entail some serious disadvantages, most importantly the characteristic that they are hard to work. In the dry season the soil shrinks and the surface becomes extremely hard with deep cracks.5 In the rainy season the soil becomes muddy and sticky. The tillage operations are laborious as they are undertaken with the long-handled hoe (emeleko) as the only implement. Farming does not have a long tradition in Toposa (see Eriksen, 1978). It is during the last 40 to 50 years, and especially after the first civil war (‘Anyanya’),6 that the importance of arable farming has increased substantially. In the beginning of the 1970s cultivated crops had become the most important source of food for the poorer sections of the Toposa. The increased importance of farming can be attributed to three main conditions. Firstly, the cattle sector has been in deep trouble because of recurrent droughts and continuous raiding activities since the latter part of the 1960s.7 This means that cattle numbers have seriously diminished. Secondly, the international borders have rather rigidly restrained mobility, with the consequence that it is more difficult to utilise the natural resources optimally. Thirdly, the Toposa have gradually developed a taste for cereals which they earlier acquired in trade with their farming neighbours.

5 Due to these characteristics it is common to call these soils ‘cracking clay soils’. 6 Anyanya is a name commonly used in the Southern Sudan for the civil war between the Northern government forces and the Southern rebel forces (Anyanya) lasting from 1956 to 1972. 7 According to a report from UN, the whole Sudano-Sahelian region has experienced three major environmental predicaments in the last 30 to 40 years: drought, desiccation and dryland degradation (Warren and Khogali, 1992).

NUPI OCTOBER 00 242 Arne Olav Øyhus

In contrast to the other peoples on the East Bank where farming has a typical communal character signified by a high degree of joint decision-making and collective labour, farming in Toposa is a typical individualistic enterprise. Each adult woman cultivates her land alone, or with some assistance from her unmarried daughters, young sons and husband. It does happen that two or more neighbouring women, or sisters, assist each other for some purposes. But this assistance is purely voluntary and between individuals, and it never leads to any stronger corporations. Regarding farming it is therefore correct to claim that there exists no institutional level of co-operation above the household.

Animal husbandry While the rainy season cattle camps are situated in areas more or less overlapping with the area for the permanent settlements, the dry season cattle camps (awi) are found at a much further distance from the settlements, generally at one–two days walking distance. Adult men, ranging from about 15 to 45 years of age, do, in general, live in these cattle camps for the major part of the year. The awi are found in areas with dry season pastures and water points. In the dry season most of the women and children also stay in the cattle camps since the supply of food (blood, milk and meat) is often better there. Before the onset of the first rains they return to the permanent compounds to prepare for the cultivation season. Some time after the rainy season has commenced, and the vegetation has started to yield grass and leaves (usually in April/May), a large proportion of the cattle (mainly milking cows) are brought from the dry season camps to the wet season camps close to the settlements. The milk from these cows is an essential source of food for the cultivators during the cropping season. After the vegetation cover has dried out (usually in October/November) the cattle are brought back to the dry season camps. There is a tremendous variation regarding the number of cattle controlled by each family. When describing the size of their herds the Toposa will never mention the number of heads but the number of ‘stables’ they manage. One stable is a unit of livestock herded together. The number of cattle in one stable will, on average, constitute somewhere between 200 and 300 heads. Even if the herd in each stable is usually mixed from different herd-owning units, the richest herd owners may control cattle equivalent to three stables. During a community survey that we performed, we were told about

NUPI OCTOBER 00 The Innocent Developer 243 herd owners controlling 6–7 stables (up to 2,000 heads of cattle). The poorer strata of the population will have from zero up to 10–15 heads of cattle. As an average, an educated guess would be that a herd-owning unit controls about one stable. This number is based on interviews with local herd owners; there exists very little valid statistical information that can confirm this number.8 Smallstock, even if much appreciated as a source of food, are in lower esteem than cattle. While a lot of religious and cultural idioms are related to cattle, you will find nothing similar related to smallstock. Castrated billy-goats are the only exemption to this rule, because they can be used in certain rituals instead of bulls. According to information we received from the community survey, an average household controls 100–200 goats and sheep. As a rule, there seemed to be little difference in the preference between the two species. The prevalence of one species over the other is usually related to the feeding habits of the two. In areas with open grass- land vegetation (mostly to the west and north), the grazing sheep were more prevalent. In drier areas with thorny shrubs (mostly to the east and north-east), the browsing goats were dominant. Also the pastoral sector in Toposa has a typical individualistic character in the sense that there is an obvious lack of communality and institutionalised co-operation over and above the (extended) family level. The enterprises connected to animal husbandry are either performed by a man and his adult sons, or by a group of brothers (with sons). It is also common for a man to have a ‘bound friend’ in another section with whom he herds his cattle together. But this is based on individual relationships and choices, not on communal commitment. The most important communal feature related to the livestock sector is the land tenure system where the separate sections own separate portions of land used for cultivation and grazing. Any herder, being a member of a section, can utilise his section’s pasture freely. If a herder is granted permission, it is also possible for him to move to other sections’ pastures.

8 According to the ‘Sudan National Livestock Census and Resource Inventory’, Vol. 29 (1976), the mean number of cattle per herd (stable) in their survey was 207.50. It was also found that the mean number of cattle per family head was 15.4. There are, however, certain problems involved in applying these numbers to describe the actual situation. First of all, the sample of Toposa families participating in the census was very small (only six families), and secondly, the (nuclear) family in Toposa is generally not a separate herd-owning unit. We should also add that from our own experience we know that pastoralists are not very eager to divulge the amount of cattle they own (due to taxation, debt relationships, etc.).

NUPI OCTOBER 00 244 Arne Olav Øyhus

With the exception of some important tribal rituals, raiding and defence activities are among the few communal undertakings ever pursued by the Toposa. The latter are commonly performed by young men in the age group of warriors. During the colonial period the borders between the countries of Kenya, Uganda, Ethiopia and the Sudan created few obstacles for the mobile pastoral groups of the ‘Karimojong cluster’ (see below). All the way up to about the mid 1970s these groups regularly crossed the existing international borders in order to utilise the natural resources found within each other’s territories. Firmer control over the borders within the last couple of decades has meant that some of the eastern Toposa sections have had a limited access to parts of their traditional grazing land. Stricter borders have also led to an increased sociocultural ‘distance’ between the Toposa and their neighbours. This has had some obvious negative effects on the Toposa livelihood. The most important consequences are that mobility and flexibility have been constrained, and that an earlier mood of co-operation has been replaced by a prevailing state of antago- nism. The increased difficulties that the Toposa have met due to their encapsulation into a modern state system are typical for many pastoral groups in Africa. Several authors claim that increased incorporation into modern states negatively affects their control over natural resources, and that this leads to a progressive political, economic and cultural marginalisation (Azarya, 1996; Markakis, 1993; de Bruijn and van Dick, 1999).

Settlement pattern According to the population census from 19839 there are 188,153 people living within the area of Kapoeta Rural Council (covering the Toposa ethnic area). This number includes 5,344 persons living in Kapoeta town. The Toposa tribe is divided into sections or sub-tribes (nyakop) with a corresponding geographical area under its jurisdiction. Each sub-tribe is named after the clan which first settled that particular area. New sections are usually formed when old sections grow too big. The sections are again divided into subsections, or village areas (ngitialai), which again are divided into neighbourhoods (ngiarea). Each neighbourhood consists of from one to about five homesteads or compounds (nyere) inhabited by a certain number of families

9 Division of Statistics, Census Office, Juba 1983.

NUPI OCTOBER 00 The Innocent Developer 245

(nyakal). While we may say that nyere means home, nyakal means family or household. Physically, a compound (nyere) is a fortified enclosure made from trunks and thorny bushes, and containing a varying number of huts and granaries. Each nyere comprises a kinship segment, usually a group of brothers with wives and children. The mother and father of the brothers, if still alive, will also live in the compound. Each nyere is led by an elder, usually the father, if alive, or the eld- est brother. The compound will be called by name of its head. The people actually residing within the nyere are generally the wives of the group of brothers with their unmarried daughters, sub- adult sons and other dependants. The physical size and composition of the nyere are determined by the number of wives living there, each wife occupying a hut and at least one granary. Since some rich men may have more than 20 wives, we may find compounds that are inhabited only by the wives of one man. Although some co-operation takes place within the nyere (fencing, house-building, etc.), each and every wife with dependants will compose the basic unit for production and consumption, i.e., a household (nyakal). As such, we may say that each nyere contains a variable number of uterine families (mother-centred units) (ref. Dyson-Hudson and McCabe, 1985). The kinship segment of a nyere is also the herd-owning unit, and the head of it will be in charge of the common herd (what Neville Dyson-Hudson, 1980, calls ‘the herd complex’). The unity of the compound is frequently broken when the old father dies. One or more of the group of brothers may then decide to leave the compound and settle somewhere else with his family and his portion of the family herd. It is not possible for a man to leave the nyere of his father as long as the father is still alive.

Cattle and the sociocultural system Culturally, the Toposa belong to what Gulliver has denoted the ‘Karimojong Cluster’, including also the Jie, Dodos and Karamoja of Uganda; the Turkana of Kenya; and the Nyangatom of Ethiopia (Gulliver, 1952). Other authors denote this group ‘Ateker’, including then also the Iteso of Uganda (see, e.g., Dyson-Hudson and McCabe, 1985; Müller, 1989). A common feature for the mentioned groups is that they can all be characterised by what Herskovits (1926) has called the ‘East-African Cattle Complex’, and Evans- Pritchard (1940) has called the ‘Bovine Idiom’. Both characterisa-

NUPI OCTOBER 00 246 Arne Olav Øyhus

tions imply that livestock plays a significant role in their economy and culture. Even though many men may participate in farming activities, their existence and attitudes are overwhelmingly determined and dominated by livestock, particularly cattle. Among the Toposa we may analytically discern four major areas in which cattle play a dominant role. The first is economic, securing the supply of essential foodstuff, such as milk, blood and meat. The second is political, maintaining and increasing the social units’ strength, size, prestige and power through marriages and other forms of social alliances where cattle are exchanged. The third is ecological, providing a means of adaptation to a harsh and unpredictable natural environment by converting non-consumable natural resources into edible resources for human consumption. The last is ideological, bestowing meaning, motivation and explanation to Toposa customs and style of life (nyepite kangitoposa, see below). All these areas are closely interwoven, and – in essence – they all add up to securing the social, spiritual and material livelihood of the group both within a long- and a short-term perspective. There are several rules connected to a man’s rights to his animals. A Toposa man can be rich in cattle in the sense that he commands a big herd divided between many stables. But this wealth is not a personal and private matter, and he cannot use the individual animals for whatever purpose he likes. Several other individuals regularly have a ‘pledge’ or ‘share’ in each and every head of cattle under his management. There is thus a series of criss-crossing pledges between individual herd owners. These pledges can lay dormant for an extended period of time, especially if the relationship between the pledger and the pledgee is good. But it can also be activated under certain conditions. The herd owner hence ‘owns’ the cattle as long as he utilises or manages them according to pre- scribed standards. If he breaks these standards, i.e., if he utilises the cattle in an unacceptable manner, he might run the risk of losing his usufruct rights over them. If he, for instance, should sell a head of cattle at the market, or use it otherwise inappropriately, the pledges would – most certainly – be activated. As Erik Eriksen describes it: ‘We must bear in mind that neighbours and close kin know every individual animal a man owns, so every movement one makes in the management of the herd can be checked’ (Eriksen, 1978: 14). A general feature of sociocultural systems based on cattle keeping in arid environments is that they are totally dependent on mobility and flexibility. After a broad literature survey on pastoral peoples, Jere Gilles and Jerome Gefu conclude: ‘The need for mobility…and

NUPI OCTOBER 00 The Innocent Developer 247 the risk resulting from a hostile environment all influences the cultural and social organisation of pastoral peoples’ (Gilles and Gefu, 1990: 102). This is also typical for the Toposa. Heavy, rigid and powerful social structures would quite certainly have a contra-productive impact on economic efficiency and ecological adaptivity since herders depend on quick movements and few restrictions to utilise the natural resources, especially pastures and water, optimally. Also among the Toposa’s ‘brothers’, the Turkana, this organisational ‘flux and flexibility’ is said to represent an adaptive strategy to cope with the adverse environmental conditions (see Dyson-Hudson and McCabe, 1985). One particular sociocultural aspect connected to cattle keeping should be emphasised in this connection. Under the prevailing social, technological and economic conditions of the Toposa, it is not ‘rational’ for a family to accumulate big tracts of land. Since culti- vable land is a plentiful resource for anybody to use freely, a family can only utilise as much land as it can dig by itself. Regarding livestock the situation is quite opposite. It is practically possible to accumulate livestock in almost immeasurable numbers as long as the herding capacity of the herding unit is adequate. And herding capacity can easily be increased by use of various means, for instance marrying an additional wife, or by utilising the labour capacity of the poorer segments of the population. In addition, it is also purposeful to accumulate large numbers of animals since they can be converted into wives, children, food, social security, prestige and power. Therefore, it is a goal among all Toposa men to accumulate as much livestock as possible. And while some in the community, by chance, luck, skill or heritage are tremendously rich, we find that others are exceedingly poor. The poor are not at all outcasts in the Toposa society but play an important role as herders for the rich. In many respects they exist together in a kind of patron-and-client relationship. No institutionalised hierarchy of social strata based on cattle wealth is found. A poor elder is, for instance, as much a member of the village council as a rich. An individual’s economic status is, in principle, never fixed. The son of a poor herder has the possibility of building up a large herd; a rich herder can lose his herd completely due to theft, drought or disease. It is also possible for a poor herdsboy to marry the daughter of a rich herder, and through this relationship get access to animals. It is important to notice in this context that it entails no ideological stigma to be either poor or rich, but obviously, rich herd owners are a pride both to their community and their clan.

NUPI OCTOBER 00 248 Arne Olav Øyhus

Toposa socio-political structure

Organisational groupings in the Toposa society Within the Toposa society we find a broad variety of organisational groupings based on three basic, but criss-crossing and integrative principles: kinship, age and territory. These principles serve a common function of regulating the composing social units’ relationships to each other, to their neighbours and to the natural environment.

Kinship Within the kinship system, the clan (nyateger) is the most dominat- ing feature. The clan is to be looked upon as a huge family where each member has a personal responsibility to the clan as such, and to each of the other members according to their individual status within the clan. A clan in Toposa is a multifunctional social unit dealing with religious, judicial, political and economic affairs. Each clan identifies itself as a people with the same ancestral origins. The clans are exogamous, they have a particular cattle brand sign, and a set of regulations, customs and rituals (ngitalio) to be observed. Each clan has a clan leader, and clan leadership is hereditary. The leader has the ultimate power to pronounce a curse (ngilam) on whoever breaks clan regulations. He represents the clan in the elders’ council. At certain intervals particular clan meetings (nyabela) are arranged. A clan meeting is a very formal forum where clan elders discuss topics of common concern. Most of the clans are distributed throughout all sections and subsections of the Toposa as sub-clans or lineages. At these intermediate levels, the oldest clan member represents the sub-clan in the local councils. His work is particularly associated with the marriage of clan daughters, movements of cattle during the dry season, officiating clan rituals and ceremonies, representing clan members in trials regarding punishment, compensations and justice. If a clan member is injured or killed, the leader, in consultation with other elders, determines the type of compensation to be made. The clan leader at each level officiates the ritual when the cattle are branded with the clan mark. The clan, represented by its leader, is – in principle – the owner of the complete herd of livestock owned by the separate clan members. The lowest level of the clan is the (extended) family inhabiting a compound (nyere).

NUPI OCTOBER 00 The Innocent Developer 249

Organisation according to age The organisation of age is based on two fundamental principles, succession of generations and grading of age. These principles are concretised into two types of age divisions, generation-sets and age-sets. The former is, in essence, rather simple as it implies that sons follow the generation of their father. The latter is a bit more complicated as it includes several age gradings. The structure and content of the Toposa age organisation can only be comprehended in the context of the essential values of nyepite kangitoposa (see below).

Generation-sets Generations or generation-sets (ngasepicho) are the most basic form of age organisation at the sectional and subsectional level. According to Toposa mythology the first people on earth, the ngisiokop, dropped from heaven and formed the first generation. The sons of the first generation formed the second generation, and so on it has continued up to the present generations (Kitonga, 1985). What we see, in actual fact, is that each generation-set is composed of real and classificatory ‘brothers’, and all their sons constitute the following generation-set (Müller, 1989). The same generation-sets exist throughout all the sections of Toposa. To understand the age composition of a generation-set, we should recall that the Toposa have a polygamous marriage pattern. A fundamental feature of this pattern is that a man has to pay a certain number of cattle to his wife’s family as bride price. Due to the fact that old men have most cattle, it is not uncommon to meet rich, old men who have a substantial number of wives. These wives may vary greatly in age. While the oldest might be 60 years old, the youngest can be as young as 14. Although these wives are not always impregnated by their husbands (i.e., the old men are not always the genitors of their children), these men will always and unquestionably be the lawful fathers (paters) of the children born by their wives. Consequently, we may find that an old man at, for instance, the age of 70 years may have sons ranging in age from 50 to 1. All these sons belong to the same generation-set, namely to the generation-set next to their father’s. As this skewed age distribution among brothers within one generation-set may be transferred and repeated by their sons in the next, and then again in the generation thereafter by the sons’ sons, we do experience that the age

NUPI OCTOBER 00 250 Arne Olav Øyhus

span within a generation-set can be extremely wide, even to the degree that the oldest members of the generation may have past away at the time when the youngest are born into it. As both the (classificatory) ‘grandfathers’, ‘fathers’ and ‘sons’ may produce children at the same time, the result is that you find ‘grandfathers’, ‘fathers’, and ‘sons’ at the same physical age, or even that a ‘son’ is physically older than his ‘father’ (Müller, 1989). At any point of time there will at least be four generation-sets, the ‘retired’ generation (‘the grandfathers’ generation-set); the ‘ruling’ generation (‘the fathers’ generation-set), the up-coming, ‘adult’ generation (‘the sons’ generation-set); and the generation of the immature boys (‘the grandsons’ generation set). The ‘seniority’ of a generation is mainly related to its control over the dynamics of the different elements of the cattle complex, i.e., ‘fathers’ control the ‘sons’ because they are in charge of the basic power resource – the cattle. The transferring of power from one generation to the next takes place at nyasapan, the Toposa generation-set initiation. At this ritual the ‘adult’ generation becomes the ‘ruling’ generation, and is formally given the power and authority to rule the sections and subsections. The former ruling generation then retires. At both the sectional and subsectional level each generation-set will have a leader, a nyekarikon. The nyekarikon is never elected but holds the leadership position by virtue of his seniority. The most powerful person within a section is the nyekarikon of the ruling generation. Also the younger generation-sets have a leader, but not at all with the same influence and power as the leader of the ruling generation. Beside the nyekarikon the speakers (ngikatukok) are the most influential leaders of a generation-set. The speakers are not elected but become leaders because of their eloquent speaking abilities (and their cattle wealth).

Age-sets Each generation-set is subdivided into a number of age-sets, nyanyamet (‘those who are eating meat together’, i.e., each age-set contains members of approximately the same physical age). The basic principle is that coevals are combined into groups, and that there is a consecutive hierarchy of these groups. The reason for dividing the generation-sets into age-sets can be explained by the fact that the generation-set comprises members with a broad varia-

NUPI OCTOBER 00 The Innocent Developer 251 tion in age. For the daily functioning, the members of the generation-set assemble themselves into groups of coevals. The age-sets that belong to the same generation-set can theo- retically be compared to a sequence of brothers divided into suc- ceeding age levels. The oldest brothers will belong to the first age-set of a generation-set, the youngest will belong to the last age-set. Regarding the relationship between the age-sets we recognise that the power of the senior sets over the junior sets is substantially related to their larger share in the ‘herd complex’ compared to their younger ‘brothers’. The number of age-sets within each generation-set will vary with the stage of development that the generation is in, from one–two in the youngest, to about 9–10 in the oldest generation (Eriksen, 1978). In general it seems that the formation of a new age-set happens every fourth or fifth year when a group of young men reach the age of about 16–22 years. At this time they will consider themselves fit for adult tasks, and will have the strength to claim their ‘rights’.

Organisation according to territory We may say that the Toposa political structure is composed of four territorial levels, each with a formal authority structure. These are the tribe, the section, the subsection and the compound.10 It is at these levels that power and authority are executed. We might therefore say that the different principles of the socio-political structure (age, kinship and territory) are co-ordinated and executed at these levels. Elders used to act as the only political authority in Toposa before the British colonial government took the area under direct administration in the 1930s. At this stage governmental chiefs both at sectional and subsectional levels were instituted in the area. Now- adays, the chiefs (of sections) and sub-chiefs of (subsections) are elected among the adult male population. Very often the candidates are from the largest and most influential kinship groups. In general, there is a positive correlation between the importance of the different kinship groups and the size of their herds, and this correlation is in turn mirrored in the election of chiefs. The chiefs have few means to enforce their authority and need the positive co-operation from

10 This corresponds more or less completely with the four levels of social organisation that Neville Dyson-Hudson (1980) has described among East African savannah pastoralists.

NUPI OCTOBER 00 252 Arne Olav Øyhus

their ‘constituencies’ to achieve their tasks. In this respect it is easier for a chief from an influential family to command respect for their authority. Despite the formal role and function of the chiefs, the elders are still the most important political force in the Toposa community. At each territorial level, the elders, representing their kinship group, form a kind of board or council. The council meets at ngikito (‘the tree’, i.e., the meeting is conducted under a shady tree) to discuss and make decisions on communal matters, such as the movement of animals, raiding activities, initiations, offerings and village ceremonies. As the elders have no authority over each other’s family groups, they have to discuss at length to reach a consensus. Even then some families may choose not to follow the agreed directions. Although the elders do not have any direct means to sanction devi- ating opinions and actions, families and individuals that choose not to follow the major directions of the council of elders may run the risk of being cursed.

Nyepite kangitoposa A concept of particular analytical relevance to Toposa society and culture is nyepite kangitoposa11 (‘the Toposa way or pattern of life’) (Kitonga, 1985). This concept, or idiom, is most important in order to understand the Toposa sociocultural system as the complete Toposa sociocultural pattern can, for most purposes, be sum- marised into this concept. Whatever a Toposa can do or not do, is explained according to nyepite.12 Nyepite is a system of norms, values and beliefs; it incorporates a cosmology and a guide for individual behaviour. The Toposa divide mankind (ngitunga) into two major divisions, ngitoposa (the Toposa themselves) and ngimoe (strangers or enemies). What separates the two is mainly that ngimoe do not live according to nyepite.

11 In the Toposa language ka means ‘to’ or ‘for’, ngi means people, i.e., ngi Toposa means the Toposa people. The term can be written differently, ka ngi Toposa; ka ngi’Toposa; or kangitoposa. Here, I follow Kitonga (1985) kangitoposa. Nyepite can be translated with traditions, way of life or pattern of culture. 12 The Toposa do not cultivate cassava even though they know it very well from the colonial era when they were forced to cultivate it (after which cassava is still called ‘by law’). During the big drought in the mid 1980s the NGO tried to re- introduce cassava as a security crop, but this was not at all readily accepted as cassava was not considered to be a part of the nyepite kangitoposa.

NUPI OCTOBER 00 The Innocent Developer 253

Nyepite is in each generation handed over from the elders to the youngsters. Although God (Nyakuj) is the original source of life and the creator of the Toposa people, the ancestors, acting as middlemen between Nyakuj and the living Toposa, have created most of the content of nyepite. Therefore, they are the ultimate guides and protectors of it. In the everyday life of the Toposa it is the elders who have the closest association with the ancestors, and who have the right and responsibility to see to it that nyepite is kept. Those who break it will be punished by cursing; those who protect it will be rewarded by blessing. It is this ritual power, combined with the fact that the elders are also the formal owners of the cattle – the core material item of nyepite – which make them the most influential authority in the society. From an outside perspective it is possible to argue that the content of nyepite is not fixed, but eternally changing. From an inside perspective, this is not the case. For the Toposa nyepite is, in prin- cipal, eternal and unchangeable. Nyepite of yesterday is the nyepite of today, which again will be the nyepite of tomorrow. What in actual fact takes place is that any new cultural item which is adopted by the Toposa automatically becomes a part of nyepite; and as such, it has always been there. Consequently, the cosmology remains while the composing components change (see, e.g., Barth 1987).13 This description of nyepite may give the impression that the Toposa society and culture have an extremely communalistic character with a high degree of social control in the hands of the elders. This is, however, not the case. First of all, as we have already seen, the Toposa livelihood system is characterised by a high degree of ‘flux and flexibility’, including a high degree of mobility and a very scattered settlement pattern. This makes it difficult to establish powerful, overarching, political structures. To a great extent each compound and its corresponding cattle camp (awi) are an autono- mous economic and political unit making all important decisions regarding the daily running of its activities. I would say that the only social relationship in Toposa which is typically authoritarian in character is that between father and son. Secondly, Toposa culture also incorporates strong individualistic features, i.e., nyepite emphasises individual will and freedom as much as an individual’s rights and responsibilities towards the community.

13 When asking an old Toposa man when they started to grow sorghum, he answered that sorghum has always been there, growing sorghum is nyepite. Look- ing into colonial records it is obvious that Toposa did not cultivate any type of crop on a regular basis in the first part of the twentieth century.

NUPI OCTOBER 00 254 Arne Olav Øyhus

4. Relief in Toposa In 1972, right after the end of the first civil war in the Sudan, a European NGO started relief operations on the East Bank of the Eastern Equatoria Province, the Southern Sudan. At the end of the decade, the relief operations had turned into a huge development programme covering most of the East Bank (about 450,000 people inhabiting 75,000 km2 of land). In the beginning of the 1980s the NGO established an aid programme among the Toposa.

Why aid to the Toposa?

Drought and cattle raiding Since the 1960s two external factors have repeatedly affected the Toposa society in a negative direction: drought and cattle-raiding activities. While the poorer segments of the population suffered almost continuously throughout this period, even the middle segments had great difficulties in coping with the situation in certain years. In the first part of the 1980s some of the eastern sub-tribes (sections) had lost most of their animals. Families who had lost their animals were in an extremely vulnerable position with their physical livelihood seriously threatened.

Seasonal hunger gaps While the access to food seriously affected the majority of the Toposa households in years when droughts occurred, lack of food was experienced on a more regular basis for the poorer segments of the population, i.e., the cattleless households. This was particularly the case during the annually occurring ‘hunger gaps’. A key condition creating these hunger gaps was the low productivity of the farming sector. The producers were not able to cultivate enough food to carry the households through the whole annual cycle, i.e., until a new yield was obtainable. Quite regularly, last year’s harvest was already consumed when next year’s cultivation process was at its peak. As these poor households could not depend on livestock for additional food, their livelihood situation was precarious. The traditional Toposa mode of adaptation meant that the majority of the people could utilise the food resources in the peripheral areas during the dry season, when they were living in the cattle camps. By this, they could save a substantial portion of the crops produced in the central area. In addition, most of the milking cows

NUPI OCTOBER 00 The Innocent Developer 255 were brought to the permanent settlements during the rainy season. Both of these sources were very important as food supplies during cultivation. However, in the present situation with a substantially lower livestock population, and with farming adversely affected by a drier climate, food insecurity had become a regular feature for an increasing number of households. To sum up, what we therefore see is that the Toposa livelihood system had entered into a vicious circle with a steadily increasing number of households seriously undernourished during the production process. This situation was adversely affected by the fact that at regular intervals the women, instead of cultivating their fields, had to leave their homes and families for long periods in order to search for ‘wild’ food – roots, leaves, berries, etc. – to feed their families. When the annually occurring hunger gaps were combined with the negative effects of droughts, regular situations of famine emerged. Many Toposa families suffered tremendously, and a lot of people died from hunger, especially the very young and the very old. This was the situation both in 1981 and during the big Sudano- Sahelian drought in 1984 and 1985.

What type of aid?

Community survey Due to the gravity of the situation the NGO intensified its activities in the Toposa area from the early 1980s. In order both to have an effective planning instrument, to map available resources in the area, and to make sure that aid activities would be relevant to the community, a community survey (‘baseline study’) was conducted. The survey covered household living conditions, agro-ecological factors and sociocultural issues. The study clearly revealed the central economic and cultural importance of the pastoral sector. However, instead of concentrating its activities on the pastoral sector the NGO decided that development efforts should be geared towards the farming sector. Three arguments influenced this decision. Firstly, the NGO’s development experts had limited competence in tropical animal husbandry, especially regarding the nomadic type performed by the Toposa. Secondly, it was believed that an emphasis on agricultural improvements would give the quickest results. Thirdly, the same experts believed that in the long run the farming sector had a greater development potential than the pastoral sector. An argument which was used to prove the lack of deve-

NUPI OCTOBER 00 256 Arne Olav Øyhus

lopment potential of the pastoral sector was that the Toposa were utterly unwilling to sell their animals, even when the demand from the market was high, and prices good. There was also an underlying assumption that a stronger emphasis on agriculture could lead to a sedentarisation process, which by many was looked upon as positive in a developmental perspective. Through the study several constraints affecting the productivity of the farming sector were demonstrated. The most important were associated with the difficult climatic conditions. It was revealed that the area, on average, was struck by drought every third year. An- other constraint pertained to work efficiency due to the character of the soils. Although the soils are fertile, they are very heavy and sticky when wet. For the women, equipped only with a hoe, cultivation was an extremely strenuous job. So, despite the high soil fertility, productivity was quite modest.

Development objectives and activities Based on the survey a development plan with two main objectives was prepared. The first objective was to improve the working conditions for women; the second was to increase food security, especially among the poorer strata of the population. To reach these objectives it was decided to focus on three technological, innovative activities. The first was to substitute some of the medium to long maturing local sorghums (average growing periods of six months) with quick maturing grains (‘stop-gap’ crops with an average growing period of three months) in order to alleviate the worst effects of the hunger gaps. The second was to introduce drought-resistant root crops (cassava and sweet potatoes) as ‘security’ crops. The third was to introduce ox-cultivation in order to increase work efficiency. Although all three activities were considered equally important, the activity which created most excitement was the introduction of ox-cultivation. Before embarking on this enterprise a warn- ing was given by some researchers knowing the Toposa culture: the possibility of utilising oxen for cultivation was slim. Various studies and surveys pointed unambiguously in one direction – oxen were ‘sacred’ for the Toposa. Most Toposa men had, for example, a ‘favourite oxen’, i.e., an oxen that symbolised and represented its owner. It was an identity relationship between a beast and its owner. The young men made songs to praise their oxen, and it was common that both the man and his beast were called by

NUPI OCTOBER 00 The Innocent Developer 257 the same name (see also Müller, 1989; Eriksen, 1978; and Kitonga, 1985). An interesting fact about these oxen was that they were not, as first believed, the breeding bulls within their respective herds, but quite the contrary – they were all castrated. This was an important technological detail since castration was a precondition for using the oxen for ploughing.

The introduction process Based on information from the survey it was decided that the development experts should use an extended amount of time to discuss the ox-cultivation project with the local councils of elders. This strategy of ‘making haste slowly’ had a typical universalistic ethical explanation, learned from different text books and manuals regarding extension activities in tropical countries (see, e.g., Savile, 1965). This literature clearly proclaimed the idea of ‘local participation’, i.e., that nothing can really be attained if community structures and institutions are not involved in the introduction of a new technology. The importance of community leaders, in this respect, was particularly acknowledged. The experts participated in several ‘sittings’ together with the councils of the elders within the different sections. Here, they were exposed to various technical and cultural constraints raised by the elders. During this whole process it was the elders who led the discussions. It was they who had to decide whether the innovation could be accepted within the framework of their sociocultural system (nyepite kangitoposa) or not. After more than two months of discussions a conclusion was reached – ox-cultivation was not contradictory to nyepite. Conse- quently, before animal traction was introduced to the Toposa it had been thoroughly discussed and ideologically accepted by the indigenous authorities. A social consensus was reached allowing animal traction to be admitted into their sociocultural system. After the completion of this awareness and acceptance process, a yoke of trained oxen was brought to Toposa for practical demonstration. The demonstration was a big event gathering several hundred people. The elders were particularly impressed by the fact that the oxen could take commands from the ox-drivers. Following the demonstration a local chief, as the first one, listed one of his sons for training. In the months to come the list was filled up with more and more trainees.

NUPI OCTOBER 00 258 Arne Olav Øyhus

The gender contradiction Even before the Toposa had accepted the new technology the development experts saw several constraints pertaining to the introduction ox-cultivation. Some of these were of a typical mechanical or economic nature. One constraint was, however, considered to be more important than the others; namely that women, the traditional cultivators, were completely separated from the cattle sector. Even milking was considered to be a typical male activity. To introduce oxen into the farming sector could lead to a situation where the women were excluded from the same sector. Taking the contradiction between the male-dominated cattle sector and the female-dominated farming sector into consideration, the development workers saw that at this stage of the introduction process two different options existed, either to engage men in farming, or – alternatively – to engage women in animal husbandry. As men had a full foot in the livestock sphere, and a toe in the farming sphere, it was felt that it would be easier to trigger the technological adoption process by giving priority to them.

Correcting the male bias As the ox-cultivation programme was progressing and expanding, the recognition that an asymmetrical relationship between men and women had been brought into the agricultural system became more and more evident for the experts. If the new system of production was allowed to continue it was feared that women, in the long run, would lose a lot of their social and economic authority and control. The only way to secure the continued importance of women would be to introduce them to ploughing. The first step that the experts took in this direction was to arrange a new series of discussions with the councils of elders. This time, however, the response was negative. The combination of women and oxen was certainly not nyepite. Although the elders had no formal authority to reject female ploughing, they would never allow their daughters or wives to be trained for this purpose.

The dilemma Now, the external development experts faced a dilemma. On the one hand, according to their own development ideology – and text book theories – it was fundamentally important to respect local sociocultural structures and institutions. And there was no doubt

NUPI OCTOBER 00 The Innocent Developer 259 among the elders, representing both the highest spiritual and secular authority within the Toposa society, that the idea of training women for ploughing was ‘ethically’ wrong, since it signified a break with nyepite. On the other hand, women issues were also of vital concern to the experts’ development ideology. And obviously, women played a key role in the Toposa economy as agricultural producers. For the poorer strata of the population women were possibly the main economic agents. The economic importance of women also gave them a strong social influence. If one recognised the core role of women in Toposa economy and society, to block their access to the new technology would most certainly undermine their socio-economic position. This was certainly not ‘ethically’ permissible according to the experts’ ideology. After carefully balancing these two positions the development experts made a choice: in this case they had to work against the will of the elders rather than create a situation where women could lose their role as important socio-economic actors. Consequently, it was decided that a particular training programme for female ox- cultivators should be introduced. As no elder wanted to register his daughter or wife for training, the focus of attention was put on a group of women living at a nearby Catholic mission station. Most of the adult women were widows, most of the younger were orphans, and all of them were typically marginal to the Toposa sociocultural system. 11 women of the right age and status (they were all mothers) who were interested and motivated were selected for training. Thereafter, the training programme for female ox-cultivators started. As time went on, Toposa women in increasing number approached the NGO staff in order to be admitted into the training programme. Due to the high demand for training of both men and women, the NGO established several training centres throughout the Toposa area.

5. Consequences of Adoption Due to extraordinary circumstances it is not possible to judge the full consequences of this technological development programme. During the later part of 1985 rebel forces from the SPLA14 entered

1 4 The Sudan People’s Liberation Army, the military wing of the SPLM, the Sudan People’s Liberation Movement. The SPLA started their military campaign against the Sudan government in 1983.

NUPI OCTOBER 00 260 Arne Olav Øyhus

the Toposa area. While the Toposa fled to the cattle camps, the NGO closed down its activities. The ‘successful’ programme of introducing a new technology to the pastoral Toposa came to an instant halt, and has never since been resumed. Due to this particular situation we may argue that the consequences for the Toposa society of adopting the new technology were minor. But it is possible to hypothesise what the probable outcome could have amounted to under more ‘normal’ circumstances, and what type of ethical considerations would have come into focus. The ethical considerations I want to raise for discussion here will, for analytical reasons, be related respectively to sociocultural and adaptive consequences of adoption.

Sociocultural consequences of adopting the ox-plough With the ox-plough the Toposa were introduced to a technology which had its origin in a quite different mental, social, technical and economic setting than their own. The analytical point to underline here is that no technology is valueless, it is never culturally ‘neutral’. For a group like the Toposa to adopt a new technology was to adopt a part of a cultural pattern that they had no knowledge about in advance, and which was completely different from the pattern they knew. Still, it can be argued, the plough was formally discussed and approved by the local authority institutions, following their own rules of the game. Consequently, the adoption was a ‘voluntary’ action by the Toposa. This is formally correct, but it does not mean that it is ethically right. The new cultural item was accepted by the elders who, in reality, did not know what they accepted. It was far beyond their indigenous ‘corpus of knowledge’ to perceive the consequences of the new technology. The actual agents behind the decision to adopt the plough were, in actual fact, the outside experts. Their interest in promoting the plough must be understood in the perspective of two basic conditions. Firstly, their technological comprehension, perceiving ox-cultivation as a technical means that could substantially improve food production among the Toposa. Secondly, their universalistic ethical standards, stating that they had a moral obligation to assist the unfortunate Toposa in the present, dramatic situation, even at the ex- pense of some possible negative effects – as articulated by the elders. Let us now look at this situation from a ‘relativistic’, or idiosyncratic, perspective. In Toposa culture and language there is no such

NUPI OCTOBER 00 The Innocent Developer 261 concept as ‘progress’ or ‘development’: yesterday, today and tomorrow represent a kind of continuous, circular ‘now’. People are born, people mature, people breed, people wither, people die, but nothing really changes. The Toposa world and their existence remain the same. How then could the elders, the living guides and protectors of nyepite, accept the plough as it was obviously (and objectively) a new cultural item? A culture is a flexible, mental grid where items come and items go even if the bearers of the culture themselves are not fully conscious about this fact. When the plan for introducing ox-cultivation was presented to the Toposa elders, what they probably discovered during their discussions was that the plough had been there all the time, but that it was a kind of ‘dormant’ item which had existed in the cultural shadow because they had had no actual need for it. Due to the changed circumstances the ancestors had now brought the item into the cultural limelight. Even though the Toposa do not have any concepts regarding progress or development, they know very well that circumstances shift, from family to family, from district to district, from day to day, from season to season, and from year to year. Now, the present circumstances characterised by drought, warfare and famine im- plied that some formerly, non-utilised cultural means had to be activated. One of these means was the plough, originally coming from the ancestors, but now brought to their mind by some strangers. Reaching this conclusion, the Toposa felt sure that the plough could only have beneficial consequences as any nyepite item per definition is good. Then, after accepting this item, a new cultural theme was presented to the elders, women as ox-cultivators. Women and cattle, which from times immemorial had belonged to two distinct spheres, were now to operate together in a unified cattle-farming sphere, or rather unified masculine-feminine sphere. The Toposa elders treated the subject seriously, discussing it for a lengthy period of time, but the ‘verdict’ was clear, female ox-cultivators were not nyepite, quite contrary, they represented a serious break with it.

Adaptive consequences of the new technology It is a long tradition among anthropologists to interpret pastoralists’ spatial and social behaviour as adaptive responses to specific features of the natural environment (see, e.g., Barth, 1956; Evans- Pritchard, 1940; Gulliver, 1951; Gulliver, 1968a; Gulliver, 1968b).

NUPI OCTOBER 00 262 Arne Olav Øyhus

When it comes more specifically to the East African savanna pastoralists, we notice that Neville Dyson-Hudson has described them as living under environmental conditions of risk and uncertainty. To survive they have therefore developed a ‘pastoral strategy’ where they manipulate the human population, the livestock population and the physical environment in order to respond to changes in the environment (Dyson-Hudson, 1980). As already stated, the natural environment of Toposa is dry, hash and vulnerable. More than half of it is actually a semi-desert that can only be exploited by the use livestock (JIT, 1954; Lebon, 1965; Øyhus, 1992). Within this environment the Toposa have for generations developed a particular mode of adaptation very much characterised by Dyson-Hudson’s ‘pastoral strategy’. Earlier, when some factors in the social or natural environment shifted, the in-built flexibility of their pattern of adaptation had meant that the viability of the Toposa as such had never been seriously threatened. Under traditional, ‘normal’ circumstances the Toposa had, by applying their skills, knowledge, technology, partners and alliances, been able to utilise the environment optimally in order to secure the viability of the comprising social units. But now, due to external and uncontrol- lable changes in their social and natural environment, and articulated through the means of the modern state; modern weapons; and adverse climatic conditions, the livelihood for a big segment of the population was under an acute and chronic threat. In a long-term perspective the increased vulnerability meant that the viability of the Toposa society as such was at stake. To cope with the shifts in the environment, changes in their traditional mode of adaptation were required. The Toposa had to adapt to a new environment. But how? Did ox-cultivation represent a type of technology that could secure their future? On the positive side, it is possible to perceive that ox-cultivation could increase agricultural efficiency to such a degree that both in good years and less-good years yields would increase substantially. On the negative side, the introduction of ox-cultivation could start a process of sedentarisation among the Toposa, a process similar to what we have already seen happen to several pastoral groups both in East and West Africa. Even though this process by many NGOs, governments and multilateral organisation is looked upon as favourable, it certainly has its weaknesses. Several studies report negative social and economic consequences of pastoral sedentarisation (Galvin, Coppock and Leslie, 1994; Hill et al., 1985; Fratkin, 1997). Among the Rendille in Kenya, for instance, it was found that

NUPI OCTOBER 00 The Innocent Developer 263 settled children were more malnourished than nomadic children (Nathan, Fratkin and Roth, 1996). An important aspect of the Toposa pattern of adaptation is that the animal sector can utilise marginal, non-palatable, natural resources over a vast geographical area, and convert them into food resources with a high nutritional content. Paul W. Baxter has convincingly argued that pastoral systems ‘are certainly the most efficient means yet devised to produce food in arid zones’ (Baxter, 1993: 147). What sedentarisation most probably would imply for the Toposa was either that the livestock sector would disappear altogether, or that it would shift in structure from nomadism to sedentary ranching (or perhaps short cycle transhumance). Any of the two possibilities would certainly imply that the animal sector would diminish in importance since a large proportion of the natural resource foundation no longer could be exploited. A new pattern of adaptation with a less important livestock sector would most possibly imply a weaker economic foundation for the whole society, especially in years with rainfall below normal. Par- tially, it would be possible to compensate for this change by the application of more drought-resistant crops. But in years with rainfall far below normal, no cultivation at all would be possible (as the situation was in 1984). The lower limit for crop cultivation in Toposa is 6–700 mm rainfall per year if the rains are evenly distributed (Øyhus, 1992). With less rain, or with erratic rainfall, the crops will not grow and mature. But grasses, shrubs and bushes will grow even with rainfall down to 1–200 mm per year, giving nutrition to goats, sheep and cattle. This situation is regularly experienced both in East Africa and in the Sahel zone (see, e.g., Gulliver, 1951; Gulliver, 1968; Dyson-Hudson and McCabe, 1985; Bernus, 1990; and Lancaster and Lancaster, 1990). As Neville Dyson-Hudson argues, the herds are not only the instruments for exploiting the environment but the means by which some of its fluctuations are dampened down (Dyson- Hudson, 1980). Therefore, what we experience is that in the Afri- can drylands the animals represent a higher level of food security than do cultivated crops. A corresponding negative effect of a more sedentary pattern of animal husbandry would, most probably, be a higher concentration of livestock around the permanent settlements. Potentially, this may cause important damages to the natural resource foundation since a high concentration of livestock within a limited area will increase the competition between farming and husbandry for scarce resources. A situation with serious land degradation might be a probable outcome.

NUPI OCTOBER 00 264 Arne Olav Øyhus

We must also keep in mind that cropping increases ecological risk. By removing the natural vegetation and replacing it with ‘alien’ crops, the cultivator introduces an element of instability into the system. Generally speaking, the more artificial the ecosystem, the higher the instability and the more it will be dependent on outside resources to compensate for this instability (de Bruijn and van Dijk, 1999). For the Toposa, a possible (and probable?) consequence of a ‘modernised’ pattern of adaptation could therefore be a less flexible livelihood situation signified by fewer and less secure sources of food, and food with less nutritional value. Thus, instead of securing the viability of the Toposa society, the new technology could possibly make it less robust and more vulnerable.

6. Conflicting Ethics – Conflicting Perceptions The introduction of ox-cultivation to the pastoral Toposa represents a typical conflict in ethics. From an external, ‘universalistic’ perspective, introducing ox-cultivation to the Toposa was good, introducing it to women was even better. From a ‘relativistic’ perspective, i.e., according to the idiosyncratic Toposa worldview, the introduction of ox-cultivation was good since it was considered to be in accordance with nyepite, but introducing it to women was not at all good since this represented a break with nyepite. As already mentioned, the Toposa have a strong belief in the power of Nyakuj (God) and the ancestors regarding the protection of nyepite. Any evil thing happening in the society can be explained in terms of punishment from Nyakuj or the ancestors for breaking the rules of nyepite. During the process of training women for ploughing a serious incident took place in Toposa – they were in- vaded by the SPLA rebel forces. We should here add that the in- vading SPLA troops consisted almost entirely of soldiers from the Dinka ethnic group. The Toposa have been in hostile conflict with this group for decades regarding the utilisation of the important pastures north and north-west of Jebel Kathiangore, situated approximately 100 kilometres north of Kapoeta town. From a ‘universalistic’ perspective, this invasion was perfectly explainable as a coincidence related to the on-going civil war. It had nothing to do with the Toposa people, as such. From a Toposa perspective such a serious coincidence could never happen by acci- dent, but only according to the will of Nyakuj and/or the ancestors. Consequently, it would be logical for the Toposa to interpret the rebel

NUPI OCTOBER 00 The Innocent Developer 265 invasion, headed by their arch enemies, as the ancestors’ punishment for some serious offence against nyepite, for instance that women had entered the masculine cattle sphere against the will of the elders. As Dyson-Hudson and McCabe (1985) argued regarding the Turkana, in a harsh environment where disasters are frequent, there are many validations of a curse – drought and famine, enemy raids, illness and death of people or livestock, and so on.

7. Conclusion Today, the ethical foundation of mainstream development aid is characterised by a universalistic perspective where development practice to a substantial degree is seen in the light of modernism. Basically, modernism perceives technology as something valueless and neutral, and technology development as a process that is both inevitable and necessary. Modern technology can, in this respect, be compared to a man on a bicycle, if he does not tread the pedals, he will fall over. The point of departure for a relativistic, ethical perspective is that technology is both a system in itself, and a component within a more comprehensive cultural system. To understand the role of technology we have to analyse both its inner dynamics and its con- nections to the broader social and cultural processes. In this respect, modern technology constitutes a particular challenge to traditional cultural patterns since its inherent dynamics quite often is contradictory to the limitations that cultural identity and a common system of values and norms imply. All societies go through a continuous process of change. Even small, tribal communities show this inner dynamics. As every society changes, so does the world. In relativistic terms, the hope is that the direction of global change should mirror the values that the different peoples and nations want to achieve, i.e., that the development process becomes plural rather than singular. As Fredrik Barth argued many years ago, cultural variation is utterly important because different cultures represent ‘experiments in living’ (Barth, 1972). But cultural variation is a global resource that today is threatened by the ‘technological imperative’ of modernism (Goulet, 1989). The introduction of ox-cultivation to the Toposa is an example of a development practice initiated by external experts with a technological understanding based on universalistic, development ethics. Using their own value system as a reference, initiatives were taken and decisions made which only took the practical aspect of the

NUPI OCTOBER 00 266 Arne Olav Øyhus

new technology into consideration. Although experiencing local resistance the introduction of the technology was perceived, more or less unconsciously, to be a culturally neutral activity without any ethical bearings. A relativistic critique of this particular development practice will proceed along two directions. Firstly, regarding the development experts’ perception of the new technology as an important means to enhance food security and social development among the Toposa. Several authors claim (see, e.g., Ellis and Swift, 1988) that due to the particular physical environment of the pastoralists, development policies should build on and facilitate traditional strategies rather than constrain them. This is because traditional practices are more appropriate for arid, unstable ecosystems than other, more modern practices. Secondly, the critique is directed against a view or a perception that it is possible to introduce an alien, technological item with no, or insignificant, consequences to the local symbolic system and pattern of adaptation. It is a general fact that people constantly have to respond to new circumstances. They do so within the framework of their cultural understanding of reality (Croll and Parkin, 1992). This implies that people utilise social and cultural means such as organisations, knowledge, religion and ideologies to counter insecurities. We have few reasons to believe that solutions generated by development experts that do not take account of these cultural means will be viable. As Mirjam de Bruijn and Han van Dijk (1999) argue, we must take into consideration that pastoral societies have borne the brunt of drought, civil war, economic marginality and development disasters over the past decades. Those who have survived are those who have been able to deal with these conditions by applying their local means, including their indigenous ethics.

References Ambrose, S.H., 1984, ‘The Introduction of Pastoral Adaptations to the High- lands of East Africa’, in J.D. Clark and S.A. Brandt, eds., From Hunters to Farmers: Considerations of the Causes and Consequences of Food Produc- tion in Africa, Berkeley: University of California Press. Azarya, V., 1996, Nomads and the State in Africa: The Political Roots of Margin- ality, Aldershot: Avebury. Barbour, K.M., 1961, The Republic of the Sudan, London: University of London Press. Barth, F., 1956, ‘Ecological Relationships of Ethnic Groups in Swat, North Paki- stan’, American Anthropologist, Vol. 58, No. 6.

NUPI OCTOBER 00 The Innocent Developer 267

Barth, F., 1972, ‘Ethvert samfunn må forstås på sine egne premisser’, Forsknings- nytt, 17. årg., Nr. 4. Barth, F., 1987, Cosmologies in the Making, Cambridge: Cambridge University Press. Barth, F., 1991, Andre liv – og vårt eget, Oslo: Universitetsforlaget. Baxter, P.T.W., 1993, ‘The “New” East African Pastoralist’, in J. Markakis, ed., Conflict and the Decline of Pastoralism in the Horn of Africa, London: Mac- millan. Bernus, E., 1990, ‘Dates, Dromedaries, and Drought: Diversification in Tuareg Pastoral Systems’, in J.G. Galaty and D. L. Johnson, eds., The World of Pastoralism. Herding Systems in a Comparative Perspective, New York: The Guilford Press. Bjørtuft, S., 1984, Norwegian Church Aid – Sudan Programme: Annual Report 1984, Oslo: Norwegian Church Aid. Croll, E. and D. Parkin, 1992, ‘Anthropology, the Environment and Develop- ment,’ in E. Croll and D. Parkin, eds., Bush Base: Forest farm, Culture, En- vironment and Development, London: Routledge. de Bruijn, M.E. and H.J.W.M. van Dijk, 1999, ‘Insecurity and Pastoral Develop- ment in the Sahel’, Development and Change, Vol. 30. Dyson-Hudson, N., 1980, ‘Strategies of Resource Exploitation among East Afri- can Savanna Pastoralists’, in D.R. Harris, ed., Human Ecology in Savanna Environments, London: Academic Press. Dyson-Hudson, N. and R. Dyson-Hudson, 1982, ‘The Structure of East African Herds and the Future of East African Herders’, Development and Change, Vol. 13. Dyson-Hudson, R. and N. Dyson-Hudson, 1969, ‘Subsistence Herding in Uganda’, Scientific American, February. Dyson-Hudson, R. and N. Dyson-Hudson, 1980, ‘Nomadic Pastoralism’, Annu- al Review of Anthropology, Vol. 9. Dyson-Hudson, R. and J.T. McCabe, 1985, ‘South Turkana Nomadism: Copy- ing with an unpredictable environment’, Vol. 1, HRAFlex Books, Connecti- cut: New Haven. Ellis, J.E. and D.M. Swift, 1988, ‘Stability of African pastoral ecosystems: Alter- nate paradigms and implications for development’, Journal of Range Manage- ment, Vol. 41, No. 6. Eriksen, E.,1978, ‘The Ritual Meat Feast – Nyakirit – as a Factor on Food Dis- tribution among The Toposa’, in E. Eriksen, ed., Aspects of Agro-Pastoral Adaptation in East Africa, Bergen: Department of Social Anthropology. Evans-Pritchard, E.E., 1940, The Nuer, Oxford: Clarendon Press. Franke, R. and B. Chasin, 1981, ‘Peasants, Peanuts, Profits, and Pastoralists’,The Ecologist, Vol. 11, No. 4. Fratkin, E., 1997, ‘Pastoralism: Governance and Development Issues’, Annual Review of Anthropology, Vol. 26. Galvin, K., D.L. Coppock and P.W. Leslie, 1994, ‘Diet, Nutrition, and the Pas- toral Strategy’, in E. Fratkin, K.A. Galvin and E.A. Roth, eds., African Pas- toralist Systems: An Integrated Approach, Boulder: Lynne Rienner. Gilles, J.L. and J. Gefu, 1990, ‘Nomads, Ranchers and the State: The Sociocul- tural Aspects of Pastoralism’, in J.G. Galaty and D. L. Johnson, eds., The

NUPI OCTOBER 00 268 Arne Olav Øyhus

World of Pastoralism. Herding Systems in a Comparative Perspective, New York: The Guilford Press. Goulet, D., 1989, The Uncertain Promise: Value Conflicts in Technology Trans- fer, New York: New Horizon Press. Gulliver, P.H., 1951, ‘A Preliminary Study of the Turkana’, Communications from the School of African Studies, University of Cape Town, New Series No. 26. Gulliver, P.H., 1952, ‘The Karamojong Cluster’, Africa, Vol. 22, No. 1. Gulliver, P.H., 1968a ,‘The Jie of Uganda’, in Y.A. Cohen, ed., Man in Adapta- tion, the Cultural Present, Chicago: Aldine Publishing Company. Gulliver, P.H., 1968b, ‘The Turkana’, in Y. Cohen, ed., Man in Adaptation, the Cultural Present, Chicago: Aldine Publishing Company. Hall, C.A.S., 1990, ‘Sanctioning Resource Depletion: Economic Development and Neo-Classical Economics’, The Ecologist, Vol. 20, No. 3. Herskovits, M.J., 1926, ‘The Cattle Complex in East Africa’, American Anthro- pologist, Vol. 28. Hill, A.G. et al., eds., 1985, Population, Health and Nutrition in the Sahel, Lon- don: Routledge & Kegan Paul. Holl, A.F.C., 1998, ‘The Dawn of African Pastoralism: An Introductory Note’, Journal of Anthropological Archaeology, Vol. 17. Ingold, T., 1986, The Appropriation of Nature, Manchester: Manchester Univer- sity Press. JIT (Jonglei Investigation Team), 1954, The Equatorial Nile Project and Its Ef- fects in the Anglo-Egyptian Sudan, London. Kitonga, D., 1985, Toposa Culture and the Communication of The Gospel. M.Sc. thesis, Daystar University College, Nairobi. Lancaster, W. and F. Lancaster, 1990, ‘Desert Devices: The Pastoral System of the Rwala Bedu’, in J.G. Galaty and D. L. Johnson, eds., The World of Pas- toralism. Herding Systems in a Comparative Perspective, New York: The Guilford Press. Lebon, J.H.G., 1965, Land Use in Sudan, Cornwall: Geographical Publications Limited, Monograph No.4, Bude. Little, M.A., 1980, ‘Designs for Human-Biological Research among Savanna Pastoralists’, in D. R. Harris, ed., Human Ecology in Savanna Environments, London: Academic Press. Little, M.A., 1989, ‘Human Biology of African Pastoralists’, Yearbook of Phys- ical Anthropology, Vol. 32. Markakis, J., ed., 1993, Conflict and the Decline of Pastoralism in the Horn of Africa, London: The Macmillan Press. Müller, H., 1989, Changing Generations: Dynamics of Generation and Age-Sets in Southern Sudan and Northern Kenya, Saarbrücken: Breitenback Publishers. Nathan, M.A., E.M. Fratkin and E.A. Roth, 1996, ‘Sedentarism and Child Health among the Rendille Pastoralists of Northern Kenya’, Soc. Sci. Med., Vol. 43, No. 4. Perrings, C., 1989, ‘An Optimal Path to Extinction?’, Journal of Development Economics, No. 30. Repetto, R. and T. Holmes, 1983, ‘The Role of Population in Resource Deple- tion in Developing Countries’, Population and Development Review, No. 4.

NUPI OCTOBER 00 The Innocent Developer 269

Savile, A.H., 1965, Extension in Rural Communities, London: Oxford University Press. Schneider, H.K., 1957, ‘The Subsistence Role of Cattle among the Pakot and in East Africa’, American Anthropologist, Vol. 59. S.D.I.T. (Southern Development Investigation Team), 1955, Natural Resources and Development Potential in the Southern Provinces of the Sudan, London: Sudan Government. Smith, D.V. and A. Ojetuk, 1985, The Way of Fire and Water: The Lotuho Way of Life, Geneva: The Lutheran World Federation. Sperling, L. and J.G. Galaty, 1990, ‘Cattle, Culture and Economy: Dynamics in East African Pastoralism’, in J.G. Galaty and D. L. Johnson, eds., The World of Pastoralism. Herding Systems in a Comparative Perspective, New York: The Guilford Press. Swift, J., 1976, ‘Desertification and Man in Sahel’, Africa Development, Vol.1, No. 2. Toulmin, C., 1991, Bridging the Gap Between Top-Down and Bottom-Up in Natural Resource Management’, in P.W. Baxter, ed., When the Grass Is Gone, Seminar Proceedings No. 25, Uppsala: The Scandinavian Institute of African Studies. Warren, A. and M. Khogali, 1992, Assessment of Desertification and Drought in the Sudano-Sahelian Region 1985–1991, New York: UNSO. Øyhus, A.O., 1992, ‘Adaptivity and Rationality: The Didinga Ethno-Agrarian Mode of Adaptation’, Doctor Scientarium Theses 1992: 19, Agricultural University of Norway.

NUPI OCTOBER 00 270 Arne Olav Øyhus

NUPI OCTOBER 00 Forum for DevelopmentCorruption as Studies Everyday | No.2-2000 Practice 271

Corruption as Everyday Practice. The Public–Private Divide in Local Indian Society

Arild Engelsen Ruud

1. Introduction Corruption has become salon-fähig in the development debates. The focus is mostly on countries in the developing world, which are perceived as being more corrupt, and in part this focus constitutes a new way of constructing a defining division of ‘us’ from ‘them’. From a culturally sensitive point of view it can be argued that the focus on corruption as a ‘problem’ in the developing world prevents us from understanding that these are practices developed within a fully mature normative system of no less moral validity than any other normative system. Moreover, the focus on the corrupt act as an isolated object for study disregards the parallels in other social practices. The act of corruption seen in isolation then becomes difficult to understand. This is particularly actute because the definition of what constitutes a corrupt act is based on a particular body of ideas and values, whereas its practice is in many societies understood on the basis of a different and wider body of ideas. The fundamental starting point for this article is that in the wider context of social life, the simple act of corruption (for the moment understood as an illicit deal involving the holder of a public position) is only one among many outcomes of habitual practices. The corrupt act involves habitual networking, negotiation and mani- pulative application of ideas and moral arguments – it just happens to involve a holder of a public position.1

1 An earlier version of this essay was circulated as SUM Working Paper 1998:4. It has been given at a seminar on ‘Corruption and development’ at Uppsala University in 1998 and at a workshop on Public Administration in the Develop- ing World, at NIBR, Oslo. Insights and valuable comments have generously been offered by Therese Tjeldflaat, Kathinka Frøystad, Ketil Fred Hansen, Dan

NUPI NOVEMBER 00 272 Arild Engelsen Ruud

Compared to widely used definitions of corruption, this approach shifts the vantage point from which we try to understand corruption to the level and places where corruption is practised. I focus on everyday forms, in backwater towns and rural societies, on the widespread petty corruption that involves ordinary people – villagers, lower level bureaucrats, petty businessmen and the local politicians. The material was collected during fieldwork in rural West Bengal, a state in eastern India. West Bengal is not particularly known for widespread corruption although some has come to light lately. The fact that corruption in West Bengal is not as endemic as, for instance, in neighbouring Bihar allows us a view of people’s manoeuvring in a situation of ‘ordinary’ and ‘medium-range’ corruption rather than in an unusual and extreme case. Besides, there is a general awareness among ordinary villagers as well about the illegitimacy of bribe-taking, embezzlement, and favoritism. The second aim is that by switching the focus to the manner in which the common actor relates to such acts and to the moral background to these practices, we may gain new perspectives to the study of corruption.

2. The Common Definition of Corruption In the main, corruption studies have been the domain of political scientists or economists (Price, 1999).2 Anthropology and similar disciplines have not achieved the place in corruption studies that their potential insights and perspectives should warrant. Anthro- pological literature on corruption is scant and very case-study oriented, with little endeavour to arrive at generalisations.3 There are possibly good reasons for this; ethnographic data on corruption are not easily available, and methods such as ‘participant observation’ are not applicable to this particular line of inquiry. Typically the present article is based on ethnographic data that mostly came my way in an incidental fashion. It is also possible that the communication gap has been too wide. The cultural sensitiveness, ethnographic minutiae or esoteric concerns of anthropology do not lend

Banik, Pamela Price, Morten Løtveit, Harihar Bhattacharya, Razmik Panossian and two anonymous referees. All inconsistencies are mine. 2 The otherwise thorough overview of corruption issues in Rose-Ackerman (1999) is strikingly thin when it comes to corruption as a ‘cultural problem’. 3 This is an oversimplification, but it is probably correct to state that anthropology has not been concerned with corruption as such, only with related topics such as networking, kinship, gifts, etc.

NUPI NOVEMBER 00 Corruption as Everyday Practice 273 themselves easily to the policy-oriented requirements of donor- driven corruption research. Communication across disciplinary divisions, however, is best achieved when basic views are made clear. Definitions of corruption abound. One often employed is the standard definition formulated by the World Bank which focuses on the abuse of public power for private benefit (or profit). Conven- tionally definitions centre on this distinction between the individual in his (or her) private capacity as a family member or greedy individual on the one hand, and, on the other, that same individual in his or her public capacity. The ‘public’ element of the different definitions may vary in the degree of emphasis on formal institution or on the state, some definitions may be more legalistic than others. There are other definitions too, but they are variations over a theme. Such definitions can easily be repudiated from a culturally sensitive reading of say African or Indian material. Although they may resonate with deep understandings concerning the role of individuals in the institutionalised state apparatus in the western world, such sentiments may not be readily shared in other societies. One of the few ethnography-based articles on corruption has argued precisely that from the participant’s point of view this distinction is unclear, the division between state and society is ‘blurred’ (Gupta, 1995). Such state- centric definitions have a cultural bias tending to ignore the ‘cultural embeddedness’ of practices grouped under the term corruption. Particular social norms widely represented in modern Africa, for instance, ‘communicate’ with or influence the practices of corruption (Olivier de Sardan, 1999: 26). The practices of Indian politicians, which include a measure of corruption, may be properly understood as a development from precolonial forms of rulership (Price, 1999). The role of the politician is not unascribed but car- ries expectations that are culturally informed by the historic role of the king, for whom the distinction between public and private was irrelevant. In short, the very idea that corruption can be endemic, general and systemic without resonating with deep sentiments is unsustainable. When corruption is widespread, writes Samuel Paul, it reflects ‘fundamental flaws … in the values and traditions of the people of that society’ (Paul, 1997: 1350). The definition of corruption becomes something of an ethnocen- trism. Many different practices are grouped together under the same umbrella: bribery, graft, embezzlement, kickbacks, nepotism, favouritism, extortion, fraud, bending of rules, gifts and ‘considerations’, and ‘commissions’ – all grouped under a term which literally means

NUPI NOVEMBER 00 274 Arild Engelsen Ruud

‘decay’.4 Although these many phenomena appear to have certain traits in common – the blurredness between the private and the public is one of them – they are still very dissimilar phenomena. They are dissimilar in terms of how they are conceived, how they are morally evaluated by the actors involved, and how they are culturally constructed. The moral basis varies widely for each of the different practices. Simply put, to use one’s contacts to help an unfortunate nephew to get entry to a preferred school is not understood to be the same as charging a poor peasant money for hand- ing him a form or embezzling money from a building construction. To term all such practices ‘corruption’ is to paint black what is in reality a variety of colours. This ethnocentric exercise condemns actions and the normative basis of those actions simply on the basis that they are dissimilar to our norms.5 Most of these terms and phrases refer to actions that have separate moral justifications and entrenched normative bases. To help a nephew in need is to most people in the world a laudable thing to do. The notable exception to this rule is people exercising their office in formal institutions – precisely the focus of most definitions of corruption. However, this constitutes a breach in the application of an otherwise universally applicable moral standard – to help your kin. Legally based restrictions on family obligations are not unequivocally accepted in all cultures. Even where the legal restriction is well-known, the moral obligation may well be stronger. From the strong sense of family obligation follows a practice that incidentally conflicts with the needs of the modern state. Is that sufficient to term it ‘decay’? The aim here, though, is not to defend corruption as a practice. Nor do I wish to argue that the focus on corruption is futile or even harmful – although such arguments can be made with some validity. The aim is rather to investigate the practice of the distinction between private and public in a non-western society and equally the practice of crossing and straddling this dividing line. The first part will elaborate on the practice of corruption, in particular the formation and usage of networks, and the second part will seek to deduct some more general implications from the material. The material is centred on a few complex and multipronged examples. The basic thrust derives from situations that reveal how social prac-

4 Certain types of corruption do not necessarily fall within the World Bank definition. Embezzlement, for instance, is legally a form of theft. However, in practice the distinction is unclear. 5 They also have unfortunate consequences for the efficiency of the modern state, but so do many other social practices in our society.

NUPI NOVEMBER 00 Corruption as Everyday Practice 275 tices of networking are helpful in penetrating a recalcitrant state. Such practices are not only helpful as strategic tools that people employ, they are habitual practices (habitus) employed in any number of difficult situations, as routine strategies resorted to. In the discussion, the point will not be to underline the significance of the culture-sensitive point of view, but to explore the practice of the crossing in a situation where the awareness of its illegitimacy is widespread.

3. Networking The practice of networking creates a pool of people that can be expected to be of use when issues arise in the lives of an individual: When searching for a bride for one’s son an extended network is mobilised; when on the look-out for a reliable wholesale dealer for the harvest, inquiries are made to friends and relatives in similar circumstance; when desiring entry to a preferred school for a son or daughter, distant relatives may be solicited for information and contacts; when in need of assistance in dealing with a disagreement with the police, powerful contacts are approached, etc. Nor- mal life represents any number of difficult situations that the individual alone cannot solve. He or she relies to a considerable extent on the social network for economic survival and protection. The network is of course also a matter of social standing, prestige and status, as well as, last but not least, a matter of personal satisfaction and comfort. It is somehow incidental that this widespread social practice also allows the individual to penetrate on occasion the formal and impersonal machinery of the state, whether this may be its bureaucracy or the political system. The difference here is that the difficult situation that requires solution happens to involve institutions of the state. The first example here, involving Uttam and his quest for a schoolteacher position, will illustrate how useful networking may turn out to be. In spite of his university degree, Uttam found it difficult to get a position as a teacher in primary school. The number of vacancies is far less than the number of available well-educated candidates. After years of waiting and applying to individual schools he went to the Calcutta-based government employment exchange in order to have his name placed on lists of available candidates, lists that are sent to primary schools with vacancies. But the intimidating bureaucracy dismissed him with a host of rules and demands for certifi- cates, and then (after weeks of back and forth) proceeded to place

NUPI NOVEMBER 00 276 Arild Engelsen Ruud

him so far down the list that he would have the prospect of a job offer only after several years. Dispirited he lingered in the corri- dors, chatting with the other hopefuls there. He accidentally met someone who – during the obligatory chatting over tea in which Bengalis so eagerly engage – turned out to be the distant relative of someone Uttam knew from his university time. Uttam was an excellent ‘networker’. He kept close tabs on a very large number of people, kept updated address lists, corresponded widely, and had a particular flair for names. He was also quite charming in his own way and was for instance invited to more weddings than most people. This connection that Uttam could make to a fellow student several years back was sufficient for the chance acquaintance to divulge how he had found his way around the same intimidating bureaucracy; the right person to bribe, and the right way of approaching him. The acquaintance also went out of his way to arrange a meeting for the following week and helped negotiate the details of the transaction (standard sum, standard procedure). The meeting was successful and Uttam then got listed to several schools over the following months. The problems that the individual actor faces when addressing government institutions are very often insurmountable, and so he is locked in a situation of permanent disadvantage. However, networking, establishing social relations and maintaining these, is a crucial coping strategy that Indians employ in dealing with and penetrating the formal and distant government machinery. Bengalis and other Indians spend hours chatting (adda mara), meeting friends or relatives at tea-stalls or under the shade of a tree.6 They talk about current events, national and village politics, gossip about common acquaintances and discuss the practicalities relating to an enormous range of issues. They eagerly include the friend of a friend in their group of gossipers, and interrogate with gusto any stranger recently introduced about his or her background. This seemingly meaning- less and tedious patter is an important and crucial social mechanism in which friendships – which are also contact points – are made. On these occasions, people map each other out, place one another on a social map of relations, which is hopefully completed with ‘en- tries’ on relations, contacts and background. Later on they can, if need be, draw on this knowledge to find contacts or information in different spheres of society. This wide net of acquaintances, friends, relatives and colleagues – in Bengali known as the atmiya-swajan,

6 Chatting, gossiping, is a favorite Bengali pastime. Chakrabarty (1999).

NUPI NOVEMBER 00 Corruption as Everyday Practice 277 one’s kin and likes7 – is actively and continuously expanded, maintained and exploited. A brother’s class-mate, the in-laws of a newly married cousin, the friend of a friend, or the brother of a friend’s friend – these are contacts that may be potentially very useful at a later point, yielding important information, furnishing further contacts, opening points for penetration of intimidating bureaucracies. Generally in Bengali society, the special value attached to personal contact and personal relationships is particularly marked. On a practical level this allows for a wide resource basis of personal contacts for problem solving and often has clear economic benefits. The practice of a normative system that emphasises mutuality, obligations and the morality of kinship as basic to social being brought Uttam closer to information and access to state services. It still took him more than a year to land a teacher’s job; still this was a strategy that came easy and naturally. Other strategies, that he was to pursue later on, included contacts in the dominant political party (a measure of political involvement) and building of direct relationships with personell at schools that had vacancies. On one occasion he brought the visiting anthropologist to the school as an excuse to gain access. That other strategies were more useful as well as legitimate underlines and perhaps emphasises how useful networking can be in such situations and how easily it can lead to situations where bribe-giving is contemplated as the next act. The dynamic of the situation was in itself not different from similar situations where information was sought. His huge network was not maintained with state penetration foremost in mind, but it allowed him, incidentally and when the occasion arose, to do that too. Particular contacts and general networks may be maintained and cultivated with a strategic consideration also. Kondos (1987) shows this in a case from Nepal, and networking can be seen as the back- bone of political engagement in India – whether locally or at a higher level. Even to a non-political individual, certain contacts more than others are maintained for their potential usefulness. However, it is crucial to understand that in cases where the invisible border between private and public, between the acceptable and the more dubious, is about to be crossed, it is often with a bit of reluctance, with some hesitation, with an eye to other possibilities. When actually crossed, the act is very often shrouded in the atoning idiom of reciprocity.

7 For an investigation see Inden and Nicholas (1977).

NUPI NOVEMBER 00 278 Arild Engelsen Ruud

4. Reciprocity and Bribes The case of Kalo will illustrate the practice of active use of one’s network, a case where the network is pushed and employed widely with a clear strategic goal. The case will also illustrate how the traits of ability and cunning enter into the picture, and how this game of ‘personal contacts’ somehow reluctantly comes to involve an exchange of money. Kalo was a technician of sorts, a hard- working man, who for various reasons had given up his paid job to start living off the land that he owned. The expanse of land was not very small and he, his wife and their two children could live the lives of a middle-class peasant household. But they also had (urban) middle-class aspirations, in particular for their children for whom they wanted to secure a good – which means costly – education. In order to secure a more steady income, Kalo applied for a position as a Class IV employee at the local hospital. There were 30 vacant positions, and several hundred applicants. The interviews were held over several days. Kalo had reasonably good qualifica- tions, but so had many others. Educated unemployment is one of the scourges of West Bengal. As one who was not exceptionally qualified, he knew that only informal sources of influence would help. Over the month or so that went by from the formal interviews to the declaration of the results, Kalo – who was a close friend of mine – came to spend the night at our apartment in town on four occasions, and on at least seven occasions came by for lunch or afternoon tea. The reason for these frequent travels into town from his village home was searching for information and contacts that would secure a positive handling of his application. In the end I could not follow the intricacies of his often frenetic search, but at least in the beginning it went something like this: He had initially two promising contacts, the first being an old friend of his who was employed in the municipality and who allegedly knew personally one of those in charge of the selection process. Accom- panied by this friend, Kalo went to the person in charge of the selection process, who, however, turned out to be quite unwilling to be approached in the matter. He declared himself opposed to favouritism and dismissed them. The second promising contact was his cousin, whose neighbour was a highly placed hospital clerk. In the company of his cousin, Kalo visited this clerk one evening, and the clerk kindly promised to do his bit to help. He warned them, however, that this would involve money because he would need to make payments to certain well- placed people. Kalo expected him to take a reasonable cut for him-

NUPI NOVEMBER 00 Corruption as Everyday Practice 279 self as well. Some days later, having talked to the people in question, the clerk informed Kalo of the sum that was required. This was a rather steep sum (some thirty thousand rupees), but the argument was that many palms would have to be greased and anyhow many other applicants had already been paying out in order to secure themselves a position. Kalo gave up pursuing this line because of the amount involved. The price demanded was prohibitive, so other strategies were called for. When bribing is not an alternative, other strategies are needed. And Kalo used his energies to pursue these. The friend with whom Kalo had visited the person in charge of the selection process, was possibly a little humiliated by his own inadequate knowledge about the man’s views. The status attached to a good network of contacts in high places often leads people to exaggerate the quality of their relationships. His friend subsequently gave Kalo the name of another person to approach. This person, he explained, was extremely well-connected in the municipality and could arrange almost everything. This man worked out of town, but returned each evening to Burdwan by train. Kalo met him twice, by the bicycle-stand next to the railway station, and on both occasions late at night. Their conversations, once over tea and once walking the street with their bicycles, centred on price, contacts and possible return favours from Kalo. Kalo’s brother-in-law (Nikhil, to whom I shall return shortly) was a Calcutta-based journalist with a wide network and one or two exceedingly highly placed friends. It was this connection, in addition to the money, that this man seemed to be wanting to take advantage of, for future use. He wanted Kalo to introduce him to Nikhil. Kalo hesitated about involving his brother- in-law, possibly because he thought his own influence with Nikhil was limited. In between their two meetings, Kalo made his own inquiries about this man through some other contact that he had in the municipality. These contacts again suggested that the man was not entirely to be relied upon, and that he might well be unable to secure the job for Kalo in spite of everything offered. In the meantime, it had turned out that the in-laws of Kalo’s cousin were close personal friends of one of the town commissioners.8 After some ado, Kalo was accompanied by his cousin to meet the commissioner (or was he an ex-commissioner? no one seemed to

8 Commissioners are elected representatives to the municipality. Their clout in city matters lies not so much in their position as commissioners as in their position in the dominant political party. Not all commissioners are well-connected and powerful; some are very marginal.

NUPI NOVEMBER 00 280 Arild Engelsen Ruud

know or care). The commissioner listened carefully but explained that his own position was not one of great influence in the matter (which may have been to say that he was not particularly interested). He promised to write a letter to the right person recommend- ing Kalo. Evidently he did not wish to jeopardise his relationship with his friends, whose son-in-law had brought this person. Yet to all present this meant that he would not apply his clout in this matter. Kalo’s cousin suggested that it would be quicker and less burden- some for the commissioner to just pick up the phone and call the right person. But the commissioner insisted that in these matters it was important to follow the correct procedures, otherwise things could become misconstrued. Kalo’s adventures over the last two of the four weeks eluded me since we were out of town. The last strategy that I know of was that he implored his brother-in-law, Nikhil, to come to his assistance. Nikhil, who lived in Calcutta, was the well-connected journalist with contacts high up in the party hierarchy. Kalo phoned Nikhil, explained the situation to him, and told him to go see his exceedingly highly placed contact and personal friend. Such an influential person would only have to pick up the phone and tell the man in charge of the selection committee that this Mr so-and-so is a very reliable man and you may want to consider him closely – and the man would be employed. Highly placed individuals do not give ord- ers in such matters, they just hint. The muted code is that if the official in question does as hinted, the highly placed individual will remember his name at some later point. However, at this point a different concern entered into our story. In a highly corrupt society there is no reason why such a chain of events would not be set in motion. But corruption of this kind – patronism, nepotism, or whatever it is called – is illegitimate to most actors, at least in West Bengal. Different ideological considerations overrule the use of contacts. Nikhil was a journalist in a communist paper, and the ruling Communist Party, to which he belonged, prid- ed itself in being incorrupt. The idea of providing a clean and clearly pro-poor government was a strong ideological guideline for many members of the Communist Party. Kalo knew this, and he knew Nikhil’s stand (which is probably why he had hesitated about involving Nikhil), but was growing increasingly desperate. Nikhil, however, was reluctant, promised once or twice that he would, but eventually did not. Twice Kalo went all the way to Calcutta (a full day’s project) to meet him, but Nikhil fled his home rather than face him. For Nikhil – when I spoke with him later – it was too embar-

NUPI NOVEMBER 00 Corruption as Everyday Practice 281 rassing to ask this very senior politician to do such a petty thing as securing a relative a Class IV position.9 It was simply too humiliating to have to ask. Nikhil’s dilemma shows us that different considerations apply to such cases. His opposition to nepotism was mellowed by family obligations. However, his moral dilemma in the end led him to refuse to interfere on his brother-in-law’s behalf. He declined to do it, declined to help his sister’s family, and jeopardised his relationship with her, his favourite sister. They were not on speaking terms for a long time. The logic of reciprocity and obligations creates pressures on the individual official – including the honest ones, who have to make sacrifices in order to stand their ground. Kalo’s other strategies failed and he did not get the job. But his case is still interesting as an example of how the habitual practice of networking operates and may help penetrate the state. Network- ing is explicitly employed as a strategy to cope with the bureaucracy. Kalo’s working the net was an unsuccessful attempt but an attempt none the less. (With 30 positions and several hundred applicants, most attempts would have to be unsuccessful.) Bribing was a possible strategy, but only one among several, and not a very reliable strategy. And even where the issue of a bribe did appear, it was invariably in the light disguise of kinship relations, wrapped in the morality of the widespread net of friends and relatives and relatives’ friends – the atmiya-swajan. Kalo worked this system, with its innumerable points of contact, its morality of mutual obligations, and with its various paths and severalty of strategies. Perhaps he just lingered too long, vacillating between options. Perhaps he was just not a good player, or did not have much to offer in return. In any event his story shows the painstaking search for possible points for penetration of someone who had become engaged with the bureaucracy, a search which involved multipronged strategies, where an appreciation of the importance of building personal relationships is a must – even where an exchange of money takes place – and where finesse and understanding of hidden meanings are required tools of trade. Kalo was engaged in difficult and subtle negotiations; aiming to be successful he employed his skills in persuasion and he exploited his network. Fundamentally, his case shows the unclear or plastic distinctions between ‘public’ and ‘private’. The following pages will seek to use this ethnography in order to extract a more refined understanding of the practice of corruption.

9 For those not familiar with the Indian administrative system, this is pretty far down in the hierarchy, albeit not at the bottom.

NUPI NOVEMBER 00 282 Arild Engelsen Ruud

5. The Puzzle of Inconsistent Condemnation It can plausibly be argued that the state as it appears locally in the eyes of its citizens is coloured by the circumstance of corruption as one of the means of mediation between the state and individuals. Moreover, studies show that corruption is condemned by a majority of the population in most third world countries (Rose-Ackerman, 1999: 91). My own experience from India supports this impression. News- papers are commonly very critical of the political establishment and regularly publish articles on alleged corruption. In private people are even more critical. Corruption is universally condemned as bad and I have personally not come across anyone who would be willing to defend corruption – that is, even from the native’s culturally sensitive point of view. At the same time though, a great many people will participate in acts that can be labelled ‘corrupt’, and that too with ease and dexterity – viz. the above examples with the exception of Nikhil. This then is one of the puzzles of trying to understand corruption. The very vocal opposition against corrupt activities may seem to contradict my arguments early in this article, but not necessarily.10 Ideas do move on different levels of consciousness and application. A voiced general opposition does not preclude exemptions for particular situations. Nikhil the journalist’s hesitation is a case in point. A moral universe is rarely unitary or even allows for only one answer to difficult situations. Again we should keep in mind the basic plasticity of discourses and arguments, and also that the nuances of interpretation applied by different parties to an act are not static. Bribe-taking, rule-bending and favouritism were universally condemned as morally base and unacceptable by all my friends and informants. The understanding of the causes of widespread corruption in their society, however, was somewhat defeatist. Attitudes would be expressed in phrases such as ‘Indians are inherently corrupt’; ‘The bureaucracy was not corrupt under the British but now it is manned by Indians’, or ‘We Indians, we always elect the wrong [meaning corrupt] leaders’.11 Politicians and bureaucrats were in-

10 From Africa Olivier de Sardan points out similar contrasts. He explains the gap by pointing out an absence of ‘ethic of public service’, the lack of a ‘public domain’ tradition, and the strong sentiment of shame that prevents people from blowing the whistle. The social price of conflict is simply too high. Olivier de Sardan (1999: 30–31). 1 1 The first is not a direct quote but a paraphrasing of several statements. The two latter are direct quotes. The first was made by a bureaucrat and the second by a businessman. All, however, mimic widespread sentiments. although, with reference to the second of the three statements, some would also hold that corruption was brought to India by the British.

NUPI NOVEMBER 00 Corruption as Everyday Practice 283 variably talked about in negative terms (Ruud, forthcoming). The state machinery as such was considered corrupt and corrupting, and popular attitudes to the state were mostly negative because of that. Terms employed to describe corruption suggest how shady and unsavoury such activities were viewed: ghush (in Bengali, meaning ‘bribe’), from what is low, suppressed, secret (Dev, 1989); and bhrastachar (in Hindi, meaning ‘corruption’), from bhrasta – ‘fall- en, depraved, spoilt’ (Das Gupta, 1977). These are common terms applied to situations where there is illicit transgression of official roles. In the case of Bengali villagers, outright bribery is held distinct from other forms of corruption, such as forms of nepotism. The term used for bribes (ghush) does not cover forms such as nepotism. Other expressions are used in connection with kickbacks and embezzlement (e.g. taka khaoya, or ‘eating money’). As such, bribes appear to have a special significance, a specially negative meaning. For one thing, there is the power gap in many situations that enables the bureaucrat to demand a sum fixed by himself from the pleading poor peasant. The poor peasant has often a very limited option but to pay, and to pay the rate more or less dictated to him. This situation is very different from what we associate with nepotism, namely the morally more or less enforced bending of (administrative) rules for the sake of family. Bribery may then be held to be morally different from nepotism. The intent behind the move is important to its legitimacy. However, I would argue that outright bribery is not a distinct type of activity for our purposes. Consider this: A villager has to pay a bribe to a government service in order to get a service done. He will call it ‘a bribe’ (ghush), and he will probably be quite frustrated about it. But his neighbour may well manage to get the same service done without having to pay for it. Instead he would have used his contacts, sources of influence, channels of contact-generation, or his powers of persuasion, manners, references to caste or community. This last act will be termed and interpreted differently, yet the crossing of the public–private dividing line will have been the same. Perhaps the blurredness or softness of the public–private division does not stem just from poorly understood categorisations but from the fluidity of interpretations of an act of crossing. The ration- ale behind the crossing and the manner in which the dividing line is crossed constitute crucial elements in how such an act is legitimised and accepted (or not). One clue to the lack of general opposition to corruption is possibly found here, in that there are ways and means

NUPI NOVEMBER 00 284 Arild Engelsen Ruud

by which the ordinary villager circumvents the obstacles posed by an impersonal bureaucracy. Even when not demanding bribes but instead following rules and regulations, a state official may still create problems for individual villagers, problems that he will seek to overcome one way or the other. It is not the poor as such who pay the bribes, it is those without adequate contacts (albeit the poor are mostly disadvantaged in terms of what they have to offer in a relationship – poor and low-quality contacts constitute a marginal ‘capital basis’ with which to extend networks). What they do is they ‘manage’, or myanej kara, an English term used in Bengali to denote the sort of wheeling-dealing activity that sorts out problems through use of a wide range of contacts and different means of persuasion. ‘Fixing’ seems an adequate translation. The clever ‘fixer’ is applauded among his peers. He plays a potentially crucial role in their lives, as a middleman who furnishes contacts and expedites deals. To outwit the bureaucracy suggests cunningness and ability, qualities that are sought after. Such a person cultivates contacts with friends, to whom he may again be useful, whom he may help, and who in turn may be useful to him. Other players may be less clever, and there are degrees in the level of activity that people exercise. However, as the case of Kalo would indicate, even the not-so-successful players know the basic rules and practices of the game.

6. Between Bribes and Obligations Kalo’s endeavours demonstrate that the distinguishing line between a bribe and a friend’s friend’s influence can be wafer-thin and often overlaps. In his case a bribe was only the last resort, a blunt and unsafe tool contemplated where other means proved inadequate. A bribe is but one way of persuading recalcitrant bureaucrats to do their job or do it differently. Only in the most difficult of circumstances does one pay a bribe; in most cases any ordinary and reasonably well-informed villager will seek out other possibilities first. People do not primarily seek to achieve objectives vis-à-vis the state through the use of money or other types of ‘grease’. They work the system – middlemen, contacts, friends, relatives – seeking energetically to avoid paying a bribe but in the process trying to achieve the same objective through other means of persuasion or influence. ‘Non-bribe influence’ will, when successful, still be a form of corruption, but it will appear to many as a less stark and more acceptable form. In many instances, it is likely that influence is a preferred method of persuasion, not only as compared to brib-

NUPI NOVEMBER 00 Corruption as Everyday Practice 285 ery but also instead of a long and complicated legal process – perhaps involving costly fees. Moreover, in actual practice the distinction between a bribe and using one’s contact is not clear. It is not easy to draw the distinction between paying up, in cash, and repaying later, in the currency of return favours. Between the two extremes of paying a bribe and doing favours for your next of kin lies a spectrum of different types of interactions, with shades of meaning and intention. Transactions will be open to interpretation, and interpretation is important because it is about what legitimises the act. This of course also opens to contestation and differing opinions. The meaning of paying money for a service is not necessarily imbued with the stark notion of a pure monetary transaction. A certain mutual understanding has to be reached, and the money is paid on the understanding that other hands will have to be greased, sometimes perhaps as a ‘consideration’, or in view of expenses that may be incurred or inconveniences or risks experienced while doing the favour. But the reciprocation for inconveniences or risks may well be done in other currencies, in return favours, delayed and of a completely different order.

7. Strategic Relationships We may consider the distinction as a continuum of shades, from gestures of thanks to unashamed bribes. At one end of the scale, next to gestures of thanks, one would find the symbolic gift of thanks, a cigarette, some food item, or a courtesy visit. Next there would be the gifts of reciprocity, that cement the relationship rather than end it. Such gifts may range from token to substantial, and the exact nature of the gift and return gift may be negotiated. It depends much on the nature of the relationship, the stakes involved, and the relative status of the parties involved. Moving along the continuum one might find the gifts that honour, that add to the status of the receiver as a man of substance fit to be given fine gifts, thus defining his superior status. Such gifts also range from token to substantial. Here a smaller or larger element of strategy may be involved, from vying for favours to buying favours. At the far end of the continuum, just before bribes, we find return favours explicitly for services rendered. In one of the few culturological studies of corruption in South Asia, Alex Kondos narrates a story from Nepal about how a relationship to someone of importance can be created, intentionally and strategically, yet it has to be a personal relationship which cannot be pressed for un-

NUPI NOVEMBER 00 286 Arild Engelsen Ruud

called-for benefits (Kondos, 1987). This takes place within an accepted cultural frame. Through certain symbolic acts and much time spent, an individual may enter into the circle of clients/friends of a major figure in a culturally meaningful way. The relationship is one of reciprocity, yet has a strategic element and allows the individual to penetrate the maze of government rules and role sharing. But the interested nature of the relationship is by necessity on a morally sustainable basis.12 This seems to be a rather unusual case, although active establishment and maintenance of relationships are known and commonly practised. For one thing it is necessary to acknowledge that some people are better players than others – they have more charisma, are more intelligent and charming and have less of a problem in establishing relationships. Whether or not the relationship is pursued with a strategic motive is not easily decided. More- over, the personal bond is the one that legitimises the relationship and any transactions within. Fictitious kinship is constituted not only by a liberal use of kinship terms but by an active personal interest in each other’s personal lives. The personal aspect of the relationship will always be seen to be strong, even where, to both parties, the relationship is clearly strategic. The motive behind an exchange of gifts (or services) becomes crucial then. The more outrageous forms of corruption, such as outright bribing among strangers, are illegitimate in an Indian or Bengali context, whereas kind reciproc- ities among close friends or relatives are not. However, the defining line between friends and strangers is not clear-cut. We need to look closer at how the practice of corruption is constructed and rationalised, how the exchange of money somehow incidentally enters into the process. The moral imperative to provide immediate and unselfconscious support for members of one’s extended family is a well-known feature of South Asian society, valid naturally for ordinary people as well as members of the political system and bureaucrats. Interest- ingly, the concept of the ‘extended family’ may also encompass individuals who are not family, in a strict or even rather loose sense. A term such as the Bengali atmiya-swajan spans widely to encom-

12 Bourdieu (1977:181) refers to the dominant’s position but the same could be said from the other vantage point also: ‘Wastage of money, energy, time and ingenuity is the very essence of the social alchemy through which an interested relationship is transmuted into a disinterested, gratuitous relationship [...]’. Olivier de Sardan, on the other hand, talks of ‘predatory authority’ (1999). The ambivalent position of someone at the client end of a close patron–client relationship is touched upon in Ruud (forthcoming).

NUPI NOVEMBER 00 Corruption as Everyday Practice 287 pass friends, neighbours with whom one is on good terms, long-term subordinates (labourers, servants), and even occasionally people rather distant in terms of status but with whom a close relationship is actively maintained. The morality of the extended family is thought to include and encompass also people with whom one has a strategic relationship. Such a relationship still builds on certain cultural presumptions that mitigate and hide its rawness. We could see that construct at work in the situation of Kalo visiting the commissioner – who was Kalo’s cousin’s in-laws’ friend. This rather thin connection was enough to get Kalo an encounter with the commissioner and an opportunity to present his case before him. It was not enough to get the commissioner sufficiently interested, but then the connection was stretched rather thin. Had their relationship been closer his attitude might have been more forthcoming. Possibly, when a relationship is rather stretched in terms of intimacy, the need for favours to be compensated for soon is increased. The moral obligation of reciprocity is not strong when the other person is the cousin of one’s friends’ son-in-law.

8. The Performative Aspect A tendency to regard an act of corruption as an ‘exchange’ in an economic sense (even if involving types of capital such as social or symbolic capital) conceals an important aspect. In seeking to understand the negotiation, we give due consideration to the moral environment in which the act is negotiated, to the cultural constitution of the ‘items’ involved, and to the limits set by access to information. Yet, the tendency is often to regard the act as something exchanged for something else, where a ‘price’ is fixed relative to the individuals involved and the service rendered. Such a view, however, ignores the ‘performative aspect’ of the negotiation, the deft (or not) handling of the process itself. There may, for instance, be a certain ‘tempo’ to the process, hidden codes and a muted language that constitute the negotiation process. These are elements subject to interpretation and manipulation. The ‘real’ message may well be the opposite of what is being voiced, but to know and be able to use these is a matter of competence and intelligence. Each and every ‘player’ may not be equally competent in this, it is an unevenly distributed form of ‘capital’. Knowing how to pay, or how not to pay, a bribe, or to get a bribe, or otherwise to circumvent rules and regulations, are feats that have their own heroism attached, a game with its own rules and heroes.

NUPI NOVEMBER 00 288 Arild Engelsen Ruud

Although it has been held that corruption at today’s level is something relatively new to India, caused by the British, or by overdone centralisation of decision-making after Independence (‘the Licence raj’), or by the logic of democracy in an undemocratic society, or by liberalisation of the economy,13 it can also be held that a notion of a public–private distinction is quite old. Equally it can be held that the problem of transgression of the public–private distinction is old. The problems of corruption are described in the Arthashastra (ap- prox. 200 BC)14 and in historical studies of eighteenth or nineteenth century British rule.15 Whatever reasons for high or increased levels of corruption, we may suggest as not entirely implausible that crafty interaction with the state machinery has a long lineage in the history of Indian society. The fact that the state’s capacity to enforce its own view of things was also relatively limited has probably contributed to the development of informal procedures of interaction between lower level officials and members of the population. The state has been there for a long time, and people have learnt to live with its potentially oppressive presence yet seek its limited but valuable resources. Put differently, through the state’s constant presence in a society’s daily existence, a mode of accommodation has been worked out and has – over time – given rise to certain informal yet somehow standardised procedures. We can understand corruption as part of a broader pattern of interaction between bureaucracy or state and the population in general, constructed as and evolved into a partly separate sphere of activity, a sphere with its own rules, codes of conduct, rewards and measures of success or failure. Let me try to describe it. The process of negotiation for getting a job done by a bureaucrat – including the negotiation for the size of the possible bribe – is to many a type of game, and it requires knowledge, wit and intelligence. First of all, these matters are delicate, and have to be approached with a certain finesse. The main problem is mostly that one may not even know where to put the bribe, whom to approach, or how to breach the issue. It is extremely foolish to walk up to an

13 The different arguments have been carried by respectively Paul (1997), Kaviraj (1994), Saberwal (1996), and Harriss-White and White (1996). 14 See in particular paragraph 2.9.34 (Rangarajan, 1992:283). 15 See, for instance, Frykenberg (1965). The establishment of an Indian Civil Service was an effort to shield the administration from other than official interaction with the population. Britain of course had its own forms of corruption in those days. The wheeling-dealing that went on with glee under the noses of the incorruptible ICS officers is shown in Chatterji (1981).

NUPI NOVEMBER 00 Corruption as Everyday Practice 289 official and propose a bribe. He would in all likelihood take great offence, and in the presence of colleagues (many officials approach- able by the public sit in the presence of colleagues) to accept a bribe would in most cases be unthinkable. To be able to master this game is certainly a matter of knowledge, experience and contacts. Even where a contact has been established, the issue of a bribe needs to be breached, the amount negotiated. To be capable of ably playing that game also enhances one’s standing among one’s peers, just as being manipulated by a more able player is a source of great em- barrassment. In Akhil Gupta’s 1995 narrative from a North Indian land record office, two young men were easy prey to the bemused manipulations of the record keeper. It was the two young men’s complete lack of understanding of the game which left them entirely at sea – not their structural position. Gupta – whose phrase I have borrowed – talks about a ‘performative aspect’ of corruption, and writes that ‘The “practice” of bribe giving [is not] simply an economic transaction, but a cultural practice that required a great degree of performative competence. [When complaining about corruption, villagers were also] expressing frustration because they lacked the cultural capital required to negotiate deftly for those services’ (Gupta, 1995: 381). We may note in particular the play- fulness with which the land record keeper treats the two inexperi- enced visitors. Moreover, to be able to know where to grease, to push the right button, to know whom to talk to, these are eloquent measures of a man’s political competence, his ability of knowing his way about. It is not knowledge that comes easy. It is an acquired skill based on intelligence, experience and learning, and a special kind of talent, or at least some finesse and knowledge of the rules of the game. The essence of the bribe negotiation is not only to arrive at a mutually agreed sum. It is a process in which the two parties try out each other’s negotiation skills, cunningness and relative power and status. And it is a matter of getting the stakes lowered or increased. The hard negotiation over the size of the bribe will most often be rather understated, muted in coded messages and recognisable patterns of argumentation, which skill and knowledge alone can decode. The subtleties of the process are constituted in known or at least knowable procedures. This comes out in Wade’s description of villagers in South India approaching irrigation department officers to negotiate for the timely release of irrigation water.

NUPI NOVEMBER 00 290 Arild Engelsen Ruud

The interaction between farmers and officers commonly goes something like this: a few farmers representing one village [...] will approach the AE [Assistant Engineer] in charge of the Sub- division in which their land is located, and put to him a request. The AE tells them, as they expect, ‘No, I cannot agree. You please see Supervisor and see what he can do’. The AE, now informed, tells the Supervisor his price for giving them what they want. The Supervisor in turn tells the farmers how much it will cost them, perhaps with an initial show of reluctance. If the farmers do not agree some bargaining may follow, with the Supervisor sending them away with a coded refusal meaning that they should see him again after he has consulted with the AE. [...] (Note that the AE takes care never to be seen asking money of farmers) (Wade, 1982: 296, italics added).

Wade here gives us the description of a negotiation process which is muted and coded, unstated yet somehow very real; it takes experience to understand and take part in it. It could be envisioned as a system of flexible rules, performed ad hoc, guiding citizens in their interaction with representatives of governmental services. Bribery may be a specific type of activity, a monetary exchange for particular, distinct types of known services. But in most cases bribery is the last resort for those with no other means of persuasion. As an outright monetary exchange – a buying of a service – it constitutes a defeat on the buyer of the service, the proof of his inadequate contacts in the system, his inability to play the game that matters, where the greatest asset is to have contacts or contacts who have contacts. Most people go to great lengths in order to avoid having to pay a bribe.

9. Peer Applause The ‘performative aspect’ has a twin in the peer applause of the able player. To be a good performer, able to make good deals, is a matter of no small pride or praise. In the context of West Bengal, civil or public servants who amass conspicuous wealth will be condemned by all but their nearest kin. And yet the able villager who is cunning enough to get things done without paying a bribe or otherwise outwitting someone in the state bureaucracy will be applauded and praised by his circle of friends and most other people who might happen to hear of it. To take a bribe is considered bad, and to be forced to pay one is a pain. To be able to get something done

NUPI NOVEMBER 00 Corruption as Everyday Practice 291 through a reliable contact in the bureaucracy, however, is praise- worthy and a good sign of being clever. This adds to the ‘puzzle of inconsistent condemnation’ we encountered above and shows that at least to some extent alternative motives may be at play. Corruption, it can be argued, can be understood as a set of actions with some commonality, a certain behavioural characteristic, including perhaps the secrecy and often glee with which stories of these shady activities are told and retold between friends. What is corruption and what is not, is a matter of shade, where morality in an almost imperceptible way becomes daring immorality. It is a matter of contestation – not only between the haves and the have- nots, but between the high ground of ideological and moral purity and the daily quest for survival and pursuit of narrow, private interest – and it follows certain patterns, patterns that are improvised upon, but patterns none the less, certain standard procedures that enable people to penetrate what was supposed to be an impersonal state. In a way it resembles what Michel de Certeau in his study on the practice of everyday life calls ‘poaching’. In describing how people use an imposed system, he writes:

People have to make do with what they have. In these combat- ants’ stratagems, there is a certain art of placing one’s blows, a pleasure in getting around the rules of a constraining space. We see the tactical and joyful dexterity of the mastery of a tech- nique. [...T]here is a skill that has its connoisseurs and its aesthe- tics exercised in any labyrinth of powers, a skill ceaselessly recre- ating opacities and ambiguities – spaces of darkness and trick- ery – in the universe of technocratic transparency, a skill that disappears into them and reappears again, taking no responsibility for the administration as a totality (de Certeau, 1984: 18).

10. End Notes This article has suggested that the division between public and private – the basis for common definitions of corruption – does not carry the same moral weight in all societies. This does not mean that the division is unfamiliar or entirely alien. It is well known at least in a society such as the one I have drawn my case-studies from, but the application of the distinction in individual situations may be restricted by other more weightier moral considerations. When the dividing line is crossed – which would constitute an act of corruption – certain traits may be observed. In many cases indi-

NUPI NOVEMBER 00 292 Arild Engelsen Ruud

viduals are drawn into situations in which the division is crossed in an incidental manner, by employing their network as they would to solve most other predicaments. When this does take place the distinction between different forms of corruption is not clear in practice. The different forms overlap, they have their own vocabularies, and the moral legitimation is open to contestation and manipulation. Not the least, the ‘bribe’ that is offered in return for a service may be an ill-defined future obligation. This is important, particularly since people often spend much energy seeking and trying out alter- natives to a bribe, although the end-result may constitute the same act of corruption. The competent player who achieves without paying is applauded by his peers and I have argued that corruption has crucial ‘performative aspects’ with its own procedures and rewards – an ‘art’. De Certeau may be criticised for overstating the delight felt in the encounter with the machinery and drudgery of the dominant force. The delight stems only from overcoming what is an impos- ing and powerful system – like poaching is a delight only if not hunted down by the guards and the hounds from the manor. It is also important to note that he ignores how this ‘art’ and its appreciation may change with wider ideological changes in society. The performative aspect mentioned above is interesting because it allows for a view of the practice of negotiation – and of corruption – as non- static. Different elements may be brought into the negotiations, such as elements of law or morality, about, for instance, what top politicians are ‘known’ to be doing or have been exposed as doing. At a certain level, discourses and ideologies have an element of plasticity that allows for new ideas and evaluations to be brought in. The sense of a border between public and private is subject to the same currents of ideas and may be manipulated or influenced by events or renewed emphasis on moral interpretations. These procedures follow gently from other established procedures, the patterns of obligations, mutuality and expectations that are integral to every other aspect of daily life. These procedures are replicated in a natural fashion both in the search for a suitable marriage partner for one’s daughter and in the search for information, contacts and assistance. Only with reluctance do these procedures shade off into the murky waters where money is exchanged and private interest quests become blaringly evident. But those instances are sought hidden, couched in words of double meaning – if you know how to read them – in hints that allow for salvaging one’s honour. Corrup- tion, in its many forms, is tricky to understand even for the practi-

NUPI NOVEMBER 00 Corruption as Everyday Practice 293 tioner, because it is rarely purely wrong; it is a special case of a wider practice – mutuality and dependence – which allows it to remain ambiguous.

References Bourdieu, P., 1977, Outline of a Theory of Practice, Cambridge: Cambridge Uni- versity Press. de Certeau, M., 1984, The Practice of Everyday Life, Berkeley: University of California Press. Chakrabarty, D., 1999, ‘Adda, Calcutta: Dwelling in Modernity’, Public Cul- ture, Vol. 11, No. 1, Winter, special issue on ‘alternative modernities’, pp. 109–145. Chatterji, B., 1981, ‘The Darogah and the countryside: the imposition of police control in Bengal and its impact (1793–1837)’, The Indian Economic and Social History Review, Vol. XVIII, No. 1. Das Gupta, B.B., 1977, Hindi-Bengali-English Dictionary, Calcutta: Das Gupta Prakashan. Dev, A.T., 1989, Students’ Favourite Dictionary (Bengali to English), Calcutta: Dev Sahitya Kutir. Frykenberg, R.E., 1965, Guntur District 1788–1848: A History of Local Influ- ences and Central Authority in South India, Oxford: Clarendon Press. Gupta, A., 1995, ‘Blurred boundaries: The discourse of corruption, the culture of politics, and the imagined state’, American Ethnologist, Vol. 22, No. 2, pp. 375–402. Harriss-White, B. and G. White, 1996, ‘Liberalization and the new corruption’, IDS Bulletin, Vol. 27, No. 2. Inden, R.B. and R.W. Nicholas, 1977, Kinship in Bengali Culture, Chicago: Uni- versity of Chicago Press. Kaviraj, S., 1994, ‘Crisis of the nation-state in India’, Political Studies, Vol. XLII, pp. 115–129. Kondos, A., 1987, ‘The question of “corruption” in Nepal’, Mankind (New South Wales), Vol. 17, No. 1, pp. 15–29. Olivier de Sardan, J. P., 1999, ‘A moral economy of corruption in Africa?’, The Journal of Modern African Studies, Vol. 37, No. 1, pp. 25–52. Paul, S., 1997, ‘Corruption: who will bell the cat?’, Economic and Political Week- ly, June 7. Price, P., 1999, ‘Cosmologies and Corruption in (South) India’, Forum for De- velopment Studies, No. 2, pp. 315–327. Rangarajan, L. N., 1992, Kautilya: The Arthashastra, New Delhi: Penguin clas- sics. Rose-Ackerman, S., 1999, Corruption and Government: Causes, Consequences, and Reform, Cambridge: Cambridge Unversity Press. Ruud, A.E., forthcoming, ‘Talking dirty about politics: A view from a Bengali village’, The Everyday State and Society in Modern India, C. J. Fuller and Véronique Bénéï, eds., New Delhi: Social Science Press. Saberwal, S., 1996, Roots of Crisis: Interpreting Contemporary Indian Society, Sage: New Delhi.

NUPI NOVEMBER 00 294 Arild Engelsen Ruud

Wade, R., 1982, ‘The system of administrative and political corruption: Canal irrigation in South India’, The Journal of Development Studies, Vol.18, No. 3, pp. 287–327.

NUPI NOVEMBER 00 Fish as VehicleForum for for Economic Development Development Studies |in No.2-2000 Namibia 295

Fish as Vehicle for Economic Development in Namibia

Ussif Rashid Sumaila

1. Introduction Until 1993 freezer trawlers were the dominant vessels employed in the harvesting of the hake stock in the Namibian Exclusive Eco- nomic Zone (EEZ). At the time of independence in 1990, the government of Namibia saw its rich fishery resources as one of the vehicles available to it for badly needed economic development. An important fishery policy decision taken by the government called for the restructuring of the trawler fleet in favour of wet fish trawlers (see section 2 below). The fishing sector is an important part of the economy of Namibia. It has been estimated (by the Ministry of Fisheries and Marine Resources, MFMR) that hake contributed about 7.4 per cent of Namibia’s estimated exports in 1994. It contributed about N$ 951 million to Namibia’s GDP in 1997 (MFMR, 1997).1 It should be noted that these figures include only the direct contribution to GDP, additional contributions from secondary indu- stries and the multiplier effects of spending hake-related incomes are not included. In 1992 wet fish trawlers landed only about 5,000 tonnes out of total reported hake landings of 87,498 tonnes. The declared policy was that 20, 40 and 60 per cent of the total quota for hake should be allocated to the wet fish trawlers in 1993, 1994 and 1995, respectively. The ultimate aim was to maintain an allocation of 60:40 in favour of the wet fish trawlers. Performance against stated objectives had been quite good up to 1994: in 1993, 19.9 per cent of the total quota was allocated to wet fish trawlers. The correspond-

1 One US$ was equivalent to about N$6 in October 1999.

NUPI OCTOBER 00 296 Ussif Rashid Sumaila

ing allocation for 1994 was 48.9 per cent, well over the target of 40 per cent. However, the target of 60 per cent could not be achieved from 1995 to 1998 mainly because there was no increase in the total allowable catch (TAC) in these years. The new government policy sought to restructure the fleet in favour of the wet fish trawlers, mainly to encourage onshore processing, and thereby reap benefits such as increased employment for seasonal workers from the north of the country. This move would make companies process their catch on land in Namibia. In this way, the country hopes to ensure participation in the fishery by Namibians, with the attendant positive effect on their economic welfare. An interesting question to ask here is to what extent this new development is economically rational; if it is not, are there any reasons other than economic that may justify this move? For example, is it the case that gains in employment due to restructuring can compensate for the resulting economic loss, if any? Seeking answers to such questions is the main purpose of this article. The article thus sets out to study the exploitation of the hake stock off the Namibian coast, with a view to finding out the proportion of the total annual quota that should be landed by the wet fish and the freezer trawlers, respectively. The principle underlying this work is economic efficiency, in other words it is assumed that the primary objective is to harvest and process the stock in the most economically efficient manner. This assumption appears to be plausible in the case of Namibian fisheries, which unlike the fisheries of most developing countries are mainly industrial.2 Thus, most of the complications that usually arise due to the community-based nature of certain fisheries are simply not present here. There is, therefore, an excellent opportunity for pursuing and, indeed, achieving economically optimal management of the resources of the Namibian EEZ. Moreover, the results to be computed herewith are meant to serve as benchmarks for determining the trade-offs between different government policies: we should, for instance, be able to discover what is being sac- rificed in economic terms due to a government policy that is geared towards increasing employment in the industry, as against one based purely on economic efficiency criteria. In more concrete terms, this article seeks to: (i) test the govern-

2 Although small coastal communities caught fish in coastal lagoons during precolonial times, the only indigenous fishing tradition amongst the peoples of the interior was freshwater fishing in the streams and rivers of the north.

NUPI OCTOBER 00 Fish as Vehicle for Economic Development in Namibia 297 ment quota allocation policy target of 60:40 for the wet fish and freezer trawlers, respectively, to see if it is optimal in an economic sense,3 (ii) determine what discounted economic benefit would accrue to society at large under the optimal allocation regime, (iii) find out what is the optimal number of both wet fish and freezer trawlers needed to achieve these objectives, and (iv) look at the employment-generating capacities of the wet fish and freezer trawlers. The trade-offs between the economic gains and the employment-generating capacities of the two classes of vessels will also be discussed. There are at least two levels at which the employment-generating capacities of the two types of vessels can be analysed. First comes the direct employment generation from the harvesting and processing. Second, the indirect employment generation in the overall economy. Only an informal discussion of the latter is given in this article. On the other hand, a full analysis of the former is carried out. In the next section, we briefly discuss the hake fishery. Section 3 presents the descriptive bioeconomic model; the more formal model is relegated to the appendix. The numerical computations are carried out in section 4. This section consists of two parts; the first sub section discusses the data used in the analysis, while the other sub section presents the computations and the results. We then proceed in section 5 to discuss the results of the study.

2. The Namibian Hake Fishery

General The hake stock is one of the three most important fish species of the highly productive Namibian EEZ. The others are horse mack- erel and pilchard. The main reason for the high productivity of the Namibian EEZ is the Benguela upwelling system prevalent in the coastal zone of Namibia and other Southern African countries. Among the species of hakes inhabiting the Namibian EEZ, that is, Merluccius capensis (also known as cape hake), Merluccius paradoxus (deep-water hake) and Merluccius pollis, only the former two are of major importance to the fishery. These two spe-

3 By an optimal quota allocation we mean the allocation that would maximise the social planner’s, that is, the Namibian government’s overall economic benefit from the resource.

NUPI OCTOBER 00 298 Ussif Rashid Sumaila

cies are so identical in appearance that they are often treated as one and the same, both in fisheries statistics and in studies such as this one. However, their biology, ecology and abundance differ considerably (Wysokinski, 1986). Both species are relatively long-lived, reaching ages up to and over ten years. Hakes are usually found close to the seabed during daytime but rise to intermediate water during nighttime, probably following their prey. Hake catches reached a maximum of over 800,000 tonnes in 1972, averaging some 600,000 tonnes annually during the period from the late 1960s to mid-1970s. As could be expected, this period of high catches was followed by lean years, with average catches of less than 200,000 tonnes from the mid-1970s to 1980. The catches, however, rose again and remained relatively stable in the range of 300–400,000 tonnes for most of the 1980s. Hamukuaya (1994) stated that during those years of high catches there was a large proportion of young fish (2–3 years old) probably accounting for the low catches in later years. At independence in 1990 the total allowable catch (TAC) for hake was set at a low of about 50,000 tonnes, mainly to help build up the stock. This has, however, been adjusted upwards each year since then, reaching a maximum of 150,000 tonnes recently. These increases were due to the apparent improvement in the hake biomass in the years immediately following independence.

Structure of the hake fishery A large variety of fishing vessels are used to harvest hake, differing in their gross registered tonnage (GRT), engine horse power

Table 1. The structure of the Namibian hake fisheries in 1994

No. of Hake as per Reported vessels cent of repor- hake landed Quota allo- Method and fleet licensed ted landings (tonnes) cation (tonnes )

Demersal hake trawl 66 93.00 99,152 130,913 monk/sole trawl 18 31.53 4,261 7,919 longline 13 96.91 4,265 7,168 total 97 86.46 107,678 146,000 Mid-water 25 0.16 525 4,000 Total 108,203 150,000 Source: Ministry of Fisheries and Marine Resources (MFMR), Namibia

NUPI OCTOBER 00 Fish as Vehicle for Economic Development in Namibia 299

(HP), processing equipment and freezing capacity. Hake are pre- dominantly caught by wet fish and freezer bottom trawlers. For instance, in 1994, out of a total of 108,213 tonnes of hake landed, 99,152 tonnes were harvested by wet fish and freezer trawlers. This is well over 90 per cent of the total landings of hake that year. The rest were landed using monk/sole trawlers, long liners, and mid-water trawlers. The structure of the Namibian hake fisheries for 1994 is given in Table 1. Data from 1995 and 1996 show that the dominance of the bottom trawlers in the hake fisheries contin- ues unabated (MFMR, 1996). As a result of the overwhelming dominance of the bottom trawlers in the demersal hake fishery, we focus our attention on these vessels and divide the wet fish and freezer trawlers into two separate and distinct entities designated as w and f, respectively. This is clearly a simplification, since different types and sizes of bottom trawlers owned by different entities are actively fishing the waters off the Namibian coast. Our simplification makes the analysis trac- table without compromising the essence of the study, which is to find out what share of the hake TAC should be allocated to the wet fish and freezer trawlers as separate groups. In other words we are interested in determining the inter- rather than the intra-vessel allocation of the TAC.

3. The Model A typical freezer trawler is usually larger than a typical wet fish trawler. It fishes in deeper waters, probably catching larger and more valuable fish. In addition, it can stay offshore for longer periods than the wet fish trawler. The freezer trawler is equipped fully for catching, freezing and packaging at sea. Therefore, all the processes needed, from actual harvesting to packaging in readiness for export, are undertaken offshore. There are two kinds of participants in the wet fish business: factory owners who catch and process hake in their factories, and private concessionaires, who catch hake and sell their catch to the factory owners for further processing. In this article, w, that is, the wet fish group, refers to the former group of participants. The factory owners undertake two distinct activities, namely, offshore harvesting of hake and onshore processing. We therefore develop a two-agent model for the hake fishery of the Namibian EEZ, the two agents being w and f. Potentially, any situation where two or more agents with conflicting interests

NUPI OCTOBER 00 300 Ussif Rashid Sumaila

jointly exploit a common property renewable resource is capable of being analysed using a game theoretic framework (see, for instance, Munro, 1979). However, due to the assumptions underlying our model, listed below, it turns out to be a straightforward optimisation one. In addition to the fact that these assumptions seem reasonable, there is a policy-needs argument in support of the approach. In terms of policy, it can be argued that the trade-off between landing hake by wet fish and freezer trawlers is one of the burning policy issues facing the MFMR at the moment. The following assumptions are underlying the models:

1. Annual quotas are assumed to be optimally and exogenously determined by the MFMR. Hence, this study does not seek to give advice on what the optimal quota for hake should be, but rather on what percentage of the decided quota should be harvested by the wet fish and freezer trawlers, respectively. The implication of this assumption is that it eliminates interaction at the level of the stock, thereby eliminating dynamic externality.4 2. It is assumed here that there are no interactions between the two agents at the market-place. This assumption is reasonable because the agents sell their landings at competitive markets where prices are exogenously determined. The implication of this assumption is to eliminate interaction by the agents at the market-place (that is, market externality). 3. It is further assumed that there are no significant natural interactions between the hake species and others. This implies that externality due to say, predator prey relations are ignored.5 Given the lack of adequate studies on interspecies interactions between the species living in the Namibian EEZ, this assumption is considered to be a pragmatic one, which will be relaxed as more biological information becomes available. 4. The model is deterministic in the sense that all parameters of the model are assumed to be known with certainty. Also, future quotas are assumed to be known. Clearly, these are strong assumptions. In the case of future quotas, for instance, we know that yearly allocations are based on both scientific knowledge concerning the biomass of hake and policy-related considera-

4 Levhari and Mirman (1980) and Sumaila (1995) are studies where dynamic externality is present. A possible extension of this article would be to relax this assumption. 5 See Fisher and Mirman (1992) and Sumaila (1997) for analyses that incorporate the natural interaction between different species.

NUPI OCTOBER 00 Fish as Vehicle for Economic Development in Namibia 301

tions, both of which are sure to vary from year to year. A future task would be to introduce uncertainty into the model.6

The overall implication of assumptions (1) to (3) above is that even though there are two agents exploiting a common property resource, we end up with a ‘trivial game theoretic model’. To have a truly game theoretical situation, the agents have to interact either in the market-place, at the level of the stock, or else there have to be natural interactions between hake and the other species in the habitat. Thus, what we have in this model is a straightforward constrained optimisation problem for each agent. Note that both players are jointly constrained by the total quota available to them in each year.

Modelling the price of hake The assumption of no interaction at the market-place necessarily implies that both wet fish and frozen fish are supplied at given prices, implying that the price the fishermen receive for their produce is inelastic to the quantities of fish they supply to the market. It should be noted that the main market for Namibian hake is Spain. This is a large international market supplied by many other sources, of which Namibia is only one of many suppliers. For Namibia or any of the other suppliers to be able to influence the market price one of two things will have to happen. The supplier will have to withdraw from the market a large proportion (if not all) of its output, or else there has to be a sudden large increase in its supply to the market, both of which are unlikely to happen under normal conditions.

Modelling the cost of landing hake In general, two types of costs can be identified depending on whether one is talking of the costs directly incurred by the agents in the model, that is, private costs or costs incurred by society as a whole, that is, social costs.7 Usually these two are not identical because of distortions in market prices and/or costs. As the focus of this study is on the benefits to society as a whole, we will be concerned mainly with social costs. Details of the costs associated with land-

6 In the meantime, the model is designed to be flexible enough to allow quick sensitivity analysis, making it possible to vary important parameters as new information flows in. 7 It is the government that is concerned about these costs private agents would usually be concerned with only their private costs.

NUPI OCTOBER 00 302 Ussif Rashid Sumaila

ing hake by the wet fish and freezer trawlers are presented in Appendix 1.

The production and profit functions In both theoretical and applied fishery economics, it is common to assume that the production (harvest) function depends on the stock size or biomass, the vessel efficiency (that is, the catchability coefficient) and the number of vessels taken out to fish in a given period. The underlying idea is that, other things being equal, the ability to harvest fish at any point in time is proportional to the biomass available in the habitat. This is particularly so in the case of non- schooling species such as hake. Details of this function are given in Appendix 2. The profit accruing to a given vessel group in any given year is the total revenue from fishing less the total cost of fishing in that year (see Appendix 2).

Stock dynamics and constraints The stock constraint in this model comes in the form of the total quotas fixed annually by the government. The players are free to maximise their profits from the fishery so long as their combined harvest does not exceed the annual quota. Given the assumption that quotas are optimally determined to ensure the long-term survival of the stock, they implicitly ensure that the underlying stock dynamics and constraints are respected all the time.8

The social planner’s objective It is assumed that the objective of the social planner, that is, the Ministry of Fisheries and Marine Resources of Namibia, is to choose ζ a sequence of quota shares, i,t (t=1,2, .. ,T), to obtain the highest possible discounted profit from the total quota, using social costs and prices.

4. Numerical Computations Three types of data are required for the computations planned, namely, quota, vessel, and economic data.

8 As mentioned elsewhere, the next paper in this series will explicitly model the stock dynamics of hake.

NUPI OCTOBER 00 Fish as Vehicle for Economic Development in Namibia 303

Quota data The average quota for the hake fishery from 1987 to 1994 is calculated to be about 75,000 tonnes, while the average for the years 1992 to 1994 is about 115,000 tonnes. The medium-term target of the government of Namibia is to achieve an annual quota of 150,000 tonnes. Barring unfavourable environmental conditions, the good resource management structure put in place by the MFMR since independence should make this target achievable into the future, hence we use this quota size in the analysis. It is, however, a simple matter to change this figure for the purposes of sensitivity analysis. Data on annual quotas for 1993 and 1994 show that while 19.4 per cent of the total quota in 1993 went to the wet fish trawlers, as much as 48.9 per cent went to them in 1994. Recall that the government has set a distribution target of 60:40 per cent in favour of the wet fish trawlers in the very near future; this target figure is what we use as the starting point for the analysis. In other words, we start the analysis by first testing the economic rationality of these target shares.

Vessel data The crucial data needed include (i) the total catch per season per vessel, assuming full-time fishing, (ii) fish landed per fishing power, which in the case of hake is the weight per horse power (HP) of fish landed (see Moorsom, 1994), (iii) the crew size on each vessel and how many of these are skilled and unskilled. All these data are either available in Moorsom (1994), or can be worked out from that paper. The averages of 1993 and 1994 fishing capacity and fishing power for the different class sizes of vessels are given in tables 2 and 3. Also included in the tables are the average crew sizes.

Table 2. Data on wetfish trawlers Size class (HP) 300–699 700–139 1400–2000

Capacity (tpa) 910 2,103 2,841 Fishing power (t/HP) 1.6 1.9 1.8 Average crew size 13 22 31

Source: Moorsom (1994)

NUPI OCTOBER 00 304 Ussif Rashid Sumaila

Table 3. Data on freezer trawlers

Size class (HP) 800–1499 1500–1999 2000–2499 2500+ Capacity (tpa) 3,312 4,742 6,590 8,381 Fishing power (t/H) 2.7 2.75 3.05 2.9 Average crew size 30 39 43 69

Source: Moorsom (1994)

Economic data: cost and price data The export prices of hake products vary considerably for different sizes of the same product. For instance, the different sizes of hake fillets went for anything from N$ 6.50 to N$ 8.70 per kilogram in 1994. Average prices are used in the analysis. To work out such prices, we make the following assumptions:

1. Catch by freezer trawlers is processed into an 80:20 mix of hake fillets and head and gutted (H&G) hake, respectively. 2. Catch by wet fish trawlers is processed into a 70:20:10 mix of hake fillets, H&G hake and fresh hake, respectively. 3. For the hake H&G product, 40 per cent are sizes 1 to 4, and 60 per cent are sizes 5 and 6. 4. For the hake fillet product, 25 per cent are sizes 3 and 4, and 75 per cent are sizes 5 and 6.

Using assumptions (3) and (4) as a basis, the average prices of hake H&G and hake fillets are calculated and presented in tables 4

Table 4. Calculating the average selling price of hake H&G Size N$/kg Average Proportion Proportionate N$/kg selling price

1 3.00 2 3.25 3 3.70 4 4.25 3.55 40 per cent 1.42 5 4,90 6 6.00 5.45 60 per cent 3.27 Average price 4.69

NUPI OCTOBER 00 Fish as Vehicle for Economic Development in Namibia 305

Table 5. Calculating the average selling price of hake fillet Average Proportionate Size N$/kg N$/kg Proportion selling price 3 6.50 4 7.50 7.00 25 per cent 1.75 5 8.10 6 8.70 8.40 75 per cent 6.30 Average price 8.05 and 5.9 We see from these tables that hake H&G commands an average price of N$4.69 per kg, while hake fillets command a price of N$8.05. To calculate the average price per kg of fresh hake, we compared the prices for frozen fillets with those for fresh hake fillets. This revealed that the latter commands a premium price to the tune of up to 100 per cent more than the average price per kg for the former. Indeed, fresh hake fillets achieved a price of up to N$ 16.10 per kg in 1994 and early 1995. From assumptions (1) and (2) and the calculations above, the average price per kg of processed fish from the freezer trawlers turns out to be N$ (0.8*8.05+0.2*4.69) = N$ 7.38, and for the wet fish trawlers, N$ (0.7*8.05+0.2*4.69+0.1*16.10) = N$ 8.18. The total social cost (both vessel and onshore) per kg of processed fish from wet fish trawler landings is calculated to be N$4.55. Similarly, the total social cost per kg of processed fish from freezer trawler landings is calculated to be N$4.76. To arrive at these figures, data from the MFMR and from fishing companies active in the Namibian EEZ were used.

The numerical results

The algebraic modelling language AMPL (Fourer et al., 1993) was used as computational aid. Combining the theoretical framework set out in section 3 and the data given in the first subsection in 4,

9 The calculations and assumptions underlying them are based on data obtained partly from the MFMR, partly from fishing companies active in the Namibian EEZ, and partly from personal communications.

NUPI OCTOBER 00 306 Ussif Rashid Sumaila

Table 6. The present value of (PV) of economic rent, employment and fleet sizes required to land different quota allocations Per cent share 50:50 60:40 80:20 100:00

RV of Rent (in bill. N$) 10.10 10.42 11.05 11.69 Fleet Size (w:f)15 27:16 32:13 43:7 53:0 (in no. of vessels) Employment (w:f) 3,976:624 4,712:507 6,332:273 7,804:0 Generation = 4,600 = 5,219 = 6,605 =7,804 (in no. of positions)

Notes: The fleet sizes and employment-generating ability are worked out using the frameworks given in appendix 2 and 3. Fleet sizes for vessels with size classes 1,400–2,000 and 1,500–1,999 HP, respectively, for w and f.

AMPL computed the results presented below. The AMPL model and data files used for the computations are given in Appendix 3. Also presented in the appendix are the frameworks for the calculations. The total present value of economic rent from the resource given the government target of 60:40 allocation of the annual quota to w and f is N$10.42 billion. On the other hand, the economically efficient allocation turns out to be 100:0 to w and f, respectively. This allocation results in a total present value of economic rent of N$11.69 billion. Thus, the economic loss due to the implementation of the current government target rather than the economically optimal share is N$1.27 billion, about 11 per cent of what is achievable. To facilitate the discussion to follow in section 5, we present in Table 6 a summary of the results highlighting economic rent, employment generation, and optimal fleet sizes for different allocations of the annual quota.

5. Discussion and Conclusion

Summary of findings The recent government policy to allocate more and more of the annual quota for hake to the wet fish trawlers is an economically sensible decision. However, the current policy target of 60:40 per cent of the quota to the wet fish and the freezer trawlers is sub- optimal. This would result in a total present value of economic rent of N$10.42 billion, which is about N$1.27 billion (about 11 per cent)

NUPI OCTOBER 00 Fish as Vehicle for Economic Development in Namibia 307 less than what is achievable under the optimal allocation, which is 100:0 in favour of the wet fish trawlers. With this allocation, a total PV of economic rent of N$11.69 billion is achievable. In terms of employment generation, more allocation to the wet fish trawlers is a good thing, as this class of vessel generates more that six times the employment generated by the freezer trawlers for the same quota allocation (see Table 6). It is possible to gener- ate up to 7,804 positions of various kinds annually from the activities in the hake fishery if the optimal solution is implemented. The fleet size necessary to land the optimal allocation of the quota is 53 wet fish trawlers of size class 1,400–2,000 HP or its equivalent. In the case of the declared government policy of 60:40 allocation, the necessary fleet sizes of both wet fish (size class 1,400–2,000 HP) and freezer (size class 1,500–1,999 HP) trawlers are 32 and 13, respectively.

Limitations and sensitivity analysis The main limitations of this study are to be found in the estimation and calculation of the parameters of the model. I discuss these below and offer sensitivity analysis, where necessary. The costs and prices used in the study are highly aggregated, derived from average prices and costs from a number of sources. I would therefore recommend that the model be re-run when more detailed data is available, most likely after the ongoing work on the fisheries database being developed by the statistics office in the MFMR is completed. To check the robustness of the present results against changes in costs and prices, sensitivity analyses are carried out. These indicate that our results with respect to optimal allocation remain valid so long as the relative prices and costs of the Freezer Trawler Group do not improve by more than 14 and 21 per cent, respectively, in relation to those of the Wet Fish Trawler Group. An interesting interpretation of this result is that the estimated prices and costs must deviate from their correct values by up to 14 and 21 per cent, respectively, for the results of the study to be invalidated. Varying the discount rate will surely affect the PV of economic rent achievable but will not affect the crucial decision on what share to be allocated to which vessel group. For instance, an increase in the real discount rate from the 2 per cent used in the computation to 3 per cent reduces the PV of economic rent to N$8.72 from N$11.69 billion.

NUPI OCTOBER 00 308 Ussif Rashid Sumaila

An increase or decrease in the projected annual quota for hake will again affect the economic benefit from the resource but not the results concerning the optimal allocation of the quota. Since management problems usually arise when the actual quota turns out to be less than the predicted, we decided to carry out a sensitivity analysis for which the quota is 25 per cent less than expected. In this case, the PV of economic rent is N$ 8.77 billion. The conversion factor (CF) from catch to processed product is a key parameter, by this we mean it is capable of turning the results of the study around if the estimated value is very different from the actual. Because of this, care was taken in estimating it.

Concluding remarks Based on the results outlined above, one may conclude that the freezer trawlers should be banned from the exploitation of hake altogether: both economic efficiency and employment generation criteria support this change. There are, however, other issues to be taken into consideration. First, we should be interested in benefiting from certain intrinsic advantages of harvesting hake with freezer trawlers. An advantage of the freezer trawlers (but a disadvantage of wet fish trawlers) which needs to be taken into account is the fact that generally freezer trawlers fish in deeper waters than their wet fish counterparts, thereby ensuring a wider spread of fishing activity than would be possible if only wet fish trawlers were employed. Such a spread is positive for the biological well- being of the habitat and the fish contained therein. Another point to note is that turning the hake fishery into a wet fish trawler-only fishery may be ill-advised. This is because building a large land- based industry in a situation where the industry can swing widely due to environmental and other human factors can be very risky. The task of isolating the risk factors and the potential benefits of using freezer trawlers is left to future research. Another important issue, not taken up in this article but addressed elsewhere (see Manning, 1998), is the question of rent capture, its distribution and Namibian participation and ownership in the fishing sector of the national economy.

NUPI OCTOBER 00 Fish as Vehicle for Economic Development in Namibia 309

References Fischer, R.D. and L.J. Mirman, 1992, ‘Strategic dynamic interaction: Fish wars’, Journal of Economic Dynamics and Control, Vol. 16, pp. 267–287, North- Holland. Fourer, R., D.M. Gay and B.W. Kernighan, 1993, AMPL: A Modelling Language for Mathematical Programming, South San Francisco, CA: Scientific Press. Hamukuaya, H., 1994, ‘Research to determine the biomass of hake’, Namibian Brief: Focus on Fisheries and Research, Vol. 18, pp. 73–74. Hannesson, R., 1993, Bioeconomic analysis of fisheries, London: Fishing News Books. Levhari, D. and L.J. Mirman, 1980, ‘The Great Fish War: An Example Using a Dynamic Cournot-Nash Solution’, Bell Journal of Economics, Vol. 11, pp. 322–334. Manning, P.R., 1998, Managing Namibia’s fisheries: optimal resource use and national development objectives, Ph.D. thesis, London: LSE. Ministry of Fisheries and Marine Resources, 1996, Fishery Statistics. Summary Statistical Report, Windhoek. Ministry of Fisheries and Marine Resources, 1997, Report of Activities and State of the Fisheries Sector, Windhoek. Moorsom, R., 1994, ‘An analysis of the fishing capacity of the Namibian fleet in quota-limited fisheries: A methodological summary’, NEPRU Working Pa- per No. 39. Munro, G. R., 1979, ‘The optimal management of transboundary renewable resources’, Canadian Journal of Economics, Vol. 12, No. 8, pp. 355–376. Reed, W.J., 1979, ‘Optimal escapement levels in stochastic and deterministic harvesting models’, Journal of Economics and Environmental Management, Vol. 6, pp. 350–363. Sumaila, U.R., 1995, ‘Irreversible capital investment in a two-stage bimatrix game model’, Marine Resource Economics, Vol. 10, pp. 163–183. Sumaila, U.R., 1997, ‘Dynamic strategic interaction: the case of cod-capelin in the Barents Sea’, Marine Resource Economics, Vol. 12, pp. 77–94. Wysokinski, A., 1986, ‘The living marine resources of the Southeast Atlantic’, FAO Fisheries Technical Paper, 178 Revision 1.

NUPI OCTOBER 00 310 Ussif Rashid Sumaila

Appendix 1: Cost functions To get hold of the costs, we split the inputs that go into the harvesting and processing of hake into the following cost elements.

Cost elements associated with the freezer trawlers

Labour costs The key variables that go into labour costs include the size of the crew on a typical freezer trawler; the number of officers and skippers on the vessel; and the cost of hiring this class of labour for a given period of time to land a certain quantity of fish. These are

used to compute the expected cost of engaging skilled ( Esc, f ) and unskilled ( Euc, f ) crew members as well as skippers ( Esk, f ) and 10 officers ( Eoff , f ) to produce a unit weight of frozen hake. Here, the subscript sc, uc, sk, and off stand for skilled crew, unskilled crew, skippers and officers, respectively, while f refers to the Freezer Industry Group. Basic economic theory postulates that the socially optimal E is the alternative cost of the labour in question.11 We assume here that the wages being currently paid out to skilled crew members, officers and skippers represent the alternative cost of their labour. This can be justified by the fact that there is no oversupply of this class of labour in Namibia. On the other hand, due to the high level of unemployment among unskilled crew members in the country, the current wages received by this class of workers are well above the alternative value of their inputs. Thus, current earnings of unskilled labour are adjusted by introducing a kind of ‘discounting’ parameter for unskilled labour costs. Now, the total annual cost of engaging the required labour force

for a given vessel, kl, f , can be expressed mathematically as =+α ϕ (1) klf,,,,,()EEEE ucf ++ scf skf offf f ϕ where f is the annual fishing capacity of a freezer vessel in unit weight; k denotes costs; and the subscript l stands for labour as a

10 It should be mentioned here that normally payments to crew members, skippers and officers are split into two, that is, a fixed and a variable part. The latter depends on actual landings. 11 Note that socially optimal is used here in an economic sense. For instance, by socially optimal cost, we mean minimum cost incurred by society as a whole to achieve a stated objective, which in most cases is different from private optimal cost.

NUPI OCTOBER 00 Fish as Vehicle for Economic Development in Namibia 311

whole. The parameter 01<≤α is the percentage of unskilled crew labour earnings that can be said to be the alternative value of this labour. Notice that in the case of private costs.

Capital costs The cost of acquiring a fully equipped freezer trawler will form the basis for calculating annual capital costs here. Suppose this cost is ϑ denoted by f , then the annual user cost of capital, kc, f (where c stands for capital), can be expressed as12 =+σ +ρ ϑ (2) kcf, ()r f f f

Here, r is the real interest rate, the rate of depreciation of a ρ typical freezer vessel, and f denotes a foreign exchange premium. The latter is necessary to account for the fact that foreign exchange is not completely determined by the market in Namibia, and that virtually all capital costs incurred in acquiring a trawler are incurred in foreign currency.13 All these parameters are in per cent of the capital acquisition value. Again, there is a difference be- ρ tween private and social costs here: f is zero in the case of σα = f 1 private costs because private agents care only about their direct private costs.

Operating expenses These include annual costs of fuel and lubricating oil, repairs and maintenance, fishing gear renewal, telecommunication and radio expenses, and management and administrative costs. Others are general insurance cover for crew, catch, cargo, hull and machinery; license fees, levies and charges; harbour fees; and provisions. All these would be counted by the agents as part of their costs, but social costs would not include license fees, levies and charges. The annual costs listed above, excluding license fees, levies and charges, are added together to obtain the annual social costs of landing and processing the average annual harvest.

12 This is true only under the assumption that capital gains or losses due to changes in the acquisition value of capital are so low that they can be neglected. 13 Labour costs are not subjected to this premium because the bulk of these are paid in local currency.

NUPI OCTOBER 00 312 Ussif Rashid Sumaila

Cost elements associated with the wet fish trawlers

Labour costs In addition to the offshore labour costs mentioned under freezer trawlers, some onshore labour costs are incurred in processing in the case of wet fish trawlers. Thus annual labour costs can be expressed as follows: =+αϕ (3) klw, [](EEuc,, w+ ou w ) + E sc , w + EEsk,, w off w + Eos, w w ϕ where w is the annual fishing capacity of a wet fish vessel in unit weight; EEou,, w and os w denote the cost of skilled and unskilled labour needed to process a tonne of hake onshore; and

Euc,,, w, E sc w,, E sk w and E off , w represent the various offshore costs.

Capital costs There are two components to capital costs here, viz., the cost of acquiring a fully equipped wet fish trawler and the cost of laying down the necessary infrastructure to process the fish onshore. These two will form the basis for calculating annual capital costs. The ( formulation of costs in this case is exactly as in equation 2 above, except that an extra cost component is introduced to capture the costs of laying the necessary onshore processing infrastructure, ϑ Ζ,w, where the subscript Z denotes infrastructure.

Then the annual user cost, kc,w, of capital can be expressed as

where the subscript v denotes vessel. Notice that, in principle, the depreciation rate and the foreign exchange premium for infrastructural and vessel costs can differ.

Operating expenses Here too there are two components, expenses related to vessel operations and expenses related to onshore processing. The former expenses include annual costs of fuel and lubricating oil, repairs and maintenance, fishing gear renewal, telecommunication and radio expenses, and management and administrative costs. Others are costs of general insurance cover for crew, catch, cargo, hull

NUPI OCTOBER 00 Fish as Vehicle for Economic Development in Namibia 313 and machinery; license fees, levies and charges; harbour fees; and provisions. Similar cost elements related to onshore processing are added to the costs stemming from vessel operation to get the total cost. In this case too the annual costs under all the items above (except license fees, levies and charges) are added together to obtain the annual social costs of landing and processing a certain quantity of wet fish landings.

Appendix 2. Production and profit functions

The production and profit functions In both theoretical and applied fishery economics, it is common to use the following production function (see for instance, Reed, 1979 and Hannesson, 1993): == (5) hit,,qxe i t it i w, f where h is the vessel and time-dependent size of the harvest, x is the time-dependent stock size or biomass, the parameter q denotes the vessel-dependent catchability coefficient, and ei,t denotes the number of trawlers of type i taken out to fish in period t. The main assumption underlying h is that the ability to harvest fish at any point in time is proportional to the biomass available in the habitat. The simple logic here is that it is much easier (and by extension less costly) to harvest fish in a habitat full of fish than one virtually empty of fish. This is particularly so in the case of non- schooling species such as hake. In the case where quotas are exogenously determined as in this model, h is necessarily equal to the quota: due to the potential for making pure profits, one would expect the agents in the model to harvest up to the quota allocated. Hence, equation (5) above can be written as ==ζζ = = (6) hit,,,qxe i t it itQt , i w, f and ∑ it ,1 i ζ Here, Q is the time-dependent annual quota for hake and i is the share of the annual quota allocated to agent i. Note that the number of vessels, e, needed to efficiently land the quota allocated to each agent can be calculated if qi and x are known. Now, the profit to a given agent in any given year can be written as π = −φ = −φ ζ = (7) i,t ( pi i )hi,t CF ( pi i ) i,t Qt CF, i w, f

NUPI OCTOBER 00 314 Ussif Rashid Sumaila

where p is the average market price per kg of the products from wet fish and freezer trawler landings, CF denotes the conversion factor from catch to processed fish, andφ denotes the average cost that must be incurred to land and process the same weight of their landings.

Appendix 3. AMPL files ###AMPL model file: Gives the model statements in the AMPL language###

param T>0 integer; # the time horizon of the model, fixed at 100 years. set Player; # the players are w and f set Periods:= 1..T; param price {Player}; param cost {Player}; param Quota {t in 1..T}>0; param CF {Player}; # the conversion factor param DF>0; # the discount factor var Share {p in Player, t in 1..T} >=0; # the variable in the model: share of quota maximise present_value: sum {t in 1..T} (sum {p in Player} Quota[t]*CF[p] * Share[p,t] * DF^t * (price[p] - cost[p]) ); subject to limit1 {p in Player,t in 1..T}: 0<= Share[p,t] <=1; subject to limit2 {t in 1..T}: sum{p in Player} Share [p,t]<=1;

###AMPL data file: Gives the baseline data for the model### set Player:= wet freezer; param T := 100; param: price cost CF := wet 8180 4550 0.505 freezer 7380 4760 0.51;

param Quota default 150000; param DF := 0.98;

NUPI OCTOBER 00 Fish as Vehicle for Economic Development in Namibia 315

Appendix 4: General framework for calculating optimal number of vessels Quota allocation = v tonnes Capacity of vessel (tpa) = c tonnes Optimal number of vessels = Quota allocation/ vessel capacity = v/c

Appendix 5: General framework for calculating labour requirements For freezer vessel Crew size per vessel = m Optimal number of vessels to land quota allocation = r Employment generation = crew size*optimal number of vessels = mr

For wet fish vessel Crew size per vessel = s Optimal number of vessels to land quota allocation = t Direct employment generation = crew size*optimal number of = st Add employment generation in processing = u Total employment generation = mr+st+u

NUPI OCTOBER 00 316 Ussif Rashid Sumaila

NUPI OCTOBER 00 Debate 317

Debate

Interdisciplinarity in Development Studies: Myths and Realities

Elisabeth Molteberg, Cassandra Bergstrøm and Ruth Haug

In development studies, as in many other academic pursuits, interdisciplinarity seems to be an intriguing and promising, but vague and elusive concept to scholars and institutions. Advocating interdisciplinary approaches as necessary in addressing complex and broad-ranging issues and problems is not only politically correct but also, for many, intuitively so. Yet, what exactly these approaches consist of, how to go about using them, and what implications they have for individuals, teams and institutions involved in these ventures, often remains unclear. This is not to say that these issues have not been addressed. The debate on interdisciplinarity is perhaps as old as science itself. During the last decades it has re-emerged as a reaction to what many experience as the fragmentation of science and its inability to address some of the world’s most pressing problems. Recent years have witnessed a proliferation of works which claim the need for such an approach and/or that an interdisciplinary way of working has been carried out. The literature is far too extensive to be discussed here, but a few examples should be mentioned. A seminal work is CERI (1972), based on results from a seminar, which presents survey results as well as a range of cases. It is perhaps more interesting for its ability to raise questions and demonstrate the complexity of its topic than for its analyses, which tend to be contradictory – un- avoidable when so many views are represented. However, several interesting points are made. Interdisciplinarity is termed a ‘polemic practice’, and the need for a common, or harmonised, vocabulary is emphasised. Moreover, the issue of different conceptualisations of interdisciplinarity is brought up, a point also discussed by Klein (1990) in her historical overview of the concept. Both works show the existence of two very different conceptualisations of interdisci-

NUPI OCTOBER 00 318 Elisabeth Molteberg, Cassandra Bergstrøm and Ruth Haug

plinarity, which are reflected in people’s scientific ambitions towards holism or reductionism, as well as through the way they choose to work. Klein argues that these conceptualisations form the extremes of a continuum of approaches. At one end is multidisciplinarity, a highly additive form. At the other is transdisciplinarity, a highly integrative form. We agree that it is important to be aware of the distinction between these conceptualisations, but disagree that distinctions are most fruitfully expressed in typologies or dichotomies. This topic is addressed below. With specific reference to development studies, Desmond McNeill recently published a paper (McNeill, 1999) where he makes a number of points that we believe are grounded in myths concerning the nature of interdisciplinarity. Here we take issue with these myths because we think they distract the debate. It is our contention that the time has come to move beyond problem-naming and static definition. As we see it, current challenges lie in operationalisation and therefore inherently in conceptualisations of interdisciplinarity which enable this. These are large-scale questions, the answers of which may be expected to differ from case to case. As such, one should probably not strive to find one universal answer, but expect continual debate about the concept of interdisciplinarity and the issues involved. That said, certain underlying general traits seem to be identifiable, enabling us to broaden and refine our understanding of the general concept. We recognise these traits as integral to the continuing debate on interdisciplinarity. We would like to discuss them and their implications briefly, with the goal of enabling, rather than constraining, a fruitful discussion about interdisciplinarity with particular reference to development studies research.

A Conceptualisation of Interdisciplinarity Discussion of interdisciplinarity is often hampered by failure to consider the ambiguity of the concept. It is crucial to clarify how one defines it, but often this is not, or is insufficiently, done. The result can be confusing and misleading argumentation. The conceptual ambiguity, while spanning several dimensions, is rooted in a failure to consider degree of integration. There is a continuum of degrees of integration involved in interdisciplinary activities, spanning from no integration whatsoever (pure unidisciplinarity) to total integration (a unification of all science). Between these extremes, which may not exist in reality, we find integration levels such as the incorporation of supportive evidence to one’s ideas,

NUPI OCTOBER 00 Debate 319 attempts to combine disciplinary results and a comprehensive unification of theory from various disciplines. Any particular interdisciplinary process, as well as its outcome(s), is found somewhere within this continuum. The higher the degree of integration, the greater the epistemological and ontological implications of the initiative. In this respect, there is a difference between two tendencies, or areas within the continuum, that can be termed additive and integrative interdisciplinarity. They are characterised by whether the concern is predomi- nantly with interdisciplinarity as praxis or as theory integration. Concrete cases of interdisciplinarity are apt to include elements of both. Additive interdisciplinarity implies merely combining pieces of a puzzle, while integrative interdisciplinarity implies acquiring a basic perspective which is different from those offered by disciplines involved – a different puzzle altogether, as it were, or in the words of Gregory Bateson, a whole that is greater than the sum of its parts. The term multidisciplinarity has often been used to refer to purely additive forms that interdisciplinary ventures can take, and mega- or transdisciplinarity to refer to forms with a high level of integration and unification. Given our recognition of the processual nature of scientific activities, which calls for emphasis on movement within a continuum, we choose instead to distinguish between concrete cases of interdisciplinary activities and tendencies or levels of integration. Interdisciplinarity should be seen as processual and dynamic. The level of integration may, as a consequence, be expected to change or fluctuate over time. For example, teams that work addi- tively several times in a row may find they approach work in a more integrative way over time, as they develop a common terminology and learn more about each other’s fields. Conversely, education programmes aiming for integration may fall short of their ambitions due to an additive syllabus and a lack of teaching staff able to integrate what is taught. Different modes of interdisciplinarity may also affect one another, such as individual, group and institutional interdisciplinarity, and short- and long-term activities. Conventional terminology of interdisciplinarity, with its focus on the multi-/interdisciplinarity distinction, may encourage conceptualisation in terms of a dichotomy, a pair of opposites. As such, it implies a static and monolithic view of multi- and interdisciplinarity as distinct, internally coherent forms. We find that discussing levels of integration liber- ates us from the restraints of this dichotomy and allows for a more refined conceptualisation of interdisciplinarity. It is our hope that the usefulness of this approach will become apparent below.

NUPI OCTOBER 00 320 Elisabeth Molteberg, Cassandra Bergstrøm and Ruth Haug

Degree of integration has many implications, including those concerning

t approach: interdisciplinarity can be considered a way of doing something, or a way of thinking. It can be a method and/or a methodology. t time span: interdisciplinarity can be a single, or a series of, one- off enterprises, or a long-term process. t personnel involved: interdisciplinarity can be an individual, group- or team-based, or a more institutional-level approach.1 Additive efforts tend to involve teams, but teams or groups can also work integratively. Institutional-level approaches tend to involve integrative ambitions, but often gravitate towards additive forms. Individual interdisciplinarity generally (but not always) involves a rather high level of integration.

Myths of Interdisciplinarity in Development Studies The vague nature of interdisciplinarity has led to the argument, widely accepted and supported, that it is also very problematic in nature, rendering it controversial in many circles. We will examine more closely two major premises for this argument. One of these premises is based on the notion of a schism between natural and social/humanistic science. This schism, it is argued, must be bridged in order for development studies – a scientific field where the study of interfaces and interactions between natural and social/human domains is central – to be successful. Yet, bridging it is exceedingly difficult. This premise supports claims that degrees of interdisciplinarity that seek to transform and unite forms of knowledge are difficult if not impossible to achieve, especially when involving disciplines that belong to different sides of the ‘cultural divide’ between natural and social/humanistic sciences. The other premise is based on the notion of a problematic relationship, a too close link, between research and policy application in development studies. This may allegedly threaten the quality and legitimacy of development studies research.

1 This three-way distinction may indicate that we are neglecting many approaches more comprehensive than those of a team or small group, such as research or education programmes, sections within departments etc., that don’t normally come to mind when talking about institutions. Our use of the term ‘institution’ is meant as shorthand for all of these.

NUPI OCTOBER 00 Debate 321

These premises derive from concrete experience of difficulties. Interdisciplinary activities do involve challenges of communication and quality assurance, which certainly merit discussion. However, such sceptical conclusions reflect an implicit view of interdisciplinarity that is not wholly unproblematic. Moreover, the level of integration discussed is often not made explicit. We hold that certain myths about interdisciplinarity have established themselves, and have become constraints to a constructive discussion of how to handle the challenges involved.

Myth 1: Disciplines are homogeneous, monolithic entities Many arguments against integrative forms of interdisciplinarity stress the difference between disciplinary ‘cultures’ and approaches to knowledge generation. These are often consistent with, or even refer to, Thomas Kuhn’s ideas about the incommensurability of scientific paradigms.2 They thus make themselves vulnerable to the same critique that has been made of Kuhn’s work: that it is anchored in an individually oriented view of disciplines as traditions in which an individual finds a home – ‘one person, one discipline’ and, moreover, that it tends toward a monolithic view of disciplines as homogeneous entities with clearly defined borders. This view fails to consider the collective nature of science, and the heterogeneous nature of disciplines. A discipline is not a static entity. In every discipline it is possible to find people who are in greater agreement with certain people outside their disciplines than with certain people within – this is a result of the arbitrariness of disciplinary borders. It is important to recognise the importance of dynamic collective processes when conceptualising disciplines and academia.

2 Kuhn (1996) argues that paradigms are not only incompatible, i.e., cannot be combined and united, but that they are also incommensurable, i.e., mutually exclusive in a profound way: paradigms are structuring principles for our views of reality and our thoughts about it, which makes it impossible to relate to more than one paradigm at a time. Kuhn concludes from this that scientists within different paradigms cannot communicate – all thought loses its meaning when it cannot refer to, or be seen in the context of, the ordering principle of reality in which it came about. Thus trying to relate one’s thoughts to someone who operates within a completely different ordering principle is impossible. The thoughts won’t make sense – or at least not the right sense.

NUPI OCTOBER 00 322 Elisabeth Molteberg, Cassandra Bergstrøm and Ruth Haug

Myth 2: Theory from natural and social/human sciences cannot be unified The individualistic conceptualisation of disciplines discussed above has implications for one’s belief in the possibility of integrative interdisciplinarity. To someone who does not believe in the possibilities of integrating and unifying theory from natural and social sciences, the whole area of development studies that addresses relations between people and the environment should then seem a lost cause, because one of the prime contributions that can be made by this field is precisely that which involves integration of social and natural science theory. However, this is a very successful area, which has made great contributions to theory, also in other fields and in more unidisciplinary research. We agree that there are ‘cultural’ differences (methods, approaches, styles) between disciplinary traditions, and that there is a divide between diagnostic and predictive sciences. But again, we think dividing lines and communication breakdowns may reflect something other than just disciplinary traditions. Moreover, there are connecting lines that criss-cross disciplinary borders and influence patterns of communication and co-operation between scientists. In this respect, it may well be more enlightening to focus on scientists’ underlying epistemological and ontological assumptions, which are to some extent part of their academic training but also a question of personal world-views. We believe that a debate about philosophical positions, aiming to find some common platform, would be a valuable contribution to interdisciplinarity of all levels of integration.

Myth 3: Interdisciplinarity is futile. Science is characterised by fragmentation processes, which increase boundaries and bar communication It is a widespread idea that integrative interdisciplinarity is a losing battle against prevailing fragmentation processes in science, which create ever new boundaries between bodies of knowledge. Often- cited examples of this include the substantial controversies between natural and cultural geographers, between sociobiologists and soci- ologists/anthropologists, and between institutional and neo-classical economists. But there are also processes in science that create new interfaces. Scientific arenas emerge where certain overarching problems are addressed (gender studies and property rights studies are two examples) that cut across disciplines and become areas of expertise in their own right, whose contributors may have back-

NUPI OCTOBER 00 Debate 323 grounds as diverse as biology, agricultural sciences, law, history, anthropology, psychology, economics, linguistics, and so on. These arenas are collective thrusts towards creating a synthesised understanding about a certain field, drawing on bodies of theory and empirical material from a heterogeneous range of disciplines. Some of the contributors to these fields become true interdisciplinarians, able to grasp and use theory from different fields to create new theory and elucidate their problem area. Development studies is another example of such an arena, though perhaps less clearly defined. Interfaces also appear when theories, models or metaphors are used or examined in new ways wholly or partly outside their ‘home’ discipline, such as the work of Darwin or Marx. When used to examine one’s own work and ideas, these different ways of conceptualising may be selectively incorporated, used creatively to rethink and further support one’s existing ideas and conceptualisations of a problem, or to transform these. Transformation enables one to conclude or think in ways that would not have been possible without these inputs. Moreover, broad academic or philosophical issues may gener- ate interesting work across disciplinary boundaries, such as the long and heated nature/nurture debate, considering whether, or to what extent, human behaviour is primarily to be understood as natural (genetic, physiological, biological) processes or as environmentally and cognitively determined (psychological and sociocultural processes). Contributors to this debate have come from disciplines such as sociology, anthropology, psychology, and biology. They have on occasion been strongly opposed to the ideas of fellow disciplinari- ans, or inspired by people from other disciplines. A recent example of the latter might be the anthropologist Thomas Hylland Eriksen’s fascination with Darwin and his book on egoism with the biologist Dag O. Hessen (1999).

Myth 4: Interdisciplinarity is a team sport; all individuals and institutions should have a disciplinary basis Failure to distinguish between different levels of integration in interdisciplinarity may lead to the implication, conscious or not, that interdisciplinary research can be defined as a team effort between individuals that belong to different disciplines. We hold that this neither reflects reality, nor the ideal. A lot of interesting and ground- breaking scientific work happens on the interface between disci-

NUPI OCTOBER 00 324 Elisabeth Molteberg, Cassandra Bergstrøm and Ruth Haug

plines. A number of academic institutions employ staff with different uni- or interdisciplinary backgrounds, aiming to address issues in their teaching and research in an integrated way. Similarly, we see no logical reason why the contributions of individuals working interdisciplinarily should be less significant than that of a team. In fact, given the foundation that needs to be laid and the threads developed, the work of an individual may arguably be more cohe- sive than the work of a team. Some examples are Amartya Sen, Johan Galtung, Gregory Bateson, Sara Berry and Sally Falk Moore. Moreover, in this day of interdisciplinary programmes and institutions, a host of individuals receive their academic training in a range of academic institutions, and they go on to do interdisciplinary research, more anchored in particular problems than in particular disciplines. There are, however, problems of academic respectability and legitimacy concerning individual interdisciplinarity. These should definitely be discussed and analysed in a debate on interdisciplinarity. An individualistic and monolithic conception of disciplines and science may also be the basis of the often-repeated argument that an individual cannot contribute to interdisciplinary research without a strong unidisciplinary background. We disagree strongly with a categorical rejection of the possibility of being a good interdisciplinary scientist. Disciplines are like languages or cultures in the sense that no one individual in a community has command or knowledge of the totality of its expressions (vocabulary and grammar, or cultural traits). There is another parallel between bi- or multicultural- ism and interdisciplinarity, too: often, bi- or multicultural persons cannot point to a primary ‘home culture’, nor do they simply add up or even blend their cultures. Rather, they often develop a cultural repertoire and life-world framework distinct from the sum of the parts.3 And just as the languages and cultures in such people’s command may differ considerably, the disciplines a person may draw on in his/her academic activities may be quite different and belong to different sides of the natural/social-human science divide.

3 This is a common phenomenon, e.g. among children of emigrants or missionaries, who are brought up by their family in one cultural tradition within a larger framework of another. These children are often referred to as ‘third culture children’, reflecting the uniqueness of their cultural experience, which they can fully share only with others who have been brought up under similar conditions. This experience is also expressed in several literary works by bicultural writers, such as Salman Rushdie, Hanif Kureishi and Amy Tan.

NUPI OCTOBER 00 Debate 325

Creating academic ‘home bases’ for individual, group and institutional interdisciplinarity is difficult within the existing structure of academia, where unidisciplinarity is the organising principle. How- ever, as we have argued, there is no logical reason to assume that knowledge produced in an interdisciplinary mode should be less valid and rigorous than disciplinary knowledge. The struggle for legitimacy and viability often faced by interdisciplinary teaching and research efforts and institutions does not necessarily derive from poor quality, but often has other causes. One important cause is a fragmentation of experience: new efforts often have to reinvent what already exists, rather than standing on the shoulders of former efforts, benefiting from their experience. Another is the general pressure for disciplinary structuring within academia, with all its concomitant problems related to understanding (a lack of criteria for evaluating interdisciplinary research and for categorising those involved in it and their products) and power (scepticism of the controversial, fear of competition and power erosion).

Myth 5: Interdisciplinary science = applied science = inferior science Considering the scientific status and legitimacy of development studies, the close relations between (particularly social) science and policy work are often a source of concern. This is development studies’ own brand of the general notion in science that applied work is inferior, does not contribute to theory development and does not merit academic recognition. The fact that development studies research is often interdisciplinary appears to increase this concern. This may be due to an assumption that an interdisciplinary approach is solely useful for solving particular problems, an assumption that arises either from a failure to distinguish between integrative and additive approaches, or from scepticism towards the feasibility of integrative interdisciplinarity, as discussed earlier. The bulk of development studies research is applied, in the sense that it addresses real-world problems and aims to contribute to their solution. It is knowledge generated with the aim of informing action; it is not ‘pure’ in the sense that it is undertaken with the sole aim of contributing to an understanding of the nature of the world. But it does take on this challenge as well – it seeks to do both.4

4 Medicine is an example of a very prestigious science where the bulk of research is applied, in a similar sense as in development studies.

NUPI OCTOBER 00 326 Elisabeth Molteberg, Cassandra Bergstrøm and Ruth Haug

This, and the fact that the same people may be involved both in research, policy and implementation activities, implicates policy/ research relations that may be too close for comfort. For instance they may, based on partly the same findings, publish reports with policy recommendations and research papers problematising policy. There is a concern that these different activities and professional identities will mesh, threatening the integrity of research. It is a challenging balancing act. Adding to this the ambiguities of interdisciplinary research, one has ample material for controversy. But distancing science from policy, withdrawing to the ivory tower, is not a viable option. The close relations offer possibilities as well, which have to do with relevance, timeliness and access to information and alternative perspectives – from within and outside academia. Concerning the scientific legitimacy of this applied brand of science, it appears to have reached maturity and a critical mass. There is a large academic community of people that write, publish and create names for themselves in this field, in prestigious institutions across the globe. There are journals, and there are peers that can evaluate both research results and recruits, at least within the broader topical fields. The problem rather seems to be the position of interdisciplinary institutions and, perhaps even more so scientists, within particular universities. One should bear in mind that the debate around these themes is as much strategic as scientific. The concern about quality of science is often used in a struggle for legitimacy, definition power, prestige and hegemony in academia. This increases both the importance and the difficulty of making a strong case for the scientific validity of interdisciplinary approaches and specialisation, and of the field of development studies in general. The sceptics’ line of arguments does have weak points. For one thing, the dichotomy ‘basic’ versus ‘applied’ science is more mud- dled than one might think, with very different connotations in different disciplines. What is the criterion for distinguishing between basic and applied science? Sciences differ in their approach to knowledge production, and so do their distinction criteria. In many natural sciences, the approach tends to consist in applying methodological procedures deductively to problems. These sciences can offer expertise consisting of these methodological procedures, which can be applied to concrete problem solving in a deductive manner, de- tached from theory. The question of whether research is basic or applied in these disciplines seems to be determined largely by whether a certain methodological procedure is applied in order to solve a particular concrete problem, or if it is applied with the aim to further

NUPI OCTOBER 00 Debate 327 general knowledge. This either-or distinction does not accommo- date cases where the results of a problem-solving exercise are used in theory development, or where basic research has direct relevance to concrete problem solving. This is also true for sciences that produce science by critically assessing others’ problem analyses, like most social and human sciences. Here there is another problem, too: these sciences work by the same procedures whether they use them for basic or applied work. Again, the terms ‘applied’ and ‘basic’ become unclear, as some people tend to distinguish between the two forms of research according to who formulated the research problem in the first place: applied research is problem-solving (the problem being formulated by the funding agent), while basic research is problem formulation. But reality is more complex. A lot of basic research, in order to receive funding, has to be directed towards pre-defined general themes and issues, and has to be justified in terms of applicability and relevance to concrete problem areas. For example, the Norwegian state funding agency organises a lot of its funding through research programmes that come with defined problems, and they expect applicants for basic research grants to justify their grant applications in terms of applicability. Conversely, in certain cases, commissioners of applied research may commission research on a broadly defined topic, giving researchers a central role in dialogues formulating the problem to be researched, and welcoming critical problem discussion in the documents presenting the outcome of the research. We conclude that there is no simple and clear-cut distinction between basic and applied research to merit the categorical statements about their difference paraphrased above. Interdisciplinary research that spans social and natural sciences may be hard to categorise because of these differences. Integra- tive research may suffer as a result – being mistrusted, ignored or misconstrued as mere problem-solving work. But while we agree that much of what is published in the field of development studies is of a specific problem-solving nature, having limited or no contribution to make to theory-building, we also hold that there is a wealth of sophisticated and innovative theory-building research in this field, with influence that reaches beyond its borders. This research is not looked down upon nor barred from academic recognition. This indicates that the source of concern may be categorisation problems rather than academic quality. It is important to distinguish between report writing and research, allowing for a distinction between specific, problem-solving activities and applied social research with a capacity to contribute to theory-

NUPI OCTOBER 00 328 Elisabeth Molteberg, Cassandra Bergstrøm and Ruth Haug

building. This helps us to argue for the integrity of development research without getting distracted by who pays for it or whether the empirical material is also used for non-research purposes. As in other scientific pursuits, a lot of policy-related work in development studies falls into the research category. The SSE programme5 might be a Norwegian example of an interesting mix of basic, applied, and commissioned research and policy work. Here there are various constellations of scientists from natural, social and human disciplines working with policymakers, engineers, end users and others on problem-solving activities which have been documented through publishing both reports with specific recommendations and scientific writings with much more general relevance.

Conclusion Interdisciplinarity, as any methodological or theoretical approach, is a means to an end, and must be judged by its merits. But we cannot see these merits clearly if we don’t dispose of preconceived notions like the ones discussed here, and try to come up with fresh ideas about how to handle the challenges and make the most of the potentials that interdisciplinary approaches offer. As we have discussed, arguments against interdisciplinarity can be grounded in a failure to distinguish between different levels of integration and types of approaches. They can also be grounded in problematic conceptions of disciplines, science and the relationship between theory-building and applied science. Common to these arguments is that they build on assumptions and observations that are not systematised and analysed. The amount of research done on interdisciplinarity is relatively modest – there is a body of interdisciplinarity theory, but without ample empirical underpinnings. It is ironic that a discussion among scientists about science is so under-researched. This is one of the challenges faced by interdisci-

5 The Sahel-Sudan-Ethiopia Programme, launched in 1985 after the 1984/85 Sahel drought to channel Norwegian assistance to countries affected by severe drought, poverty and environmental degradation. Its thematic focus was food security, ecological rehabilitation and women’s situation. The main geographical focus was on Ethiopia, Mali, Sudan and later Eritrea. The programme has three components: development aid through non-governmental organisations’ projects, development research and support to multinational organisations’ projects. In the proceedings from a summary workshop in 1996, it says in the introduction ‘mutual benefits and synergy effects are expected to result from interaction and cooperation between the three components’ (Molteberg and Utvær, 1996: 1).

NUPI OCTOBER 00 Debate 329 plinary science – to document and analyse experience more syste- matically. Individual scientists, groups and institutions working in an interdisciplinary mode face other challenges. Interdisciplinarity is, like all other research, a ‘dance’ between theory, methodology and findings. But this mode of working faces certain particular challenges and problems, concerning reflexive processes, terminology development (communication), methodological and theoretical development and approaches, and delineation of a problem area. As for team and institutional interdisciplinarity, it may be seen as a methodological process, a way of working, as much as an analytical approach. This indicates a need for more time to be spent on reflection (e.g. on our own epistemological and ontological assumptions, our communication skills, and the way we structure interdisciplinary processes) and on mutual learning, collective work on manuscripts, etc., than what one would expect in unidisciplinary collaboration and institution-building. Communication fora become more important. One of the results of more teamwork and team reflection/analysis/writing is a common terminology, a sorting out of the (to the team or individual involved) most central of the terminological problems which interdisciplinary work is fraught with: the same terms have different meanings in different disciplines, and/or the same notion or phenomenon has different terms.6 Exploring the languages of different disciplines is a time-consuming process. However, the contribution to mutual accessibility involved in this work of translation is among the important contributions that individual interdisciplinary scholars can make. They may act as academic translators, conveyors of messages between representatives and audiences of different disciplines. This can be useful in interdisciplinary pursuits, as well as in communication of disciplinary work to a broader audi- ence.

References Bateson, G., 1972, Steps to an Ecology of Mind, New York: Ballantine Books. CERI, 1972, Interdisciplinarity: Problems of Teaching and Research in Univer- sities, Paris: OECD. Hylland Eriksen, T., 1997, Charles Darwin, Oslo: Gyldendal. Hylland Eriksen, T. and D.O. Essen, 1999, Egoisme, Oslo: Aschehoug.

6 Some examples off the cuff are system, organism, organisation/organise, resili- ence, productivity, actor, agency, law, rule, selection, knowledge, sustainability, efficiency, interaction, regime and resource.

NUPI OCTOBER 00 330 Elisabeth Molteberg, Cassandra Bergstrøm and Ruth Haug

Klein, J. T., 1990, Interdisciplinarity: History, Theory and Practice, Detroit: Wayne State University Press. Kuhn, T., 1996, 3rd edition, The Structure of Scientific Revolutions, Chicago: University of Chicago Press. McNeill, D., 1999, ‘On Interdisciplinary Research: with particular reference to the field of environment and development’, Higher Education Quarterly, Vol. 53, No. 4, pp. 312–332. Molteberg, E. and A. Utvær, eds., 1996, Honne ‘96. Ten years with SSE: Past experiences – future challenges. Proceedings from the SSE workshop at Honne Conference Centre, Biri, Norway, 2– 6 September 1996.

NUPI OCTOBER 00