Discursive development order and local informal practices
A development project in Northern Ethiopia

Jon Harald Sande Lie

University of Oslo, Department of Social Anthropology
Thesis submitted for the Cand. Polit. degree


To download, print, or bookmark, click: http://www.anthrobase.com/Txt/L/Lie_J_S_01.htm.
To cite, quote this address and the download date. Not for commercial use.
2004 Jon Harald Sande Lie. Distributed with permission, by www.AnthroBase.com.
Do not remove this notice from digital or paper copies of this text. 




Chapter 1: Introduction
The Problem
    - Some Reservations
The Context and the Actors
    - Afar Region and Aba'ala
    - NORAD and Norwegian Development Assistance

    - The Development Fund
    - Mekelle University
    - The Department of Agriculture

Methodological Considerations
    - Various Types of Data
Preliminary Theoretical Overview
    - Discourse as a System of Knowledge
    - Some Critical Remarks of a Discursive Approach
    - Actor Orientation and Informal Practises
    - Brief Outline of Thesis

Chapter 2: Development conceptualised
The Development of "Development"
    - Evolvement of Positivistic Ideas about Progress and Development
Two Grand Theories on Development
    - Modernisation Theory
    - Dependency Theory

The Post-Structural Approach To Development
    - The Development Apparatus
    - Discourse and Agency

Taking Post-Development Theory Further
    - Criticism of Post-Development Theory
Summarising Remarks

Chapter 3: The formal order of the IPDP
Presentation and Representations of the IPDP and Aba'ala
    - Representations of Aba'ala
Self-Presentation and Representations of the IPDP
    - Nursery Activities
    - Soil and Water Conservation
    - Women in Development
    - Water Development
    - Veterinary Services
    - HIV/Aids Prevention Initiatives

Where and What is the Bottom?
    - 'Bottomless Development?'
    - Participatory Empowerment or Cosmetic Representations?
    - The introduction of HIV/Aids Prevention Initiatives
    - Integrating Women in Development?

Summarising Remarks

Chapter 4: Agency and development discourse
Informality Within the Formal Structures
On the Project; Off the Record
    - A Strict Plan, Flexible Implementation: The Case with the Donor Representative
    - 'He knows the NORAD format': The Case with the Consultant
    - The REST Case

Understanding Off the Record Informality
'What is there to Know?'
    - 'Cultures of Formality'
    - The 'Good' Documents of the IPDP?
    - Practical Interrelatedness of Formal Structures and Informal Strategies

Summarising Remarks

Chapter 5: The IPDP: Seeing like Ethiopia?
To See Like a State
    - Elements and Effects of State Intervention
The IPDP: Seeing and Acting Like a State?
    - Two Elements Contested
NORAD Guidelines
    - Representations and 'the Development Gaze'
Pastoralists on the Government's Scene
    - The Ethiopian State's Encounter with the Pastoralists; a Historical Account
    - Land Reforms, Producers Cooperatives and Sedentarisation
    - Sedentarisation Policy in Contemporary Ethiopia

Prescribing Ethiopia
    - SDPRP: A Cacophony of Interests
    - Diverging Results and Effects

Summarising Remarks

Chapter 6: Globalisation and development: homogenisation or hybridisation?
Development: Cultural Flattening or Local Creativity
Development Discourse and Standardisation
    - The Modern Project of Standardised Problems and Solutions
    - Standardised Planning and Implementation

Agency and Hybridisation
    - Reflexive Interface and Reproduction of the Formal Order
    - Creativity and Hybridisation

Dualism, Irrelevant Distance and Hyperspace
    - Local Agency Reproducing Global Discourse
Summarising Remarks

Chapter 7: Concluding remarks
    - Some Summarising Empirical Remarks and Exemplifications
    - Highlighting Some Theoretical Remarks

Unpublished Sources



I wish to acknowledge some of those that have contributed in making this thesis. I am particular grateful to the people associated with the Integrated Pastoral Development Project (IPDP) who not only welcomed and integrated me in their work, but also accepted to be the most important and invaluable part of this thesis; the informants.

Thomas Hylland Eriksen, Department of Social Anthropology at the University of Oslo, has offered valuable help and guidance in the process. As my supervisor, he has given fruitful comments and advices en route.

Axel Borchgrevink at NUPI (Norwegian Institute of International Affairs) has also given valuable inputs and comments to my work. His insight into anthropology and development work has been important.

I am grateful to Yayneshet Tesfay and Diress Alemu at Mekelle University, Ethiopia, for facilitating my fieldwork. The Development Fund, Norway, has also contributed in this matter.

I am grateful to the student grant holders at NUPI for interesting debates and discussions. The librarians at NUPI have been helpful. Additionally, I wish to thank the research project Transnational Flows of Concepts and Substance, at the Department for Anthropology, and particularly the students involved for interesting debates.

Despite the contributions from various people, the presented thesis and the analyses made are solely my work. I am responsible for any errors or misquotations.

Finally, and not least, I'm grateful to Elisabeth for the patience and support she has given me in course of the process.


Jon Harald Sande Lie, Oslo, January, 2004

Chapter 1. Introduction

This thesis is about discourses of development as they are featured in donor-recipient relationships within the development sector. Empirical focus is the development agents and organisations related to a Norwegian funded development project in northern Ethiopia. My argument is that development constitutes a discourse, as lessons from post-structural development critique illustrate, but that there are many discontinuities between the discourse's formal order and the local practices among recipient organisations and the actors involved. The actors involved face the gap between local knowledge and the development discourse. My combination of discursive and actor-orientated approaches to development illustrate that actors can relate reflexively to a discourse, and that post-development theory has severe shortcomings in neglecting agency and seeing development discourse as a hegemonic and homogenising system of knowledge. I argue for the plausibility of identifying a development discourse as a system of knowledge which development agents relate to in their work. As I will show, the development discourse is not irrelevant to what is going on, but it is not what is going on.

The study object is a development project in Aba'ala in the Ethiopian Afar region called the Integrated Pastoral Development Programme (IPDP). The IPDP is funded by NORAD through a Norwegian non-governmental organisation (NGO), the Development Fund (DF). Local implementers are Mekelle University (MU) and the Department of Agriculture (DoA). DF facilitates the partnership with the two Ethiopian organisations. MU and DoA are in partnership with and responsible to DF. DF, additionally, collaborates with and is responsible to NORAD, as the project's back-donor. My primary concern is DF, the Ethiopian partners, these organisations' development agents, and their comprehension of the formal order of development.

In general, the choice of theme and area dates back to 1999, when I conducted my conscientious objector duty at the Development Fund (DF). As a project assistant at DF, I helped facilitating a workshop where selected DF partners participated; amongst others Mekelle University (MU). At this workshop, the president of MU invited me to conduct my fieldwork on one of the projects MU coordinates. After I had started my post-graduate studies, taking courses in general development issues and the anthropology of development, I initiated my fieldwork. From January 2002, I spent six months in Ethiopia, more precisely in Mekelle, where MU as the coordinator of the project is situated, and in Aba'ala, Afar, where the DF funded project I study is implemented.

Anthropologists have not always been concerned with development issues, especially not applied and in practice. A seminar held in Oslo in 1982 addressed the issues of anthropologists' role in development work. The seminar approached the problem of the lack of cooperation between academics and development workers, and argued on the inclusion of anthropology, which despite 'its development-country-profile' so far had been neglected (Melhuus and Klausen, 1983). In 1989, another book (Eriksen, 1989b) addressed the problems development implementers have with including the cultural dimension in their work, in which anthropologists are to be experts. Both of these edited books argue for emphasising anthropology's knowledge and insight about 'the others', the target groups, or the 'underdeveloped' in development assistance. My approach draws more on recent literature known as the anthropology of development (as opposed to development anthropology) and post-development literature, which largely focus on 'us' and the donor side of development. Early post-development scholars see development as a western invented discourse and as a neocolonial project, mainly due to their focus on the donor's formal order and not how this order is received among local organisations and development agents. The combinations of a discursive and actor-orientated approach bridge the traditional anthropological micro-orientated focus with post-development scholars' more macro, textual and discursive approach.

This dissertation results not only from the combination of the above-mentioned circumstances, but also from the experiences and 'counter-tendencies' I've made en route. The thesis is an empirical study of a development project, but also contributes to the theoretical discussion on relations between discourse and agency, or structure and actor. The processes that have led me to this thesis mirror what Wadel describes as 'a round dance' between theory, method and data (Wadel, 1991). In my case, this round dance took place not only prior to and during my fieldwork, but also afterwards; during my empirical analysis and the writing of this thesis.

The problem

What follows in this thesis is a description and analysis of the donor-recipient relationship and its implications connected with a particular development project in Afar, Ethiopia, and the processes of how a particular knowledge is translated as it is transferred from the donor in Norway to recipients in Ethiopia. I illustrate how development cooperation and the implementation of a project function practically, and how and in what way the actual practice relates to the codified formal order of partnership cooperation and project implementation. My concern is the flow of development concepts, policy and ideas (which largely follow funds from donor to recipient) and how this is responded to locally in the encounter with the recipients. The 'development speak', or rhetoric, and policy are characterised by buzzwords, which change regularly, but get high influence as they spread and circulate fast. Among the words that characterise and infiltrate contemporary development speak are 'participation', 'bottom-up', 'community planning', 'empowerment', 'partnership', 'accountability' and 'recipient's responsibility'. These words all have in common that they address the beneficiaries of a project and their role. Thus, they also dismiss the role of the development agents involved who promote these buzzwords, and thus see themselves merely as facilitators for the participants in achieving the stipulated policy. Despite the development rhetoric and partnership idea, the donor-recipient relationship implies the transfer and flow of substances, concepts and resources. The recipient's dependence on project funds from the donor enables the donor, regardless of the formal guidelines and self-imposed rhetoric, to heavily influence the project, its goals and its policy. What is then the partnership relationship in development work about? How does it work? How do rhetoric and practice relate to each other? How are the donor's ideas and policies conceived among implementing development agents? I explore the relationship between donor and recipient, and what it implies for the constitution, realisation and implementation of the project. The partnership relation also denotes the encounter between what can be classified as different systems of knowledge, that is, the encounter between a discursive expert knowledge and local practical knowledge. I question to what extent the discrepancy in terms of knowledge affects the project's formal design and local implementation.

Drawing on lessons from the previous decade of development critique, my initial idea was to study how the development discourse shaped and was articulated in a particular project. Since one of the shortcomings of this literature is the lack of an empirical foundation, it could be interesting to study how a Western hegemonic notion of development was communicated locally and practically. Post-structural development critics largely approach development as identified in the donor countries and in various policy statements. Recipient actors and how they relate to this system are largely neglected. The most radical post-development approaches are associated with Sachs's (1995b) and Escobar's (1995) works, which postulate a total critique of development. They describe development as a uniform practice that during the last 50 years has been manifested into a massive Western hegemonic and formative discourse. They argue that the structures and knowledge that development intervention relies upon are constituted as an objective and neutral field, which give legitimacy to development intervention. These structures and the system of knowledge are manifested and reproduced in development language, practice and institutions (Nustad, 2001b).

Due to my experience in the field, I somehow had to modify my initial assumption about the relevance of the post-structural critique and the validity of a strictly discursive understanding of development. My objection to post-development scholars is their description of development as a uniform practice that is shaped by the discourse in which development agents are embedded. Their description of contemporary development discourse draw mainly on the general conception of development from the post-World War II era, and they have largely disregarded many later approaches to development practice and policy (Nustad, 2001b). Development critique among post-development scholars focuses mainly on the formal and ideational level as seen from the donor side, and consequently neglects local practices and responses.

The modifications I made are based on observations. While I acknowledge the value of seeing development as a discourse on the formal level of the relationship between donor and recipient, I emphasise the importance and value of focusing on individual actors in practical development work. Consequently, a more balanced view of development practice appears. An actor-orientated approach illustrates if and how the development discourse is translated by and articulated among development agents, how these actors relate to the formal order of development, and in what way this development knowledge is applied in constituting the project both practically and formally. To get to grips with this, I apply Norman Long's concept of interface (Long, 1989, 1992c). This implies a shift in analytical focus from the different discourses or systems of knowledge towards the various situations where these meet and become articulated. As I show, this also implies a shift from the conception of formative discourses and structures to how actors relate to them. Situations of interface generate local contextual counter-tendencies, i.e., actors' responses and methods to bridge separate systems of knowledge (Arce and Long, 2000). This will be discussed more fully in chapter three, four and six.

The relation between discourse and agency, or structure and actor, is among the classic problems within anthropology and sociology. The social phenomenological tradition emphasises discourse in the sense that the world is socially and meaningfully constructed, and can be approached hermeneutically. Berger and Luckmann (1992 [1967]) argue on the inter-relatedness of structures and actors. Actor-orientated analyses have problems in grasping the structural factors that shape agency, and structural explanations tend to be functionalist. This impasse is met by postulating that structural phenomena influence people's values and thus their choices, which again have structural and societal implications (cf. Borchgrevink, 1989: 4).

Development's formal order and organisation denote the discourse and structures of development to which actors relate in various ways. In trying to get as complete a picture as possible, one needs to take into account both the structures of development and the actors that act in relation to these structures. Approaching development as a system of knowledge that actors relate to in situations of interface enables one to draw on the insights provided by both post-development and actor-orientated scholars. This two-fold approach underlines the lesson from Barth, who is "…in no way arguing that formal organisation is irrelevant to what is happening - only that formal organisation is not what is happening" (Barth, 1993: 157). Further, Barth argues that one needs to "…trace the contexts into which people through their interpretations embed their acts, since each provides a much used, living tradition of knowledge" (ibid.: 173-174). Development discourse is one of these contexts that development agents face alongside their local 'cultural stock', i.e., the knowledge, concepts and values that actors relate to in shaping their acts and lives. The interrelatedness and situations of interface between development discourse and local practical knowledge among development agents and the implied counter-tendencies are this thesis' core themes.

Some reservations

This dissertation does not intend to give an exhaustive description and analysis of 'development' in general, neither concerning Ethiopia nor those projects funded directly or indirectly from Norway. It is about a particular development project in Ethiopia, funded by NORAD through a Norwegian NGO. Nevertheless, my understanding of the IPDP and the depiction and analysis of the donor-recipient relationship it implies might illustrate some general trends of development and partnership relations. I present selected cases from my fieldwork that draw on and reflect my general understanding and experiences. Despite the frames given and the narrow focus on one particular project, which holds a rather small position within the global discursive order of development, I believe this thesis can illuminate some general aspects about development projects and cooperation in general.

Though related and interesting, some themes fall outside the scope of this thesis. Most prominent, though implicitly and briefly touched upon, is an explicit analysis of power relations - both in terms of potential power structures in the donor-recipient relationship and actual practical power. In this respect, my initial plan, which is abandoned, was to see the donor-recipient relationship in terms of a gift economy. Another aspect falling outside the scope is a direct identification and analysis of various actor networks. An explicit account of both the actors' intentions and the multitude and variety of local cultural 'determinants' that make up their life-worlds fall without my range of study.

I've seen the necessity to protect my informants' name. If not because of any irregularities regarding the project's formal order, so at least due to the agreements I made with my informants in order to get the privilege of having them as informants. No names are mentioned. Apart from the cases of a more sensitive character, I refer to my informants in terms of the position they hold. In some cases, I've seen it necessary to denote actors as part of a larger group to which they belong, e.g. the board, and not their position, e.g. project manager. All organisations are denoted by their original names.

The context and the actors

The Integrated Pastoral Development Programme1 (IPDP) is implemented in Aba'ala wereda2 in zone two of the Afar region. Below, I give a preliminary account of some relevant actors and some general contexts. The presentations are not exhaustive, but more information is provided accumulatively in future chapters. The IPDP is funded from Norway by the Development Fund (DF), and is locally run and implemented by Mekelle University (MU) and the Department of Agriculture (DoA). DoA lies in Aba'ala town, which is the centre for most project activities. MU lies in Mekelle, in the Tigray region, approximately 60 km northwest of Aba'ala. Aba'ala lies at the bottom of the escarpment area that marks the regional border between the lowland of Afar and the Tigrean highland.

The Afar region did not exist prior to the national federative formation in 1995 and Aba'ala3 was until then part of the Tigray region. The establishment of a federative formation based on ethnic boundaries was amongst the most important cases of the revolutionary side in the 1975-1991 revolution. The Ethiopian revolution in 1974 started with the overthrow of Emperor Haile Selassie, led by a group of young radical military leaders who organised their opposition through a military coordination committee known as the Derg.4 The revolution was not only a military revolution. The assumption of power was also due to a popular rising against the absolute and feudal leadership of Haile Selassie and the increasing rate of general poverty. After internal rivalry and disagreement on policy visions within the Derg, Mengistu Haile-Mariam rose as the leader and became head of the Ethiopian republic in 1975. Whereas the government of Emperor Haile Selassie received support from the US, the Soviet Union supported the military dictatorship of Haile-Mariam and the Derg.

The 1974 revolution triggered the establishment of a Tigray-organised opposition. Tigray People's Liberation Front (TPLF) was established in February 1975. TPLF, as a political opposition and guerrilla group, argued that the revolution did not imply any positive change for the Tigrean people. From being a neglected Ethiopian ethnic group during the reign of Haile Selassie, the new government succeeding from the 1974 revolution implied a shift in governmental policy towards increased oppression of the Tigrean people. TPLF argued for each Ethiopian ethnic group's right to self-justice. Together with various other ethnic based groups, most important the Eritrean People's Liberation Front (EPLF), the Tigray initiated uprising ended in 1991 with the overthrow of the Derg. Whereas EPLF pursued their goal and Eritrea became independent from Ethiopia in 1993, TPLF formed an interim coalition government, named the Ethiopian People's Revolutionary Democratic Front (EPRDF), consisting of the various victorious ethnical revolutionary parties. Their main objective of establishing a federative governmental structure based on regions demarcated by ethnic criteria was implemented in 1995 as the Federal Democratic Republic of Ethiopia (FDRE) was founded. The government has since 1991 been led by EPRDF whereof TPLF is the most influential part, and both the positions as prime minister and president are held by Tigreans.

The Development Fund (DF) has been engaged in Tigray since 1982, when they supported the Relief Society of Tigray (REST), known as TPLF's humanitarian wing, and thus indirectly supported the Tigrean revolution and guerrilla activities. DF's food and aid assistance, supported and funded by the Norwegian Ministry of Foreign Affairs, along with several other donors, was paramount to the general population of Tigray, but also meant an indirect economic and political support to the revolutionary activities. This was particular to DF since it in Ethiopia only was engaged in Tigray.

After 1991 and the end of the revolutionary uprising, DF maintained its support to REST, which now formally was independent from TPLF. Despite this, REST was still seen as a political actor: It was only working in Tigray and there were widely overlapping in terms of interests, policy and people between REST and TPLF. Today's leader of REST is a central party committee member of TPLF. From the late 1980s and until the mid-90s, DF's support to an ethnic guerrilla group and later the indirect support to the ruling party and group in Ethiopia became a problem and a weight to bear for DF. Thus, and in order to position its political neutrality by working with others than those associated with the ruling elite, DF decided to start work outside Tigray region. After assessments of area and potential partners, the Integrated Pastoral Development Programme (IPDP) was initiated in 1998 in Aba'ala Wereda in Afar region. Partners are the local Department of Agriculture (DoA) and Mekelle University (MU) from the capital of Tigray, Mekelle. As argued below, MU's practical involvement in IPDP is crucial. The president of MU, who also is the IPDP steering committee leader, is a member of TPLF's politbureau.

Regarding the initialisation of the IPDP, its first programme manager said in my first meeting with him that he still remembers a DF representative in 1998 coming 'running across the border from Eritrea with money'. That year the Eritrean government threw out largely all NGOs working in Eritrea. Thus, DF had money left on its budget adequate to initiate on a new project outside the Tigray region.

In general, Tigray region has always been rather marginal in the Ethiopian context, but after the 1991 revolution Tigray has increased its national position, largely due to positions Tigrean leaders, as representatives for the victorious revolutionary side, acquired when establishing a new government. This also enabled an increase in the channelling of resources to the region. It is illustrated in the growth in Mekelle's population, which in 1982 was approximately 20.000, and in 2003 had grown to 128.000. This growth exceeds what is natural regarding the regular Ethiopian urbanisation. Whereas the former great inequality of resources between Mekelle and Addis Ababa has decreased and Mekelle starts to become a centre by itself, a similar schism of resources is found today in the relation between Tigray and Afar. The position and role of Mekelle University vis-à-vis the Department of Agriculture serve to illustrate the contemporary division between Tigray and Afar.

Afar region and Aba'ala

In mid-1999, the estimated population of Afar was 1.188 million, of which 27.259 were found in Aba'ala wereda (Alemu, Farah and Mbuvi, 1998). Aba'ala town has 3.300 inhabitants (Kelemework, 2000). Different tabias are scattered in the eastern direction of Aba'ala town, of some are in the IPDP's target group. Previously, the areas in Aba'ala wereda were used for grazing by Afar pastoralists, but in the late 1960s Ras Mengesha Seyoum5 started commercial agriculture by clearing the wooden bush lands in the flooded areas. Since then, people, mainly from the highlands, have settled to cultivate the flooded areas (ibid.). The changes taking place made not only people from the highland settle there, but also attracted some pastoral people. Today, the majority of the people in Aba'ala town are Tigreans. Some remote tabias also have Tigrean majorities.

The classic stereotypical socio-cultural and ethnical distinctions between the Afars and the Tigreans are that while the former group is characterised as Muslims, nomadic and pastoralists, the latter group is characterised as orthodox Christians, sedentary and farmers. Some Afars living in central areas combine pastoralism and farming, and are thus denoted as agro-pastoralists. These stereotypes are merely stereotypes. In chapter three and five, I show how development agents' static and homogeneous conceptions of the ethnical characteristics and differences are important in the conceptualisation, planning and design of the project.

In general, pastoralism and farming do not represent polar opposites, but rather ideal types of economic activities along a continuum from 'pure' pastoralism to farming. Hogg (1997b) argues that most of Ethiopia's pastoral societies pursue multi-resource economies in which the balance between pastoral and non-pastoral activities is constantly shifting in response to the circumstances. Pastoralism is thus not a way of life but a set of specialised economic activities and techniques revolving around the herding and care of livestock. Pastoral communities adapt to their changing natural environment (ibid.: 2). Many pastoral communities were first informally incorporated into the Ethiopian polity during the last century, and formally only since 1995 with the establishing of the Afar regional state (Hogg, 1997a; Said, 1998; Getachew, 2001).

Aba'ala village is the hub for governmental administration of Afar zone two. It is also the centre of most of the IPDP's project activities for the various target groups living scattered around the village. The village has experienced a rapid growth and development. In 1999 electricity and tap-water were introduced to selected parts of the town. A new and passable road connects Aba'ala to the Addis Ababa - Mekelle/ Asmara road. Thus, the traditional Thursday marked, which previously only attracted caravans and people from adjacent areas, now attracts merchantmen from the highland who take advantage of the Aba'ala market, which is cheaper than markets in the highland. This exposes the Afars to external actors.

NORAD and Norwegian development assistance

The Norwegian Agency for Development Cooperation, NORAD, is an implementing agency under the Norwegian Ministry for Foreign Affairs. In the formulation of development assistance the Ministry makes policy decisions while NORAD works out the rules and regulations for implementation. Unless otherwise specified, NORAD's and the Norwegian Government's policy can therefore be taken to be the same (Saugestad, 2001). Though financing "…initiatives and efforts prioritised by the development countries themselves, NORAD invests in human rights, democracy, environment, economic growth, education, health, welfare and equality".6 Aiming at policy coherence, NORAD distributes funds for development assistance in three channels. These channels are bilateral, that is, directly state-to-state support, multilateral, i.e., assistance given to and directed by various international organisations, and through civil society and NGOs. My primary concern is the latter channel. NORAD states that civil society is "…the formal and informal networks and organisations which operate and are found in the space between the state, the family and market and Norwegian NGOs are encouraged to support this sector of society" (NORAD, 2003b: i). Not only does NORAD promote policy coherence among its three channels, it also strives to cohere with international organisations and treaties regarding development in which Norway is engaged. I address the complex issue of policy coherence in several chapters, and most notably in chapter five and six. NORAD provides Norwegian applicant-NGOs with 90% of their funds based on the applicants' 10% share, which the applicants must collect from private donors to illustrate their role as representatives of civil society, that they are non-governmental and rely upon popular involvement. An implication is NORAD's power to delineate and affect applicants' policy, despite NORAD only is supposed to support 'initiatives and efforts prioritised by the development countries themselves'. The issue of participatory approaches in terms of policy choice, project planning and design is among my central concerns, in addition to the ambiguity between development rhetoric and practice. A central ambiguity of development rhetoric is between the widely acknowledged ideas of participation and policy coherence.

The Development Fund

The Development Fund (DF) is a Norwegian NGO and holds the role as donor to the IPDP. In 2001 DF received 31.1 million NOK from NORAD, which places DF as the eleventh largest recipient of NORAD funds allocated to Norwegian voluntary organisations (Liland and Kjerland, 2003: 250). DF states that since its start-up in 1978 it " …has been in the vanguard of progressive thinking, with was then the radical vision that poverty is best fought by aiming to enable people to help themselves",7 an idea that today is widely, if not universally, acknowledged within the development sector. DF's primary goal is to contribute to combat poverty and give support to marginalised groups in rural areas in selected countries. DF's three priority areas are food security, productive efforts and civil society, all based on the condition of environmental security and sustainable resource use. Small farmers are DF's primary target group. DF supports projects in a variety of countries in Asia, Latin America and Africa, hereof Ethiopia and the IPDP. DF works only through partner organisations in the respective target areas, and has no personnel stationed abroad. DF's partners in Ethiopia are Relief Society of Tigray (REST), Women Association of Tigray (WAT)8 and Mekelle University (MU), all situated in Tigray. Regarding the IPDP, DF collaborates with MU and the Department of Agriculture (DoA). DoA lies in Afar. My general experience from these partners is that DF is reckoned as a highly appreciated, good, stable, engaged and serious partner.

Mekelle University

Initially established as a College of Dryland Agriculture Resources in 1993, this academic institution gained status as University in 1997. The activities of Mekelle University (MU) largely rely upon business assets or funds from external donors. NORAD is among several governmental organisations that support MU. Additionally MU collaborates in, implements and coordinates a number of development project funded by various NGOs, including DF and the IPDP.

MU plays the key role in the IPDP despite formally being equal with DoA. The formal role of MU in the IPDP is largely one of coordination, technical back-up and provision of technical personnel exceeding the local staff's abilities. MU is responsible of providing DF with financial mid-term and annual reports.9 The IPDP project manager, who is an employee at MU, performs all direct communication with DF. MU controls the accountings of the project and thus makes the disbursements of salaries and other costs related to the project. MU holds the project car which is used by the manager to travel between Mekelle and Aba'ala. MU holds the most central and influential position of the IPDP board and steering committee, i.e., the chairman.10 The board rarely meets, because of difficulty of communication, restricted time-budgets and the spatial scattering of board members. Practically, MU's role and influence over IPDP exceed the formal power and responsibility initially assigned to MU, since MU holds four board members (including the chairman and the project manager), while also managing the accountings and the communication with DF. Many decisions are taken by MU representatives without conferring with other board members or their seemingly equal partner, DoA, despite "Zone 2 Department of Agriculture of the Afar Regional [sic] is the major partner of the programme".11

The Department of Agriculture

As the major 'partner of the IPDP' "[a]ll programme activities are planned, implemented and closely supervised by the staff of the Department of Agriculture at zone and Wereda levels. The department is also responsible for technical support of the project. The project's site manager is seconded from the Department".12 Other local offices are also engaged in IPDP under the supervision of DoA: Zone and Wereda Administrative Council, Zone 2 Women's Affairs Office and Wereda Health Office. These institutions collaborate with DoA in the practical implementation of selected project activities ascribed to their respective area of responsibility in Aba'ala.

The Department of Agriculture (DoA) holds office in the outskirts of Aba'ala town among a cluster of governmental and regional administrative offices. DoA has difficulties in implementing its formally ascribed tasks. DoA staff often complain about their subordinate position in the IPDP, but also acknowledge their constraints in terms of lack of knowledge on project management, their high illiteracy rate, and thus their difficulties in communicating directly with DF. In addition is the turnover rate of qualified personnel high among DoA staff, who rather want to go back to the urban highland where they were educated to work. The impression given is often that they feel as recipients towards MU, while they see DF and DF's project coordinator, who visits Aba'ala a maximum of three times a year, as their friends and assistants in their problems with MU. The difficulties DoA faces in executing its formal obligations are not purely due to internal factors, but also to the position MU has in the IPDP and towards DF, which, as argued above, exceeds its formal obligations.

In practice the DoA functions to bridge the gap between MU, the project activities and the project's beneficiaries. The people of Aba'ala are, as Afars in general, sceptical of external actors (as MU and DF) and change. The DoA's role is crucial in practical implementation regarding the communication of the project to the beneficiaries, to get local acceptance for the project and its various activities, and the involvement of external actors.

I aim to study development agents and their relations to both the formal order of development, to the donor and the implications of this relationship in project implementation. Therefore, my main study object becomes the IPDP project staff at MU, because they hold key positions in the project in terms of implementation, planning, communication, and general project management. Nevertheless, I do draw on some material from DoA and its relationship with MU, DF, and the project in general. Some other organisations are also relevant to my study. They will be presented as they emerge in this thesis. The above-mentioned organisations are the most important ones for the IPDP and my study, and thus represent my main frames for gathering data.

Methodological considerations

Development in general is a huge industry with a multitude of actors and organisations involved on various levels. I will limit my study area to the IPDP and the various actors involved in the project on various levels. The organisations mentioned above, the actors they employ and the various IPDP project documents demarcate the field. The decrease of scale by focusing primarily on the IPDP also reduces the complexity of my field. The concept of scale implies a connection to something larger, and my local empirical data thus say something about the larger structures and discourse of development in which my field is embedded. The IPDP design is not merely a result of the relations between donor, recipients and beneficiaries, but draws in many ways on a global development discourse. The development discourse is depicted in the next chapter. Occasionally I also draw on material not directly aligned with the IPDP. Most of this material relates to DF and its other partners in Ethiopia. I use written sources from the IPDP, but also other written documents and literature to illustrate the broader system IPDP is a part of, such as the Ethiopian PRSP13 process, the role of the Development Association Committee (DAC) in stipulating policy, and NORAD. I've also had three meetings with representatives of the Norwegian embassy in Addis Ababa. It all serves to give a broader picture of DF, the donor-recipient relationship, and features common to the IPDP which all contribute to illustrate the larger system of development which the IPDP is part of.

I've used several methods to collect various types of data in order to understand if, how and to what extent development agents on various levels are embedded in and how they relate to various social practices, organisations and systems of knowledge. The numerous collected data and the methods used to gain them are characterised by what Denzin calls data and methodological triangulation (Denzin, 1989). The concept of triangulation denotes the combination of multiple strategies and methodologies in the study of the same phenomenon with the purpose of providing the researcher with more reliable and unbiased data. Denzin states that triangulation can occur on four basic levels, i.e., on data, the investigator, the theory and the methodology. I have applied data and methodological triangulation. I was unable to apply investigator triangulation, since this is an independent study and product. The various approaches are more a result of a round dance between theories, methodology and data (Wadel, 1991) than something that I strove for in advance of the fieldwork. As Denzin writes, there's no magic in triangulation. It only makes the researcher aware that different approaches yield different pictures and slices of reality. Triangulation is a way for anthropologists to cope with the problems of subjectivity, biases and self-reflection.

Studies of development projects and organisations from within are scarce. This thesis is, however, such a study and was rendered possible as I, in 1999, when working in the Development Fund (DF), was invited by the president of Mekelle University (MU) to do my field study on one of the projects MU facilitates. Before leaving for Ethiopia, I requested and received a letter of recommendation from MU that said I was invited for six months to conduct my field study. I got admittance and was integrated into the IPDP due to my initial invitation and because I was, in the beginning, still associated with the Development Fund (DF), since one of DF's employees helped me in establishing contacts. I explicitly stated that this was my own academic project, independent of DF and NORAD. Consequently, some high officials and the administrative staff of MU became sceptical to my presence, and asked 'what I had to offer them'. After explaining about my project and referring to the invitation letter, I was again admitted. Later, these initial problems were paid no attention to and I was included in the parts of the MU staff working on development projects, and especially those working with IPDP, DF and NORAD. I many cases, I was also seen as a resource, as I was regularly called into the office of a senior staff member at MU. He would ask me to inform him about e.g. NORAD's role in IPDP, how to apply for a PhD through NORAD's quota programme, to update his anti-virus software (437 new viruses found!), or to teach him to pronounce the name of an employee at the Norwegian embassy who he was supposed to address in a meeting, a name composed of three of the Norwegian letter Ø.

The initial formal problems were disregarded by those who later were to become my key-informants. These are the IPDP project manager, two IPDP board members and the first IPDP coordinator, who all are employees at MU. Other informants are the general staff, board members and the network associated with the IPDP in particular. Another complementing group of informants are those working in MU and other organisations on development projects, but not the IPDP. The IPDP project manager and one of the board members, who previously lived in Aba'ala but moved back to Mekelle as the project he was running was to be phased out, are the most important persons in the practical implementation of the IPDP. They both seemed genuinely to appreciate my presence and my interest in the IPDP and their work. To them, I also was a person with knowledge about the Norwegian model and NORAD, which they had little experience with. They were also of key importance to help me with transport between Aba'ala and Mekelle. I was allowed to join in on their trips to Aba'ala, and as they went there about once a weak each, I regularly had the opportunity to visit the project area and observe how they worked and related to the beneficiaries and the Department of Agriculture (DoA). I thus became familiar with some of DoA's staff and the local field manager. All together, I stayed approximately 65 days in Aba'ala. The 25 nights I spent there were largely because I conducted a household survey. The many trips back and forth to Aba'ala, a one-way drive estimated to around 1.5 hours, were an imperative source of information, as I had the chance to spend time alone with my informants. The car became not only a means of transportation, but also a mode of inquiries and a marker of "conversational communities" (cf. Gudeman and Rivera, 1990). Nevertheless, my base was Mekelle, as almost all my informants lived there. My key informants were also crucial in providing me with various project documentation and background information of the IPDP.

All my informants, except those living in Afar, spoke English. English is also used as the 'project language' and to communicate with the donor. Consequently, both my informants and I used our second language. Obviously, this implies various constraints regarding the actors' translation from Tigrinya to English, and my translation from Norwegian to English. Nevertheless, the possibility of speaking English was crucial for my fieldwork, as it would have been difficult and time-consuming for me to obtain sufficient skills in Tigrinya or Amharic. Despite the problems of rationality and translation, which increase when using a second language, English was the natural choice as it was our common denominator in terms of language and thus had fewest negative implications regarding translation. If needed, my informants assisted me in translating from the local language. Whenever I was present, the IPDP staff generally talked English.

Various types of data

My collected data can be divided into qualitative and quantitative data. The qualitative data is based on observations of the general project work which I occasionally participated in, different forms of interviews (structured, semi-structured and unstructured), narratives, discussions and through ordinary conversations with my informants. Of a qualitative matter are also the various written sources and documents I've used. The qualitative data can be divided into formal and informal, depending on how and where they are acquired. Interviews with a tape recorder present, observations of formal meetings and workshops, and documents are of a formal manner. Informal qualitative data was collected in interviews without a tape recorder, in daily speech, regular observations in various settings, observations made as I participated in project work, and in discussions with the involved actors. There are no clear-cut boundaries between formal and informal data. I frequently confronted my informants in informal settings with data obtained in formal settings. I early abandoned formal interviews with a structured questionnaire and a tape-recorder, since my initial assumptions were not reflected in what the informants wanted to talk about; and rather I decided to follow the loops and let my informants elaborate freely when collecting data. Having the privilege to follow the project manager in largely all his tasks that concerned the IPDP, I received access to interactional data with and between MU employees engaged in the IPDP, the DoA, the beneficiaries, DF (which during my fieldwork visited its partners in Mekelle and Aba'ala), and the beneficiaries. I mostly observed, but sporadically I was also assigned to participate in project management, a workshop, and planning.

Various project documents, (i.e. strategy plans, reports, applications, terms of reference, communication, partnership agreements) are also an important intake for information about the project, since they constitute the formal order of the project and the partnership agreement. Project documents stipulate what has been done, what is to be done, and how to do it. Project documents constitute the formal order of development as a social discourse that different development agents produce and relate to in their project implementation. The formal order of the IPDP, which I present in chapter three, is important since formal institutions form a context that appears as an unproblematic truth underlying development agents' actions. The IPDP's formal order is codified in project documents. Eduardo Archetti, on the importance of also including texts in the study, argues that "…social discourses are also embedded in, or expressed through, writing" and states that one must identify how texts are produced and consumed (Archetti, 1994: 11). Since project documents stipulate the IPDP's formal order, it easily enables to distinguish "…between what people say they do, what they ought to do, and what they in fact do" (Hendry and Watson, 2001: 4).

My quantitative data are more questionable. I conducted a household survey of 58 samples with the objective of identifying the socio-cultural changes that might have taken place since the implementation of the IPDP. In retrospect, I see severe shortcomings in my survey, primarily due to my lack of knowledge about how to make it. Yet, the survey shows some general trends and provides some useful statistical material. The main profits of the household survey are that it can be seen as a personal exercise in how to make and collect this type of material, as well as giving me access to various people, places and stories. I employed a translator when conducting the household survey.

Preliminary theoretical overview

My general theoretical approximation lies at the juncture between a discursive and actor-orientated approach. More precisely, it addresses the connections between development discourse and development agents. The development discourse is depicted in chapter two, the formal order of the IPDP is presented in chapter three, while in chapter four and six I elaborate around development agents' relationship to this discourse. Below, a general theoretical account is given of the concepts of discourse and actor-orientated approach.

The concept of discourse, as applied within the social sciences, denotes the interrelation between knowledge, meaning and power, i.e., a system of knowledge or meaning that is shared by various people (Svarstad, 2001: 3). A combination of a discursive and actor-orientated approach allows us to identify various actors' relations to a system of knowledge, illustrating how actors might draw on, challenge or alternate between different discourses. Imposed discourses might be challenged by the actors' cultural stock and local practical knowledge. Discourses, as the implicit, obvious and unspoken conditions for communications can be challenged by actors' agency and their opinions, i.e., what is intentional, explicit and debatable. Focusing on actors and their relations to the development discourse, one observes interplay between different systems of knowledge. Norman Long's concept of interface denotes the critical point of intersection between different systems of knowledge. Situations of interface are articulated through actors (Long, 1989; Long and Long, 1992).

Discourse as a system of knowledge

Neumann defines discourse as a

"…system for the formation of statements and practises, that by inscribing itself into institutions and appearing as more or less normal, constitutes reality for its bearers and has a certain degree of regularity in an array of relationships" (Neumann, 2001b: 18, my translation).

The arrays of relationships that are of my concern are generally those aligned to contemporary and historical development issues, that is, the development discourse's formation, and particularly how it is reflected in the IPDP and its implied organisations. The demarcation of a discourse implies identifying the regular and systematic collection of statements and practices (Hammer, 2001: 8). My field is demarcated as the development sector, and more precisely the IPDP and adjacent elements. An institution is a symbol-based program that regulates social interaction. The institutionalisation of a discourse implies the formalisation of statements and practices, through rules of formation, which bearers of that particular discourse both represent and reproduce through their agency. A discourse refers not only to oral and written statements, but also to aggregates of social practices (Kårhus, 1992; Kårhus 2001).

Foucault proposes a discourse that is perceived as insignificant by its bearers, and emphasises the structuralising power a particular discourse has over its bearers through the discourse's conditions of existence, rules of formation and procedures of exclusion. "The discourse can appear as insignificant, but the prohibitions it is affected by reveal quite early and quite fast its connections with the desire and the power" (Foucault, 1999 [1970]: 9, my translation). The rules of formation lead to regularity in statements and practices aligned with the discourse. Actors' expressions that do not reflect or relate to the existing discursive order are sanctioned by exclusion. The actors' self-disciplinarian and self-regulating normalisation of statements and practices lead to a strengthening and reproduction of the established discursive order.

Discourse analysis takes as its primary concern to understand the processes of discursive formation. Foucault's method in revealing and exploring discourse is to identify the discourse's archaeology and genealogy. The genealogical approach identifies the history of the discourse and enables the researcher to see the present discourse in terms of the past to discover the discourse's historical conditions of validity. The archaeological approach explores how present discourses regulate and distribute its bearers' statements and practices through certain rules of formation. Foucault's main occupation is to identify the historical and contemporary discursive conditions, and how they are articulated and manifested. Foucault is not directly interested in the discourse's originator, since a discourse "…comprises a sort of anonymous system that is available to those who want or can operate it without its meaning or validity necessarily being connected to the discourse's originator" (ibid.: 19, my translation). My conception of discourse is not as exclusively reserved for selected groups, but as a system of knowledge that can be shared, learned and applied by others.

For those embedded in a discourse the discourse is the reality. For the discourse analyst this reality is perceived through its representations (Neumann, 2001b), or metonyms (Kårhus, 1992). A representation, or metonym, is a piece that stands for a larger whole in which the piece itself is a part of (ibid.: 113) and appears between the physically given world and our perception of it (Neumann, 2001b: 33). Discourse analysts are concerned with epistemological questions, i.e., how and why things appear as they do and how we have a particular knowledge about the world we live in. Identifying the representations which a discourse relies upon say something about the particular discursive realm and how the discourse functions. In chapter three I present and challenge different representations of the IPDP as they appear in the project documents.

Some critical remarks of a discursive approach

Many academics engaged in development issues have during the last two decades been influenced by and largely adopted Foucault's notion of discourse. Known as post-development theoreticians, these scholars' development critique takes advantage of seeing development as a discourse - "…as a system of knowledge, technologies, practices and power relationships that serve to order and regulate the objects of development" (Lewis et al., 2003: 545). This view is associated with, amongst others, Sachs (1992b), Ferguson (1994) and Escobar (1995). My general concern of discourse analysis, which echoes my critique of post-development, is the validity and the area of application ascribed to it and its ability to grasp the entirety and complexity of what is analysed. It's hardly a novel anthropological insight, but it echoes my argument that no single theoretical approach manages to grasp the full complexity of what is described. Discourse analysts in general, and post-development scholars in particular, largely avoid other theoretical approaches. I am critical of their conception of discourse as a monolithic, hegemonic and homogenising system of knowledge that neglects and undermines humans as reflective individuals and rather see them as subordinate and merely bearers of a discourse. Discourses, regarded as obvious conditions for communication, can be questioned under particular circumstances; they can be revealed to be constructions and therefore changeable. Not only conflicts between discourses, but also challenges to the very discourses themselves can occur as the implicit, may be transformed and thus appear as explicit and intentional opinions. I see discourse as a system of knowledge that is maintained and spread by its bearers, but the reception of this discourse among those encountering it has no a priori defined outcome. No discourse is hegemonic, and what happens in the encounter between a discourse and other systems of knowledge is an empirical question. Too rigid a conception of discourse necessarily closes out certain ways of thinking and viewing the world, while privileging others (cf. Lewis et al., 2003). As knowledge is distributed, it is also contextualised. The development discourse, as analysed from the donors' side, is not necessarily what happens locally among recipients. Transformations and translations occur as the realms of donor and recipient encounter. I focus on actors as bearers and articulators of knowledge. This enables me to identify the processes that take place in the knowledge encounter. In my case, the development discourse represents a system of knowledge development agents in various ways relate to in constituting their reality.

I acknowledge post-development scholars' identification of a development discourse, but I disagree on the exclusive formative power ascribed to the discourse. To assume that the formal order of development, as codified in project and policy documents (which is the primary concern of post-development scholars), is identical to the local implementation and practice is not correct. It attaches too much faith to formative structures and knowledge. This is among the central points in this thesis. To study the reception and the local application of the development discourse, an actor orientation is necessary. What becomes prevalent is the disjuncture and slippage between the formal discursive order and local practices.

Actor orientation and informal practices

The inclusion of an actor-orientated approach to development issues emerged because my initial theoretical and methodological approaches, largely affected by post-development theories, did not resonate with my empirical findings, nor with my post-field review and analysis. I acknowledge post-development scholars' identification of the development discourse, but I see this discourse as merely one amongst many systems of knowledge in which development agents relate to, form and are formed by. There are many discourses, or systems of knowledge. Some coexist, some overlap, some oppose each other. By including an actor-orientated approach, it becomes possible to see how various systems of knowledge are affected when challenged and encountered by others. These situations of interface are articulated via various actors.

Traditional actor analysis gives primary attention to the involved actors' intentions, motivation and to some extent see individual as purely homo economicus who pursue their own goals without regard to these being of an egoistic or altruistic kind (cf. Barth, 1993; Long and Long, 1992). I am not directly concerned with identifying various actors' intentions. My actor orientation is more a matter of where focus is put to gain data and analyse how a development discourse is received, applied, translated or rejected, since a discourse necessarily needs to be articulated through someone or something. What becomes prevalent is the difference between formal structures and informal practices, or the discontinuity between formal discursive order and local practical knowledge.

The actor-orientated approach not only helps to open up black boxes of formal, institutional and discursive developments, but also opens up and nuances post-development theoreticians' depiction of the development discourse. An actor-orientated approach offers a possibility to understand how meanings associated with development are "…produced, contested and reworked in practice - and thus to illuminate the multiple significances that the term holds for actors involved in the development process" (Lewis et al., 2003). In order to understand these processes, it is important to understand the broader picture of development, a picture offered by post-development scholars, and how the various organisations involved function formally and practically. The understanding of development as a discourse relies mostly on formal sources. What is regularly prominent in this field, is the discontinuity between formal organisation and the many informal practices that oppose but at the same time relate to the formal structure of development. An actor-orientated approach illustrates the slippage between the formal order and organisation of development, and the local informal practices that result as coping mechanisms towards the imposed formal order. I aim to provide an ethnographic study on how particular texts are produced and consumed by development organisations and agents, how they relate to or feed into a development discourse and how these influence and interact with project practices as communicated by local development agents. In many respects, this mirrors the classic question on the relationship between the map and the terrain. This much said, we are now ready to embark on the analysis proper, after first briefly outlining the thesis.

Brief outline of thesis

In next chapter, I shall present the development discourse and give an account of some main theoretical approaches to development. The chapter is not merely of theoretical value. It also illustrates the context of my general approach, and shows the plausibility of having a discursive approach to development. Thus, the chapter also has empirical value.

Chapter three shows the formal order of the IPDP and the representations that the project relies upon. The formal order of the IPDP not only stipulates the project activities, but also how to design, plan and implement the project. The socio-cultural descriptions of Aba'ala and Afar in the project documents, which formally are the knowledge the project is based on, are also challenged. Project documents present the project's components, target group and area simplistically and as legible units constituting the project.

Chapter four presents three different cases which all serve to illustrate various informal strategies arising as local counter-tendencies in the encounter between the formal structures of development and local practical knowledge. The chapter shows that what is perceived as formal and informal are interrelated, and that informality and reflexivity towards the development discourse can be as much an attribute of local development agents as of the donor representatives.

In chapter five, I give an account of the process of planning and the effects produced. In giving an account of Ethiopia's government's approach to pastoral people, I show how state and NGO intervention in many respects rely on comparable elements and produce similar effects.

In chapter six, I question what the situations of interface lead to. Do the many encounters between donors' and recipients' knowledge represent a process of homogenisation or local creativity? Drawing on previously presented material, this question is debated.

Chapter 2. Development conceptualised

This chapter presents the theoretical framework of the thesis. It also argues for the plausibility of approaching development as a discourse. My general view is on 'those who are to do the development, and not those supposed to be developed'.14 Consequently it is necessary to depict the general system of knowledge that development agents relate to, and the history of this discourse. In retrospect, the last 50 years are characterised by a rhetoric that has altered between a wide range of approaches, methods and policies to development, which all have substituted the former either due to lack of results or because of political alterations. Post-development scholars, in seeing development as a discourse, argue that these new ideas never managed to free themselves from the established development practice and that despite the rhetorical changes the old practices and approaches are reproduced.

Post-development, which dismisses the idea of development, represents a post-structural discursive approach to development and seeks to explain why so many development projects seem to fail by focusing on the underlying premises of development and the unintended side-effects it produces. According to post-developers, the development discourse represents a monolithic, hegemonic and homogenising system of knowledge with a high degree of formative power that prevails over actors involved. The discourse's structures of power captivate the agency of the actors that relate to the discourse. Since no approach manages to explain a phenomenon in its entirety, and post-development largely neglects human agency, an actor-orientated approach is included and combined with a post-structural discursive approach in trying to give a broader account of how development functions practically. This is done through Norman Long's conception of interface, which has a primary focus on processes occurring in the encounter between different types of knowledge, e.g. a development discourse and local practical knowledge. The concept of interface is fruitful for envisioning the interplay between structure and agency because it acknowledges the "…notion of multiple realities and arenas of struggle where different life-world and discourses meet" (Long, 1992b: 271). Initially, development discourse is presented since it constitutes one of the systems of knowledge to which development agents relate.

The development of "development"

Post-development scholars usually argue that the concept of 'development' as we know it today got manifested in the last part of the 1940s, i.e., after World War II.15 The US was in rapid progress, and Europe, which was in ruins, received help from the US to rebuild itself. This was manifested through the Marshall help. The 'Iron Curtain' that divided the former allied parts into a 'democratic west' and 'communistic east' triggered the idea and need of aid and intentional development as a means to secure the US's, and later the West's geopolitical interests by building up partners and future allies initially in Europe and later around the world. Aid and economic support were a means to establish and secure political interests. Development assistance was to become an important tool in the superpowers' ideological and geopolitical struggle.16 The initiation of development assistance and aid was not merely a moral concern of poverty alleviation in general, but also a means to spread scientific advances and technical progress to make underdeveloped areas grow economically so that, from the point of view of the US, the underdeveloped areas would not fall under communistic influence. This was manifested through the inauguration speech of US President Truman in 1949.17 The Truman Doctrine's Point Four created master metaphors within the development sector (Porter, 1995). The division of the world into a democratic west and communistic east was now supplemented with another axis that separated the developed north from the underdeveloped south. When tracing the genealogy of development discourse, post-development scholars ascribe its origin to the post-WW2 era and Truman's speech. Nevertheless, the concept of development as we know it today dates further back.

Evolvement of positivistic ideas about progress and development

The manifestation of the idea of development in the post-WW2 era drew largely on already exiting notions of development and progress. In the 17th and 18th century, questions related to development rose. The term development, which originates from biology, was conceived in the social sphere as "…the transformation that moves towards an ever more perfect form" (Esteva, 1992: 8). The era of industrialisation led not only people to experience rapid growth and progress but also involved an issuant division between rich and poor and the rise of a class-divided society. This largely occurred in the urban areas and exposed poor and rich groups to each other, and poverty became defined as a problem for society in general. In combination with the Era of Enlightenment's belief in science and rationality, this led to the rise of theories and methods to intervene in the society to help those perceived as poor (cf. Nustad, 2003a).

In Doctrines of Development (1996), Cowen and Shenton express that there are two different ways of perceiving the concept of development, that is, development as a latent and immanent process, and development as an intentional directed process, both of which date back to the Enlightenment Era. Development as a latent and immanent process is associated with Adam Smith, who argued that in order to disengage the immanent processes of development, the social and structural constraints preventing this had to be removed. The understanding of development as an intentional process differs from Smith's conception, in terms of how humans are understood. Whereas Smith argued that the aggregate of individuals' free moral choices would benefit society, those seeing development as an intentional process argued that this process needed to be directed and guided by someone, i.e., society's managers in accordance to certain stipulated intentions. The latter view, often associated with the Saint-Simonians, rejected the idea of progress as a natural non-intended process, and rather proposed an idea about development as an active interference in society by its managers (Cowen and Shenton, 1996; Nustad, 2001a). August Comte brought these ideas further in his promotion of rationality, planning and science.

Nustad argues that the conception of development as a process leading towards an ever more perfect form implies that someone necessarily needs to know about this form. These are also to intervene and intentionally direct society's development. A consequence is that despite certain groups (the poor and the rich) are living in the same spatial and temporal society, those defined as objects for development by society's managers are conceived as living in another time, since they do not have the same conditions of living as the managers. Nustad calls this phenomenon temporal segregation - that inequality in standard of living is explained by stating that the poor are at another evolutionist stage of development or history than the rich. Temporal segregation denotes a unilinear perception of history, where societies, or groups within a society, are placed along a two-dimensional axis leading from 'not-developed' to 'developed'. This implies that poverty is not explained with reference to contemporary political relations, but by stating that societies exist isolated from each other at different developmental stages (Nustad, 2003a).

The Era of Enlightenment also represented a fracture between a sacred and secular perception of time and history, which underlines the concept of temporal segregation. The rejection of God's omnipresence turned into great emphasis put on man's contribution to history and society. This is illustrated with the concept of travelling. What previously had taken the form of pilgrimage was now conceived as a refinement-travel for the bourgeoisie of the 17th century, which represented a journey in time and space (ibid.: 31). In 1800, J.M. Degèrando wrote that "the philosophical traveller, sailing to the ends of the earth, is in fact travelling in time; he is exploring the past; every step he takes is the passage of an age" (cited in Nustad, 2003a: 31). What constitutes the notion of temporal segregation went from being a characteristic of internal inequalities to become projected on external and remote societies and countries. The increase in scale also implied an increase in effect. Relations between rich and poor people, and later developed and underdeveloped countries were understood as temporal relations, and underlines what Nustad calls temporal segregation.

The ideas on how to cope with and solve the internal inequalities of a society were also later to be projected externally on other countries and societies. Poverty and inequalities were initially seen as a local and national problem of European capitalistic societies, but when European nations started to explore and engage in colonialism outside Europe, the problems of poverty and inequality were identified there. The conception of European inequality now became the characteristic of other remote and foreign societies, seen in relation to Europe. Not only were the same characteristics adopted, but also the same means to cope with and solve the problems in the colonies were adopted. The division between rich and poor in various European societies was now applied as a distinction between Europe and the colonies. Helping these poor nations, among other things, became a legitimisation of the colonial powers' presence. The understanding of development as possible to intentionally direct through intervention was applied. The same notion was later largely pursued by development organisations in the post-colonial era. This conception of development is reflected in post-WW2 development policy and strategies.

Two grand theories on development

Modernisation theory

It is argued that the 'invention' of underdevelopment, the conceptualisation of somebody as 'underdeveloped', and thus the Project of Development was initiated by US President Truman's speech of inauguration, 20th January 1949.18 This speech is viewed as the start of a new era, a particular historical period (Sachs, 1992a), the era of development (Esteva, 1992).19 The speech presented grand ambitions; to give all the people of the world what societies characterised as 'developed' had: better conditions of living, democracy, rapid growth in material production which coincides with a high degree of industrialisation and urbanisation (Escobar, 1995). Before Truman's inauguration speech, US Senator Herbert Hoover, at Truman's request, made a tour of 38 countries to assess global food supplies and to see how surpluses and intervention from the US might be deployed in order to help the 'underdeveloped'. His depiction of these poverty-struck countries was devastating. Hoover ended his address to the American people by saying: "But we can save these people from the worst - if we will" (cited in Hancock, 1989: 70). Truman's doctrine about development, which in many ways builds on Hoover's report, states:

We must embark on a bold new program for making the benefits of our scientific advances and industrial progress available for the improvement and growth of underdeveloped areas. More than half of the people of the world are living in conditions approaching misery. Their food is inadequate, they are victims of disease. Their economic life is primitive and stagnant. Their poverty is a handicap and a threat both to them and more prosperous areas. For the first time in history, humanity possesses the knowledge and the skill to relieve the suffering of these people … our imponderable resources in the technical knowledge are constantly growing and are inexhaustible … The old imperialism - exploitation for foreign profit - has no place in our plans … Greater production is the key to prosperity and peace. And the key to greater production is a wider and more vigorous application of modern scientific and technical knowledge (cited in Porter, 1995: 66-67).

Escobar (1995) understands this vision not only as an American idea, but also as a result of the general post-war conditions. Despite Point Four intentionally being a public relation exercise (Brigg, 2002: 424), post-development scholars regard it as the initiation to the era of development. Three crucial elements arose from this speech, which affected development thought in general and modernisation theory in particular: First, it generated a conceptual division of the world into two separate and distinguished entities, the developed and underdeveloped. Nustad states that "[d]evelopment colonised the world by ordering in into 'us' and 'them', 'the developed' and 'underdeveloped' (1998: 42). Kate Manzo applies Jacques Derrida's conception of logocentrism in describing ethnocentrism as one of many manifestations of modernistic procedures. This term describes a disposition to impose hierarchy when encountering familiar and uncritically accepted dichotomies, e.g. north and south, developed and underdeveloped, modern and traditional, core and periphery, etc. (1991). In such dichotomised terms, the latter is understood in relation to what constitutes the former. "The first term in such oppositions is conceived as a higher reality, belonging to the realm of logos, or pure and invariable presence in need of no explanation. The other term is then defined solely in relation to the first" (1991: 8). Secondly, Truman's Point Four says something about development: He largely defines what developed is and how to become developed. To become "like us", in the sense of having prosperity and peace, is reached through "…a wider and more vigorous application of modern, scientific and technical knowledge". Simultaneously, Point Four underscores the idea of development as an intentional directed process, and that the "underdeveloped areas" will become developed through application of Western elements. Thirdly, Truman says something about who is to do the development, that it is "We" that "must embark on a bold new program" because "humanity possesses the knowledge and skill to relieve the suffering…" in order to make our "…progress available for the improvement and growth of underdeveloped areas".20 Among the power-effects of Truman's speech are that it depoliticised the relations between rich and poor and consequently reproduced the largely immanent temporal segregation of the development concept. Simultaneously it stipulated how and who to bring about development to the underdeveloped (Nustad, 2003a), thus producing a room for intervention by external development agents. This established top-down technocratic approach was later manifested in Rostow's work about modernisation theory.

Scholars regard Point Four as the initiation of modernisation theory, which was to become the most influential and famous paradigm within development, especially represented by the North American economist Walt Rostow (Nustad, 2001a; Escobar, 1995; Gardner & Lewis, 1996). Development was perceived as a natural process that had to be emancipated and directed by those already developed through their intervention with the aim of speeding up the process. Development was perceived as a uniform process and was supposed to be the same everywhere, thus and as in opposition to later development theories the context, i.e., population and area, where not relevant for the 'unfolding of development'. Modernisation theoreticians saw development as immanent everywhere, and lack of it was explained with reference to the systemic and structural level of a society, which hindered the forces and process of development (Nustad, 2003a: 51). Rostow's version of modernisation theory is evolutionist. It stipulates a universal development model consisting of five phases21 that an underdeveloped society has to pass in order to become developed: Poor countries have to pass through the same phases as the industrialised and developed west did. The poor countries' lack of development is perceived as prehistorical in relation to the developed, which underlines the notion of temporal segregation. According to modernisation theoreticians, external development intervention can help increase the velocity in the processes of development by introducing certain elements that are attributes of a later developmental stage, i.e., elements ascribed to a higher degree of modernity. In the 1950s and 60s, the notion of a modern society referred to certain components such as cities, capital entrepreneurship, democracy, the rule of law, education, and science and technology.22 This notion about modernity shaped the discourse around modernisation theory. Scholarly literature on modernisation theory uses the term 'development' to mean the process of transition or transformation toward a modern, capitalistic industrial economy (Ferguson, 1994: 15). Whereas modernisation theoreticians emphasise the introduction of western elements into an underdeveloped area in order to help promote development, dependency theoreticians reject the utilisation of a western model as a blueprint when trying to develop other areas.

Dependency theory

Dependency theory emerged in the 1970s as a result of - and reaction to - the theory of modernisation and as an explanation of unequal structural and economic possibilities between north and south. It also gave an account of why many countries are poor while some are rich. Dependency theory, also signified as underdevelopment theory, draws on Lenin's theory of imperialism and Marxistic analyses of capitalistic societies (Gardner & Lewis, 1996). Dependency theoreticians were one of the first groups to see development in terms of political and historical structures, insisting on the coherence between underdevelopment, exploitation and political structures, describing underdevelopment as a politically active process and not due to a partition in time. Instead of presenting a model for development, dependency theoreticians focused on reasons to why some countries are underdeveloped and others not. The explanation was found in the 'structures of underdevelopment', such as unequal relations between north and south, the impoverishment of the south in terms of e.g. trade, and a hegemonic western control in international institutions. Dependency theory thus challenged the temporal segregation of the modernisation theory when it explained underdevelopment in terms of contemporary political and economic structures. While modernisation theory saw lack of development as a result of poor countries' lack of integration into the capitalistic world marked, dependency theoreticians stated that poor countries in fact were integrated into the world market, which largely was the reason for their underdevelopment and their difficulties of reversing these trends. If the south were to become developed, it needed the same chances and free market opportunities as the north had. The main distinction between modernisation theory and dependency theory is that while the former focuses on internal relations in a country for its lack of development, the latter gives focus to politics and structures imposed on the poor countries. Whereas Rostow sees the fifth stage in modernisation theory's model of development as the objective for all countries, dependency theoreticians see this stage as a critical point where inequalities between poor and rich countries are intensified and reproduced, since the position of rich countries largely relies on the exploitation of the poor ones.

Regarding the structures of power between north and south, dependency theoreticians argue that development is an essentially unequalising process (Gardner & Lewis 1996): While rich nations get richer, the rest inevitably get poorer. This creates a relationship between rich and poor perceived as centres and peripheries (ibid.; Manzo 1991). Structures generated and introduced in the era of colonialisation maintain these structural inequalities. They are sustained by the international system and economy that also are based on unequal structures from the colonialisation era, creating an even more stratified and divided world. In opposition to modernisation theory, dependency theoreticians argue that rich and poor countries are interconnected, and in order for the poor to become developed this relation has to become restructured. The differences between north and south are explained in terms of the south's structural lack of possibilities in relation to the north. Dependency theoreticians state that a dismissal of these structures would enable poor countries to pursue their potential regarding processes of development (Gardner & Lewis, 1996).

'Developmentalism'23 (Manzo, 1991) emerged as a result of the Truman doctrine, and in the 1970s its counter reaction came with the dependency theory, mainly fronted by André Gunder Frank (1967), Samir Amin (1976) and Wallerstein (1974). Both modernisation and dependency theory have been enormously influential in the way development is perceived and practiced. Whereas dependency theory mostly influenced theoretical understandings and obtained resonance within academic sectors, developmentalism influenced development practice and was largely echoed among development policy-makers and implementers (Eriksen, 1989). In 1991, Manzo argued that the field of Third World studies "…once again [was] in a state of crisis" (1991: 30), since both developmentalism and dependency theory largely had been falsified due to their lack of results and positive output, as an addition to the general academic critique of the two development paradigms. Manzo suggested that both theoretical schools, especially the 'dependentistas',24 should consider the critique in order to develop the theories further. Developers and theoreticians have acknowledged the shortcomings of the grand theories but only to a small extent managed to provide the theories with new paradigmatic input. This is echoed in development practice.

In his critique of the dependency theory and the explanation as to why the dependency theory did not entail a paradigmatic shift in development practice, Hobart (1993) argues that the dependistas never managed to free their rhetoric and world-view from the modernisation theoreticians. With focus on knowledge, Hobart shows how the dependency theoreticians reproduced the expert knowledge of the modernisation theoreticians, thus contributing to the generation and reproduction of a development discourse and expert knowledge about development (ibid.; Nustad, 2003a: 80-85). As with modernisation theory, dependency theory also generated a schism between the theoretical model and those supposed to benefit from it, since the intended beneficiaries (e.g. the rural farmers) were as alienated by modernisation theory as they were by dependency theory. Both theories represent a top-down approximation from the developers towards the intended target group. Consequently, dependency theoreticians continued to reproduce the schism between expert knowledge and local knowledge.

In the late 1980s, development studies took a turn towards a post-structural and discursive approach to development, labelled as post-development. Post-development theoreticians do not have a mission in generating a new theory or model on how to execute development practically. They seek to understand how contemporary 'development' functions by identifying its genealogy in order to give an account of why so many development projects seem to fail. Post-development is a radical reaction to the problems aligned with development, and focuses on the structures of development, the underlying premises and gives an account of development's unintended side-effects.

The post-structural approach to development

There have been many reactions towards development and its failures. Sachs argues that the period after WW2, the age of development, is coming to an end since the "…four founding premises have been outdated by history" (Sachs, 1992a: 1).25 Development stood as the idea that guided emerging nations in the post-war era. The template and guide for development were the US as a 'beacon on the hill', which through Truman's Point Four launched the idea of development with a call to every nation to follow in its footsteps. Today, after over 50 years of development, the lighthouse shows cracks and is starting to crumble. According to Sachs, "[t]he idea of development stands like a ruin in the intellectual landscape" (ibid.). Several other critical voices to the project of development have been raised during the last decades.

Next to the reaction fronted by the dependentistas, 'anti-development', 'beyond development', 'alternative development', 'human development', and 'post-development' stand as various reactions and critiques to the project of development. Anti-development is a rejection inspired by anger with development's 'business-as-usual' and wants to abandon the project of development as it is known today. Beyond development combines this aversion by looking over the fence after other solutions, theories and methods for approaching the issues of development. Alternative development focuses on the lack of popular participation as the reason for flawed development projects, arguing for participation as the crucial means in development. Human development addresses the need to invest in people in preference to infrastructure and the state (Pieterse, 2000). Post-development, the most radical critique which also has gained most attention, focuses on the underlying structures and premises of development and the unintended side effects that are produced. What sets it apart from the other mentioned critical approaches is on the reason for rejecting development: "It is rejected not merely on account of its results but because of its intentions, its world-view and mindset" (ibid.: 175) because it implies cultural westernisation and homogenisation (ibid.). Thus, Sachs argues that "…it is not the failure of development which has to be feared, but its success" (1992a: 3). This radical approach has lately been somewhat modified. The intentions are largely seen as good. Prominence is rather given to the structures and discourses of development and their formative power that embed and shape the processes and practices of development and its agents, and thus development's outcome.

To get to grips with development practice and its mindset, post-development scholars trace the genealogy of development, and give emphasis to its conception in Truman's Point Four, which ideas later were manifested in Rostow's work on modernisation theory. By depicting development's genealogy post-developers illustrate how development constitutes a discourse where social meaning is produced and maintained, while diverging and contesting knowledge largely is ignored and thus has small pragmatic influence. As stated above, dependency theory had small practical influence on the existing and established system of development knowledge. Post-development scholars argue that development agents, institutions and policy-makers have legitimised, constituted and reproduced the development discourse rather than considering the critique. Discursive development practice reproduces existing knowledge. Seeing development as a discourse came because development studies were in a crisis, and some intellectual circles declared the concept of development dead when entering the age of post-modernity. Pieterse argues that "[p]ost-development overlaps with Western critique of modernity and techno-scientific progress" (2000: 176). As post-modernity is a cultural and intellectual rejection of modernity (Gardner & Lewis, 1996), post-development is a rejection of development.26 Scholars who adhere to the post-structural development critique take advantage in analysing development as a discourse. Hence, these scholars are able to identify what development does and how it functions instead of presenting a solution to underdevelopment as their predecessors tried to do. The strength of post-development's discursive approach is that it allows one to distinguish between the moral aspects of development issues and the theoretical apparatus that has monopolised development discussions, solutions and interventions (Nustad, 2000: 223; Nustad, 2003a). This thesis is about the latter, that is, development discourse as one system of knowledge development agents relate to in their work, and how this is articulated locally through a project.

The development apparatus

The development apparatus is the instrumental and implementing aspect of development discourse. The development apparatus consists of an institutional and a conceptual apparatus and it thus manifests development discourse conceptually and institutionally. It is articulated through actors who operate within this apparatus (Ferguson, 1994). The development apparatus is an aggregate term that denotes the various ways the development discourse is maintained and articulated, but also signifies an important element that actors relate to in bringing around development. It consists of certain objective ways of talking about development and stipulates how to plan, design and bring out development. The development apparatus is communicated through e.g. documents, reports, applications, policies, organisations, institutions and projects. The first four reflect the conceptual apparatus, while the last three articulate the institutional apparatus. There is interplay between these two parts, and together they comprise "…the apparatus that is to do the developing" (ibid.: 17).

As discourse, the term apparatus originates from Foucault (dispositif). The dispositif is both a 'thoroughly heterogeneous ensemble' of discursive and material elements27 and the system of relations established between these elements (Brigg, 2002: 427). The apparatus emerges from the genealogy of the discourse manifested in and through the combination of meaning, power and materiality. The apparatus is a systematic network of different strategies, practices, mechanisms and procedures that are active in generating knowledge (Schaanning, 1995: 9-10). Within development discourse's genealogy, the development apparatus emerged in order to arrange and promote development. International institutional and discursive development in the post-war period allowed the emergence of a strong apparatus never seen before. This apparatus and discourse give precedence on how to act through a process of normalisation that implies the use and manifestation of power (Brigg, 2002). The apparatus delivers development problems, solutions and means to plan and intervene (cf. Green, 2003). In general, it delivers a top-down approach to development, which is reflected in the way development agents work, because "…the intervening agent has to construct the object it addresses in a way that makes intervention possible" (Nustad, 2001a: 484). The apparatus delivers methods on how to intervene which reflect and adhere to the development discourse.

The development discourse has become manifested as an objective form of knowledge through the regularity of various development institutions' practices. This system of knowledge constructs its field of intervention as a particular object and creates a structure around that object (Ferguson, 1994: xiv) which development interventions rely upon. Instead of approaching development in terms of theory, concepts and methods, a post-structural discursive approach sees development as a "…form of knowledge which, while including or making use of a series of objects, concepts, and methodological choices, are primarily characterized by regularity in dispersion" (Escobar, 1991: 666). This discourse represents the knowledge development agents act upon. In being formed by the discourse, actors' agencies thus reproduce the discourse. Ferguson denotes this process as the reproduction thesis - that "[a] structure always reproduces itself through a process" (1994: 13). By acting upon this established development discourse, development agencies and agents reproduce it. As a result, knowledge and practices are normalised. These practices feed into and thus strengthen the discourse. This makes it even more difficult for challenging knowledge and practice to get resonance.

Discourse and agency

Too strict a view on the discourse and its formative power has implications on the general view on actors and their agency. Ferguson states, regarding actors' relations with and reproduction of the development discourse, that

"[w]hatever interests may be at work, and whatever they may think they are doing, they can only operate through a complex set of social and cultural structures so deeply embedded and so ill-perceived that the outcome may be only a baroque and unrecognisable transformation of the original intentions" (ibid.: 17).

Ferguson proposes a rather static picture of actors and their agency in relation to the discourse's formative power. Actors are seen merely as representatives, bearers and reproducers of the development discourse. This is the core in the post-developmental explanation of why so many development projects seem to fail, viz. that actors' agency are constructed by the discourse these actors bear and are embedded in. Actors are deeply embedded into the structures and knowledge of what is seen by post-development scholars as a constructivist and hegemonic discourse. This view on actors and their agency is also amongst the main critiques directed towards post-development theory. The post-developers' comprehension of development discourse as a constructivist structure and system of knowledge provides no room for local variations, complexity and agency in development practice. According to post-development scholars, the formative power of development discourse makes the actors involved reproduce the discourse they relate to non-reflexively. Discursive development practise by various discursive bearers necessarily, according to post-developers' and Foucault's notion of discourse, reproduce the discourse through their agency.

Post-development theory has been opposed in several ways. Below an account is given of some of this critique. Additionally, an actor-orientated approach is offered in combination with a post-development theory to meet some of the critique but also to gain the relevance to some of the lessons from post-structural development critique.

Taking post-development theory further

A lot of criticism has been directed against the post-structural critique of development. This includes theoretical criticism on an eclectic use of Foucault, post-development's lack of instrumentality, lack of focus on agency, lack of an empirical foundation, and its lack of an alternative when dismissing development as a project. Post-development theory is fruitful in understanding development's formal order and system of knowledge. In taking some of the critique into account, mainly by giving attention to agency where the development discourse is articulated, post-development theory's approach to development also gains relevance.

Criticism of post-development theory

According to the anti-Foucauldian Ray Kiely, post-development scholars are wrong in their approach to discourse as a system of knowledge that constructs the reality: "Development is, in the Foucauldian sense, a particular discourse which does not reflect but actually construct reality. In doing so, it closes of alternative ways of thinking, and so constitutes a form of power" (1999: 31). Kiely argues that since discourses do not have agency, it is impossible for individuals to be the product of discourses. The value of a discursive approach depends upon whether agency is taken into account or not. If not, development becomes a cultural construction of post-development scholars. Including agency in the analysis also meets another weak point of post-development theory. Agency nuances the view on discourse, which is otherwise almost exclusively constructed from the analysis of donors' ideologies, policies and documents without identifying or having any real empirical foundation where the development discourse is deployed and articulated.

After identifying post-development's serious lack of taking agency into account, Kiely assesses post-development theory's notion of power. The conception of a constructivist discourse and its power is based on the assumption of Foucault's notion of power; that power does not operate over and against individuals, but rather is 'a machine in which everyone is caught', which thus neglects the agency behind discourse (1999: 36). Kiely argues that this renders the idea of power meaningless, since it does not say anything about who distributes and utilises the power or in what ways, as well as it has implications for the progressive and political aspects of power (ibid.). Power is manifested and articulated in interaction, and is not a priori latent in structures. Consequently, actors' agency and the interplay between actors and discourse must be given focus when identifying power relations. Further Kiely attacks post-development theory for being inconsistent in its anti-essentialism, because "[w]hile it champions cultural diversity and the difference as a source of resistance against Western domination, development itself is portrayed 'in terms of a monolithic hegemony'" (ibid.: 38). Irrespective of the dimensions of time and space, development, in the post-structural approach, constitutes the exercise of western power over non-western people. There are two implications related to this: First that post-development operates with only one tool when analysing development as a discourse, meaning that all development is understood and approached according to this discourse. Whereas post-developers criticise development in representing a hegemonic and constructivist discourse, their counterparts condemn the reductionistic view of post-developers on development in presenting the development discourse as a realm in which everything fits. As post-development scholars argue that development is a western construction representing a panoptic monolithic approach, post-developers are guilty for largely the same in merely having a discursive approach to development. Hence, both views close off alternative ways of thinking and seeing the world. Both views criticise the other of having merely one theoretical tool or approach: 'If the only tool you got is a hammer, the whole world appears as a nail' (Abraham Kaplan, in Nustad, 2001b: 78). Secondly, most post-development theoreticians' general notion of the world is that it consists of two entities; the evil developed west and the noble victims of the south, a view that initially gave rise to post-development theory. As a result, post-development scholars operate with the same concepts and world-view as the structures and ideas they tend to criticise, i.e., that the world is split and divided into two separate parts. Ironically, post-development scholars are themselves largely part of the development discourse and reproduce the realm they initially are criticising. Ray Kiely also raises criticism of post-developers in terms of the relation with politics, relativism, and how post-development ambiguously celebrates tradition. Post-development criticises the development apparatus for neglecting ethnographic particularism. Simultaneously, post-developments see individuals as static and subordinate to the discourse's formative power, and that individuals in general lack the capacity to oppose the discourse's power (cf. Kiely, 1999).

Morgan Brigg questions post-development scholars on their use of Foucault, and argues that most post-development scholars seem to use Foucault eclectically. Brigg puts special emphasis on post-developers' use of Foucault's dispositif (apparatus), the way the concept of power is used, while he argues that Escobar's (1995) use of Foucault is limited to 'a particular sort of style and a sprinkling of the name Michel Foucault and quotations from his work' (Brigg, 2002).

Pieterse (2000) questions the ideas of post-development on a wide range of elements. Most important, he dismisses post-development as it is known because it operates on terms of developmentalism, which post-development sets out to criticise, because it "…replicates the rhetoric of developmentalism, rather than penetrating and exposing its polysemic realities". Further, post-development "…echoes the 'myth of development' rather than leaving it behind" (2000: 188). Still, the core weakness according to Pieterse's viewpoint is that post-development does not give any answers on 'how to do development', that post-development is only an "…endorsement of the status quo…" (ibid.: 184). Schuurman (2000) supports Pieterse in post-development being flawed in terms of not offering any alternative programme for development practice.

Nustad's response to the latter argument is that the "…lack of instrumentality is not a weighty argument against the analysis itself" (2001a: 479) because the call for alternatives and post-development's attempts to demonstrate why development interventions do not work must be kept separate. Nustad addresses some of the post-development critique by including "…an examination of how development interventions are transformed in encounters with target populations…" (ibid.: 480) stating indirectly that this points a way forward for development practice.

It is to the post-development context and discourse that this thesis relate. The approach and analysis will draw upon post-developmentalism, but also address some of the shortcomings of post-development theory by including the role of actors and agency. Post-development theory is a useful approach in order to understand how the formal order of the development sector and donor-recipient relationship function. It gives an insightful account of one of the many realms and systems of knowledge development actors relate to. Despite this, I argue that a post-developmental approach do not generate sufficient information to understand and analyse a development project and what is really going on due to its exclusive focus on discourse and structures, while it neglects the multiple realities and various forms of knowledge. A traditional anthropological insight is that "[t]here is no standpoint from which a phenomenon can be grasped in its entirety" (Nustad, 2003b: 127). I complement the discourse perspective with a focus on actors and actors' agency. By including an actor-orientated approach, one meets the most serious critiques against post-development, because attention is on agency, which also gives the analysis an empirical foundation. Utilising two approaches, what Denzin calls theoretical triangulation (Denzin, 1989), give an account of actors' agency and their relation to structures imposed through a donor-recipient relationship. It also acknowledges the local contextual conditions which form an actors realm, what Barth calls cultural stock (Barth, 1993). Hence this calls for an analysis of actors and development agents in relation to different types of knowledge. On the one side, discursive knowledge, and on the other side, local practical knowledge. Discursive knowledge refers to development discourse, while practical knowledge refers to the multitude of knowledge all actors relate to without necessarily knowing it themselves. The latter is implicit, embodied, non-reflexive - a sort of taken-for-granted knowledge. It is not given that it is congruence or symmetry between discursive and practical knowledge. Discursive knowledge represents a type of expert knowledge, which in its deployment intersects with local and practical knowledge. A point is that both experts and locals can have both systems of knowledge, or alternate eclectically between a discursive and a practical understanding of the world.28 This implies that both systems of knowledge exist on the same logical level but are not necessarily identical.29 Norman Long describes the encounter between different systems of knowledge as a process or situation of interface. In combining both a discursive and actor-orientated approach, representing two different systems of knowledge, the term interface must necessarily be applied in order to describe the encounters between different systems of knowledge. "A principal reason why it has been difficult to integrate structural and actor perspectives is that they entail opposing (or at least diverging) theoretical and epistemological assumptions, similar to Kuhn's paradigms that are incompatible until a 'scientific revolution' confirms the paramountcy of one of them" (Long, 1992a: 18). The concept of interface enables a combination of a structural and actor-orientated approach in the analysis, because it takes the encounter between different systems of knowledge, i.e., the situations of interface as articulated via actors, as its primary focus.


Norman Long defines "…a social interface as a critical point of intersection or linkage between different social systems, fields or levels of social order where structural discontinuities, based upon differences of normative value and social interests, are most likely to be found" (1989: 1-2). Interface is a methodological device for studying linkages between structures and processes, and encounters between different systems of knowledge as articulated through actors. Interface helps to bridge the gap between structural and actor-orientated research. Long states that interface is an analytical tool for understanding what happens in the encounter between different knowledge systems. Long calls for a "…thorough-going actor-orientated approach which builds upon theoretical work aimed at reconciling structure and actor perspectives" (Long, 1992c: 4). This is to counter the resurgence of simplistic system thinking, stressing the importance to acknowledge and take the ethnographic particularism into account (ibid.). The fruitfulness of using interface as a methodological and analytical tool in an actor-orientated approach to the encounter between various systems of knowledge is that its "…concepts are grounded in the everyday life experiences and understandings of men and women, be they poor peasants, entrepreneurs, government bureaucrats or researchers" (ibid.: 5.). Actor-orientated research takes the 'multiple realities' and diverse social practices of various actors into account and makes it possible to get to grips with these different and often incommensurable social worlds of different actors (ibid.).

Development discourse's encounter with 'multiple realities' involves a transfer of technology, knowledge, resources and organisational forms from the more developed world or sector of a country to the less developed parts (Long, 1992a: 19). The encounter denotes a process of transformation as the formal order of development "…is transformed through acquiring social meanings that were not set out in the original policy statements" (Long, 1989: 3). Situations of interface articulate factors which cannot be directly linked to the development programme itself, but evolve as a result of the intersection of different fields of knowledge. In dealing with multiple realities, acknowledging potentially conflicting social interests, we must look closely at the issue of whose interpretations or models prevail over those of other actors, and under what conditions. "Knowledge processes are embedded in social processes that imply aspects of power, authority and legitimation…" (Long, 1992a: 27). This discussion brings out certain parallels between power and knowledge processes. Like power, knowledge is not simply something that is possessed and accumulated. Nor can it be precisely measured in terms of some notion of quantity or quality. "It emerges out of processes of social interaction and is essentially a joint product of the encounter and fusion of horizons" (ibid.). Power and knowledge must therefore be understood relationally and not treated as if it could be codified, depleted or used up. That someone has power or knowledge does not necessarily imply that others are without, nor is this the case in the development sector concerning the relations between donor and recipient. Recipients are not incapable and powerless in their encounter with externally imposed structures, rather they apply a wide range of strategies to cope with the formal order of discursive development. One cannot generalise over the multitude of local and practical knowledge. Neither is it correct to generalise about what happens in the encounter between different systems of knowledge. Whereas the development discourse reflects a formal order of development, what happens in the encounter with other systems of knowledge as it is deployed in various settings is solely an empirical question. This thesis is about the processes involved in the donor-recipient relationship, i.e. the encounter between the development discourse and local practical knowledge, of the Integrated Pastoral Development Programme (IPDP) Afar, Ethiopia.

Summarising remarks

The era of development has generated what post-development scholars call a development discourse due to the regularity and increasing normalisation of development practice despite the rhetorical alterations. Post-development scholars argue that bearers of the development discourse have a monopoly in presenting premises to development work and policy, and thus development problems and solutions have become standardised. Nevertheless, this view on development and the portrayal of the development discourse's evolution have severe shortcomings in that it does not take agency into account and it lacks an empirical foundation from where development is deployed and where the development discourse intersects with local knowledge through various development agents. The depiction of development discourse is mainly done through scrutinising development, as seen from and by the donor and not from the recipient organisations or the supposed target group of a development project.

Development discourse, as a system of knowledge, is useful in order to get an understanding of the structures and relations within the development sector, and the general system of development knowledge as seen from the donors' side. It thus often is one of several systems of knowledge in situations of interface between donor and recipient within the development sector. Local practical knowledge, which is identified by an actor-orientated approach, is the other encountering part in such situations of interface.

The next chapter shows that it is plausible to operate with the concepts of development discourse and apparatus with regard to the Integrated Pastoral Development Programme (IPDP) and the network of organisations and actors aligned to it. As that chapter shows, development intervention relies upon simplifications and representations of the field in order to generate quantifiable and legible units to plan and intervene on. The next chapter shows how the development apparatus aligned with the IPDP has constructed and articulated the project and simplistically codified the target area and population. Chapter four, on the other hand, illustrates various counter-tendencies that arise from the situation of interface between different systems of knowledge.

Chapter 3. The formal order of the IPDP

In order to intervene in a society, development agents need legible and tangible units of what is to be developed. The development apparatus constructs the project area as a particular kind of object of knowledge (Ferguson, 1994), which is based upon simplifications and representations of the multiple realities manifested in the area. Thus, what is perceived as complex and difficult to grasp becomes uncomplicated, legible and tangible. The construction of the field as an object of knowledge is based on an expert discursive knowledge of development and is articulated through representations. Representations serve to portray or show reality, but the representations are in nature simplistic towards what they seek to characterise or denote. For the discourse bearers the representations are the reality, while for the discourse analyst representations are the main intake to get to grip with the discourse itself.

The purpose of this chapter is two-fold: First, it shows the representations the Integrated Pastoral Development Programme (IPDP) is based upon, and that these socio-cultural, demographic and geographical representations are simplistic and reductionistic reflections of the different aspects of the project area. The representations reduce, diverge and to some extent oppose local practical knowledge, and the discontinuity between the representations and local knowledge is elaborated. Secondly, this chapter shows empirically the initiation of some of the project's components and how these are conceived. The widely accepted notions of participatory approaches and 'bottom-up' planning are discussed.

Combined, the chapter's two sections show the development apparatus in practice, and how development practice is standardised due to the practices applied to plan and implement development as well as it is a consequence of the standardisation of development problems and solutions (cf. Green, 2003). This standardisation of development practice and problems will be further elaborated in chapter five and six.

Presentation and representations of the IPDP and Aba'ala

The IPDP takes place in Aba'ala wereda in the Afar region. The area was "…unexposed to any sort of development activity sponsored by external donors until 1995", and "[s]ince the idea of development with external components was new for the local institutions and the Aba'la people, a tremendous amount of time and effort from Development Fund (DF) and Mekelle University (MU), have been put into creating a basis for and understanding of the processes involved" (AIPDP, 2000, emphasis added)30. An IPDP board member states that:

"The Afar people had a lack of working habit, you see them chewing chat all day. They didn't have the concepts on what projects are about, so we had to talk them into it. When they heard about a development project coming they thought they would get food and become developed for free without doing anything. We had to tell them to work, that if they want our development they have to do as we tell them" (emphasis added).

According to an IPDP review, what they 'had to be talked into' was, among other things, the role of external actors in relation to the Afar social system and environment, "…choice of problems and the development options as well the mechanics of project planning and implementation with the participation of the local people" (AIPDP, 2000:1). The IPDP application submitted to NORAD for the project year 2002 presents the project as follows:

"The Integrated Pastoral Development Programme (AIDP) started as a pilot phase in 1998 in Aba'la Wereda of the Afar National Regional State. The Programme involves integrated grassroots community development work, and is carried out in collaboration with Mekelle University, the Bureau of Agriculture (BoA), the administrative councils and community elders (Mabilo).

The AIPD has four main goals:

These goals include aspects concerning human development, human relations, and natural resource management, The goals are linked with the overall development process for the programme in Aba'la, covering important aspects of people's lives. This means focusing on the pastoral mode of production, acknowledging the fact that Aba'la is a heterogeneous physical and human environment. In order to achieve the aforementioned goals, the programme relies on its various constituent components (underlined below). Each component comprises several activities."

At their web pages, the Development Fund presents the IPDP:

"The project, which started in 1998, is based on the needs of the nomads, their way of life and their production system. The project includes measurestoprotect [sic] water and soil, a nursery, training of "barefoot veterinarians", micro-credit schemes and training that enable women to run their own businesses".31

The IPDP is a relatively small project, with an average annual budget of approximately one million NOK. Despite limited resources, the IPDP has large objectives. In addition to create popular awareness about the project, another of the initial tasks of the institutions that adhere to the IPDP was to create a 'basis for the mechanics of project planning and implementation'. One of the foundations of a development project is to generate legible and tangible units of the project area possible to intervene in, that is, to define the field as objective knowledge.

Representations of Aba'ala

According to the IPDP project application for 2002,

"[t]he pastoral production system is an efficient and sustainable way of utilising rangeland resources in the marginal drylands. In the lowlands of Ethiopia, the drylands cover an extensive area and support about 12% of the country's total population. The Afars are among the dominant pastoral and agro-pastoral groups found in the lowlands of Ethiopia, occupying the northeastern parts particularly, in the area now designated as Region 2 comprising five administrative zones".

The majority of the Afar residents are "…pastoralists and some are converted to agro-pastoral life style practising spate irrigation" (AIPDP, 2002, emphasis added). The area belongs to the dryland part of Ethiopia with an annual rainfall often less than 500mm, which coupled with recurrent draught and subsequent famine "…has severely damaged the Afar livestock dependent economy, weakened their traditional coping mechanisms, and exposed them to frequent external food aids" (ibid.). This is the only presentation given in the project documents of the socio-cultural and natural setting of the Afar people.

The household survey of 58 samples conducted in different tabias within the wereda of Aba'ala shows that the area is not only settled by pastoral or agro-pastoral people: 13 households (22.4%) state they exclusively rely upon farming methods as their household's main income-generating activity. They are not part of the project's defined and intended target group.32 Only one household (1.7 %) states to be solely part of the pastoral economic system. This group increases to 6 households (10.3%) when those who combine pastoralism and trade are included.33 The remaining 89.7 % is thus not directly defined within the target group since the IPDP is based on the 'needs of the nomads'. However, the survey shows the tendency that those most satisfied with the project are the Tigreans living in the city of Aba'ala. The reason for this, given by a board member, is that since the Tigreans were the first group to settle in the area, approximately 35 years ago, they chose to settle near the river. Thus, they have a larger potential to utilise the river. Additionally, many of the IPDP project components and activities favour farmers since the activities necessarily need to take place somewhere, which underscores the predicaments in the approach to nomads. Since the Afars that settled did that at a later stage, Aba'ala city has become divided between the Tigreans living near the river that the permanent IPDP activities are centred around, and the Afars generally living in more remote areas. In this way, the Tigreans benefit more from the project components concerned with agricultural activities than the Afars do. According to the 1994 census, the population of Aba'ala town is 3.300, of which 2.765 are Tigreans and 405 are Afars (Kelemework, 2000), which shows the initial misconception in the process of planning based on the assumption that Afar region is merely inhabited by Afars. This is, however, not only a misconception. It is also due to the need for legible units when planning, and thus to make the complex appear as legible and homogeneous. The IPDP relies upon these initial established representations and simplifications.

The IPDP codifies Afar as an object of knowledge based on simplifications of the area in terms of geography and demography through which it seeks to address the needs of the people that live there. "IPDP want to address the main problems of the pastoral people … IPDP is trying to grasp the idea of the people, the real needs", a board member explains. These representations are stipulated in formal project documents, but are also reproduced by the IPDP implementers who adopt these representations. A board member states that the project activities are selected "…by the grass roots, the poorest of the poor, in order to address the needs of the people. Asking the beneficiaries about their needs, we were able to identify and implement this. Of course, in the end it is a financial issue, and since we cannot give them all they want, we select the most appropriate for the area". This statement illustrates a general comprehension among the various IPDP staff and board members.

Self-presentation and representations of the IPDP

Initially, upon arrival to Mekelle, the work to identify the network around the IPDP began. People connected to the project were asked to explain and elaborate around the IPDP. The terminology the different actors involved use to depict the project are largely identical,34 and is characterised by their disposition to merely refer to the different project components' and activities' names as something that symbolise a larger unit, as metonyms it is superfluous to elaborate around. A high official board member states:

"Everything is there, in the project. Lets say veterinary service, water harvesting, soil and water conservation, women in development, natural recourses, capacity building and management, I mean everything is in the project. All the components to help the Afar people is in the project".

Oral explanations and elaborations about the IPDP refer almost by default to the same terminology as used in the project documents. Consequently, expert knowledge about the IPDP as an object of knowledge is confirmed and reproduced. Reproducing this knowledge and its representations imply a recycling of the project's inducement and give legitimacy to future interventions and activities.

The IPDP is structured in a taxonomic hierarchy, i.e., the project consists of certain objectives promoted through different components that are composed of different activities. The four objectives of the IPDP are "(1) capacity building of the local administration in Aba'ala; (2) preventing environmental degradation of the rangeland; (3) increasing food security situation in the programme area; and (4) strengthening the integration and relation between the Afar and Tigray peoples" (AIPDP, 2000: 3). These goals are promoted through different components: "-Nursery activities: Production of tree seedling and vegetables production. -Soil and water conservation: Terracing and planting tree seedlings. -Women in development: Pilot credit scheme and small business management training. -Water development: River diversion, ponds and cisterns development. -Veterinary Services: Training of paravets, refresher course and dipping vat. -Capacity building: Training of institutions in Aba'la" (ibid.: 7-8; and IPDP application, 2002). From 2001 a 'HIV/ AIDS Prevention Initiatives' component is added, which consists of three activities: radio broadcasting, education and poster production. The project manager elaborates about the project in general:

"The IPDP, which is an abbreviation for the Integrated Pastoral Development Programme, started in 1998, in April, and after that the program has been implementing a lot of activities which vary from year to year. And generally when we look at these activities they focus on water development, local capacity building in the form of short-term training for the office bureau in Aba'ala as for the ordinary people. The other component is nursery, which focuses on production of local trees, seedling, vegetables and fruits, which we introduce to them. And another component of the IPDP is focus on women which takes the form of training, the form of credit and the form of support in vegetable production. The other component is soil and water conservation which we can assume is a model for the community, so when we say soil and water conservation is not a large scale, it is in a small area that can be used as a demonstration for the people. The other component was water harvesting for domestic use, and water harvesting for crop production. In the case of domestic water harvesting for domestic use this again takes two forms: One is the form of improving the traditional ponds which are build by the people themselves, and the other is modern concrete build pond which is called cistern, and the diversion is in the form of diverting flood water which can enter into the cultivated fields of the farmers. So sometimes this flood diversion maybe work using concretes or sometimes it can be simply soil embankment. So more or less this is the activities the IPDP has been concentrating on since the beginning on 1998".

An external evaluation of the programme sums up that: "The AIPDP components are: nursery activities, soil and water conservation, women in development, water development, veterinary services and capacity building" (Dioli and Gebre-Mariam, 2001: 4). To understand and to be integrated into the realm of the project entails not only to adopt these terms, but also to understand the larger whole that these metonyms, or representations feed into. Drawing on Weber's distinction between acts and events, Barth (1993) argues that if one is able to grasp persons' 'cultural stock', it renders possible to understand the meaningful aspects of their events, that is, acts. Whereas events are solely the observable, acts include the meaningful elements and performances of these events. Acts are imputed with meaning. The development discourse is one of many aspects that make up development agents' cultural stock. From post-development scholars' point of view the development discourse is a uniform, monolithic and hegemonic discourse that has a low degree of inter-discursivity, i.e., that it mainly draws experience from itself rather than other discourses (Jørgensen and Phillips, 1999: 72-105). Consequently, the discourse is reproduced. In order to grasp development in general and the IPDP in particular, knowledge about development discourse and its effects is necessary. In the following the representations of the various project components as articulated in project documents are presented.

Nursery activities

The nursery is an area of 2.5 hectares of land with good access to water. It is used as a demonstration plot, to grow seedlings, while providing an area suitable to grow vegetables for selected participants. The nursery is situated directly north of Aba'ala river, near the new buildings of the local governmental offices where, amongst others, the Department of Agriculture has its office. The area is allotted to Mekelle University (MU). Directly south of the river lies the village of Aba'ala. As one of very few places in the area the nursery has direct access to water from the river nearby, which is diverted onto the site on a regular basis in order to sustain the irrigation and productivity. The site is run in collaboration between MU, the IPDP and the Dryland Husbandry Project (DHP).35 As a demonstration site, local farmers can come to observe and learn about vegetable production, irrigation practices, handling and use of fruit trees. Fruit seedlings are also produced and distributed to local farmers. 0.87 hectare of land is allocated to women for vegetable production in order to supply their diet and benefit their household's economy by selling vegetables on the local marked. "This is a good activity. It has really benefited to the diversification of the Afar's diet, but most important the women are empowered", the DF coordinator states.

There are some incongruities about these activities. One necessary premise is that the people have access to land and water in order to utilise and gain from the knowledge obtained from the demonstration site. The idea is to make the participants use their learned expertise about growing vegetables in their own compounds. Beneficiaries state that they have learned a lot from participating in this activity, but that it is hard, if not impossible, for them to get an advantage from these 'introduced techniques', primarily due to the lack of water and, secondly, because they lack fertile land. 'Not all of us live by the river, and those of us who don't are not entitled to divert it onto our plot', a supposed beneficiary states. The group that seems to benefit most from the nursery activities is the farmers who live close to the shore of Aba'ala river because they have land, access to water and the opportunity to visit the demonstration site regularly.

In the standardised NORAD application form, point 3.2. questions 'what indicators will be used to establish whether the objectives have been achieved'. Questioning the project manager on how to measure the transfer of knowledge, he states that 'we just have to count how many seedlings that are planted and distributed, how many women that participate in the vegetable production, and the number of people visiting the demonstration site'.36 To measure the activities' success based on quantitative indicators conceals the qualitative effects and whether the skills and knowledge given at the demonstration site are adopted and practised in the participants' own compound. Beneficiaries that live far from the river and the nursery site state that they have visited the nursery and received fruit and vegetable seedlings, but express that they have difficulties in taking advantage of the skills learned due to the lack of water. For some, the situation has worsened, since they had to replant their plot for their new seedlings to the detriment of their old plants. The stipulated indicators to measure the project's achievements do not open for taking this information into account.

Soil and water conservation

On the hillsides, long-stretched horizontal plateaus are built. Rocks support them in order to conserve the soil and water in the area, preventing environmental degradation and erosion. Terracing of hillsides not only minimises erosion, but also generates new arable areas. Tree seedlings from the nursery are planted on the terraces, and the plants' roots bind soil and water, which enriches the soil's biodiversity and prevent erosion.37 Hill terracing is a rather new phenomenon in Afar. The technique is introduced from the Tigrean highland where the extensive use of the labour-intensive terraces has had significant effects for the rangeland. "Since terracing benefits the Tigreans, we wanted to introduce it here as well. This component were to create awareness about rangeland management and show the Afars how it can be done", a board member informs.

There are some problems related to this component that, according to the programme manager and a board member, date back to the initiation phase of 2000 and are connected to the planning and selection of area. The first obstacle is that the selected site is close to a track that leads to a water source where the people who live nearby fetch water. It is also used as a watering point for animals, and a lot of damage is done to the area by animals that graze and browse on the way to the water point. Secondly, the seedlings planted in the area did not develop sufficiently to survive the dry period, since they were planted in the end of the rainy season due to inaccurate timing and planning. Thirdly, as the beneficiaries that were intended to guard the area did not see the profit of it, they stopped watching. Consequently, a guard was hired on the IPDP budget. An external evaluation states that the frequency of participation in this component is "[v]ery low. All the human interventions have been implemented because a payment was offered. It seems that there is no free community participation" (Dioli and Gebre-Mariam, 2001: 10). The notion of participation is challenged since the surveillance of the area is salaried labour and not based on popular involvement, which thus undermines the general comprehension of 'participation'. The lack of popular awareness about the work is also reflected in the household survey as 41 out of 51 households (7 households do not have livestock) state that they use the enclosed site as grazing area for their animals. Now the component of soil and water conservation is phased out of the project, but still it is referred to as one of the IPDP's achievements. Rangeland enclosures and hill terracing are effective in Tigray. They are less effective in Afar because of the large number of livestock and the dependence on grazing areas.

Women in development

"The Women in Development component consists of two activities; the small credit scheme and vegetable production that experts teach them having the objective of empowering women through raising their capacity on several issues, teaching them income generating activities as well as business management so they can have their own economy and business next to their husband's. That is good, and that makes them independent, empowered and developed".

The above quotation is taken from an interview and refers to how an IPDP board member elaborates about the women in development component. Women are segregated as an independent target group to become empowered by learning how to produce vegetables and establish their own business by lending money from a micro credit scheme. This component addresses a limited group of women, mostly Tigreans, who are allotted a piece of land at the nursery site where they can grow vegetables. The project provides the women with equipment and water from the river. Before the IPDP engaged in teaching vegetable production, vegetables had rarely been produced in Aba'ala. The intention is to diversify the households' diet by enabling women to grow vegetables for own consumption and for sale, which will strengthen the women's position in their households. 'Earlier, vegetables were only accessible for the Afars at the weekly market, but they were very expensive. The project has also made vegetables more accessible', the head of the women affairs informs. The 'micro credit scheme' enables women to borrow money in order to start their own 'small business management'. The activities for 2000 are summed up as: "Small business management training for 30 poor women was implemented together with an experience sharing visit to Mekelle. This was then followed by a pilot credit scheme involving small business: an initial credit of 500 birr (with 10.5% interest rate) was offered to each member" (Dioli and Gebre-Mariam, 2001: 11). In 2002 '40 poor women' are planned to receive credit.

The vegetable production that takes place at the nursery site is of good value for the women involved, largely because it enables them to produce goods to sell at the market in order to gain money, which is more sought after than an amendment to their diet. Only a limited number of women are admitted at the nursery at a time due to limited space, but the women involved rotate so that more might get the opportunity. As in the case with nursery activities, not all the women that previously had taken part in this activity are capable of cultivating vegetables in their own compound, due to the harsh climate and lack of water. Vegetable production is almost exclusively done at or near the nursery site, which illustrates the continuous lack of water in other areas. The household survey of 58 samples underlines this: Of the 10 households that have received training in vegetable production, 5 of the 6 households that grow vegetables today are Tigreans situated near the Aba'ala river. The success of this component is largely due to the involvement of Tigrean women, who traditionally are more familiar with vegetable production. However, they recognise the threat to their vegetable production from the general lack of water.

In a meeting between the coordinator from DF and the recipient organisations, the head of the women's affairs office, who is responsible for the micro credit scheme says that the repayment rate was impressively 100%, to which the DF representative expresses gratitude towards the good work laid down in the awareness process around the credit scheme. Later, in an informal setting, one of the local staff of the IPDP informs that 'the people at the meeting are fooling her [the DF coordinator]. Just go and see yourself. I don't like this. She is a friend of mine. They say they have retrieved all the money, but it is not true'.38

Water development

The fourth component is water development. It consists of three activities; 'river diversion', 'pond development' and 'cistern construction'. Before the initialisation of this component, local experts and MU undertook a feasibility study in consultation with local institutions. "The survey identified technical and socio-economic parameters required for the construction of the structures", the internal review informs (AIPDP, 2000: 10). According to a board member, the 'river diversion seeks to combat the soil erosion and formation of gullies resulting from yearly thunderstorms and floods which make land unsuitable for farming'. Further, ponds are dug in different tabias to increase the access to water for those responsible for fetching it, who largely are women. The construction of cisterns refers to the building of structures that will collect and store rain water for consumption by the use of imported materials and technicians as labour force.

Water is probably the most essential and basic need of the people living in Afar, and this component seeks to address this constraint. The initial planning of the river diversion took place in 1999, but the implementation phase did not start before late 2001 due to a conflict within the local community about where to build the diversion since some will benefit while others not depending on where the structure is built. A board member states that "[i]t is good they finally agreed where to build the diversion so we can start. The DF has asked about this several times, but we have waited long time. The feasibility study gave clear recommendations on where to put it. We should have followed that advice instead of letting the people decide. It took them two years!". The conflict arose between 'up-streamers' and 'down-streamers', and "…we said that 'we are not doing anything before you have decided yourself'. It took them two years to decide. Sometimes bottom-up development is difficult, mostly because we have to produce results", the board member continues. Nevertheless, the water development component is the most approved component among the local community, and it mobilises the community to participate. Except for the conflict in the case with the river diversion, very few had any objections to this component, and the household survey shows that almost all identify lack of water as their primary constraint and the building of ponds and cisterns as important.

Veterinary services

The veterinary service is the only component that exclusively addresses the pastoral people. It provides them with training in basic health care for animals and offers them a tool-kit with drugs and equipment to establish their own veterinary post. The participants are selected so that the 'barefoot veterinarians' are geographically scattered and thus cover a wider area. The project manager informs that they "…usually give training for approximately 25 Afars a year, including some women. We give them the skills to practise modern medical treatment, and they return to their community practising this knowledge. We call them 'barefoot veterinarians' or paravets". The participants of the veterinary training course seem to adopt the skills taught, arguing that they see more effect of medical treatment than from their traditional way of treating livestock.

The main problem raised by the participants is the lack of access to drugs once the medical kit given to them is exhausted. The project manager also stresses this problem, which is due to lack of funds. He is frustrated over the situation, since this component would be more successful if they were able to support the previously educated paravets with drugs. Instead, he argues, "…they return to their traditional methods when the medicine is consumed. As long as medicine is available, the activity builds a lot of awareness in the different remote communities. They experience that we give them good knowledge that is to the best for their livestock that they are dependent on". The beneficiaries echo both the positive sides and the problems with the veterinary training programme.

HIV/AIDS prevention initiatives

The HIV/ AIDS prevention initiatives component was added to the IPDP in 2002, and consists of activities such as radio broadcasting, education and poster production with the objective to create awareness around the emerging problem of HIV/ AIDS. The application states: "The spread of HIV/ AIDS is becoming a major threat to achieving economic growth in many developing countries including Ethiopia. … HIV/ AIDS is causing not only health problems but also having significant socio-economic impact on the overall development of the environment" (emphasis added).39

No one associated with the IPDP diminishes the problem of HIV/ AIDS, which in general is a huge problem in Ethiopia. Nevertheless, some of the IPDP staff has ambiguous thoughts about the inclusion of this component, especially since the national HIV/ AIDS programme is already taking place in Aba'ala. "Suddenly we were supposed to work with HIV/ AIDS awareness. Why? It will overlap with the government's programme that takes place there. It is better to spend the money elsewhere, on something to meet the people's needs". Further in the interview, the board member questions the inclusion of this component and says that HIV is a lethal disease that threatens people's lives, and should be understood as that, and not as an element that hinders economical growth. Later an account on how the HIV/ AIDS component was included is given.

The examination of the different IPDP components and activities as communicated in formal project documents illustrates interesting features about the project. First, it shows the simplistic and reductionistic nature of project documents, and that the stipulated representations undermine local particularities, complexity and variety. Secondly, it illustrates that the planning, implementation and measurement of a project depend upon representations and legible units of the field. This is illustrated e.g. by the nursery and women in development components of which the degree of implementation and qualitative success is measured by counting numbers of seedlings distributed and the numbers of participants in the vegetable production. Thirdly, it shows that despite the acknowledgement of Aba'ala as a heterogeneous area and the objective of strengthening the relations between the Afars and Tigreans, the various components largely address the two groups as distinct and separate. The veterinary service addresses only pastoral people, while vegetable production in practice addresses only Tigreans. The examination also shows that Tigreans benefit most from the activities since the project largely focuses on activities and production attached to a locality, and the pastoral people have bad access to good land as well as they are regularly moving to find new grazing areas. Additionally, the activities take place mainly in the centre of Aba'ala where few Afars live. Fourthly, it shows the difficulties of introducing new and unfamiliar elements into an area, especially without popular awareness. This is illustrated with the vegetable production activity, but most notably with the hillside terracing and the enclosure. Due to lack of awareness and that the people did not see the advantage, the project paid the terracing and also a guard since so many Afars let their animal graze at the enclosure.

Combined, this shows that what is stated in the project documents is not directly analogous to local reality and knowledge, largely because the difficulties in planning and describing without making codified simplifications based on general assumptions. Nevertheless, there is interplay between the documents and local practice. Project implementation is grounded on the project documents, which are characterised by the representations created of the project area. Norman Long argues that development consists of positivistic methods of research that make up a tool-box of techniques, which create simplistic systems and causalities and that ethnographic particularism is neglected and not taken into account (Long, 1992c). The tool-box of techniques is not necessarily reflected in the development rhetoric or in the project documents, but finds its echo in development practice. Whereas institutional rhetoric states that the IPDP is a 'community based, grass root orientated, bottom-up project deduced to approach the needs of the 'poorest of the poor', the practices of development planning and implementation are largely donor-guided and top-heavy.

Where and what is the bottom?

"Well, the IPDP is about, well simply if you see the abbreviation, it is a Integrated Pastoral Development Programme. Well, better to tell you how it is working in Afar. It is community based, well the main beneficiaries is the main community of Afar, around Aba'ala … It is working with the people, working on social affairs, especially with women…" (Member of the IPDP board).

That the IPDP is based on the community's needs is thoroughly articulated by the project's staff. MU annually provides DF with a new application for the subsequent year, as part of the partnership agreement and its guidelines, "…and each year we sit and discuss with the people themselves about their basic needs", a board member explains putting emphasis on 'the people themselves'.

Robert Chambers, who promotes a participatory approach, argues that the language of development rhetoric changes fast, and that development practice lags behind language. Some words lapse into history, while other persist and prevail whatever happens to the field's reality and development practice. According to Chambers, participation, which got its renaissance in the 1990s, is among those words. "So widespread is its use that some talk of a paradigm shift to participatory development" (1995, 30). The conceptual development apparatus consists of lots of words that are widely used among developers; still these conceptual representations are merely on the linguistic level, and the practice and policy these terms represent 'lag behind the reality'. The IPDP documents state that the project is community based and applies a participatory approach in planning and implementation. Later this notion is discussed through scrutinising the HIV/AIDS prevention initiatives and women in development component.

'Bottomless development?'

Chambers argues that 'participation' is largely used in three various ways. The first use is merely as a cosmetic label to make whatever proposed appear good due to donor agencies' and governments' requirement of a participatory approach. The project's designers state that a participatory approach is applied in the planning and implementation of the project. Still, the reality and practice are a traditional top-down approach that comes with the development apparatus' tool-box. Secondly, participation is used to describe a co-opting practice to mobilise local labour in order to reduce costs, meaning 'they' (the target group) participate in 'our' (the developers') project. Thirdly, the use of participation refers to a process of empowerment that enables local people to make their own analysis, to take command, to gain in confidence and to implement their own choices and decisions. This means that 'we', the developers, participate in 'their' project. The IPDP, according to the project documents, seeks to apply participation as a process of empowerment. This reflects Chambers' third range of use where the project staff is merely supposed to function as facilitators for the beneficiaries. Chambers argues that such an application of participation, which is reflected in the formal order of the IPDP, is the key to reverse the power relations between beneficiaries and development agents.

As shown in the examinations above, local reality diverges in many cases from the stipulated formal order. Despite the fact that the IPDP planners' intention is to empower the beneficiaries, the way the development apparatus functions practically rather resembles Chambers' two first conceptions of participation. This is largely because practice lags behind language, the development agents' methods to plan and implement projects, and the back-donor's requirement for results. The postponement of the river diversion illustrates the problems of actually applying participatory approaches regarding planning and implementation. As it took the target group two years to decide where to build the diversion, the IPDP and the project manager were exposed to pressure from the back-donor, NORAD, who expected to see results. The inclusion of participants and beneficiaries created problems for the IPDP, because there were no signs of progress or results. In this respect, the implementation of the formal order, i.e. the use of a participatory approach, produced negative effects, as it was difficult to explain why the activity had been postponed since they had received funds. This dilemma underlines both the ambiguity of participation and the (back-)donor's requirement of evident results, and the problem associated with a strict implementation of the formal order.

DF's notion of participation also takes a different form. In pursuing a participatory approach, DF states that it is working with the 'real experts' to secure popular participation. DF's notion of participation refers to a close collaboration with the recipient organisations in the south, as representatives for the beneficiaries and recipients. The participants, i.e., the organisations, are supposed to develop their own plans, and to decide the content, division of labour and how and who to implement the project.40 MU and DoA, both governmental organisations, are not merely recipient organisations but also participants and beneficiaries. This makes sense regarding DoA, since one of the IPDP's goals is to 'strengthen the local administration in Aba'ala'. Regarding the '40 women trained in vegetable production' and the '25 pastoral people educated as barefoot veterinarians, it is however somewhat more questionable. This shows that participation is applied in different ways, both formally and practically, and that participation not only refers to 'the empowerment of local people' as is the general notion, similar to Chambers' (1995). What, then, is the range of use of participation when the IPDP is referred to as a community based project?

Participatory empowerment or cosmetic representations?

Regarding the inclusion of the target group in the project planning process, an IPDP board member states that "[e]very year we sit down with the people discussing their needs, what they want. They always come up with a lot of stuff that we cannot do, because it is not in the frame of the project". He mentions a variety of proposals whereas some as good initiatives and some not according to the project's main objectives. He expresses concern over the multitude of suggestions and that 'they don't know our budgetary and thematic constraints':

-Member of Board (MoB): …because we couldn't accommodate every activity that was proposed, we had to prioritise. So then we [MU] prioritised and together with Department of Agriculture decided which activities to be implemented. The people suggest several things, but we decide what of the suggestions to implement. So this is based on the needs of the people.

-Me:So all the different components and activities are based on the demands of the people, or…? How can it be based on the people's choice if you, MU and DoA, choose the activities?

-MoB:Yes, it is based on the people. Well, what we do is in July we have a small gathering, sort of brainstorming. We discuss so many things, so many issues, which are related with local development. And after discussion, we finally screen which activities we shall prioritise. And based on what? Why do we need to prioritise this activity? Justifications which come from the officials, from the experts, from DF, from local level etc. So based on such discussion we finally screen those activities that shall be implemented for that particular year. And then in July we finalise this, and in August we send, we write a full proposal and send it to the DF.

-Me: Does the DF always approve the different activities, or?

-MoB:Sometimes they say no, and then we have to cut. Because not addressing the main objectives. And sometimes they don't approve the activity, and sometimes it is due to budgetary issues.

The quotation underlines that the IPDP is not a participatory project in the sense as an empowering process that enables local people to make their own decisions where the developers participate merely as facilitators. Such a use of 'participation' implies a total shift in power, giving all the power connected to project planning and implementation to the beneficiaries. Chambers, who defends and promotes this notion of participation, states that it will imply a total shift from what he calls the paradigm of things to the paradigm of people.41

"Top-down becomes more bottom-up. The uniform becomes diverse, the simple complex, the static dynamic and the controllable uncontrollable. The future becomes less predictable. The transfer of packages of technology is replaced by the presentation of baskets of choice. Most difficult, the paradigm of people implies the third meaning or use of participation, an empowering process, with a shift to power to those who are local and poor" (1995: 33).

Regarding the IPDP's formal order, it adheres to Chambers' notion, but not in practice. That is not only due to practical reasons because of the many difficulties it would imply to consider the various local particularities, but also because it would challenge the development apparatus. Chambers' characteristic does not take account for the span between the wish for a participatory approach and not being able to implement it. The concept of participation is also dependent on the stipulated objective, since recipient organisations can also be the participating beneficiaries. As the case with the river diversion shows it is also problems aligned with applying a participatory approach, since it makes the implementation process slower and thus more difficult to present clear and effective results to the back-donor. The inclusion and presence of the HIV/ AIDS and women in development components in the IPDP illustrate aspects regarding the lack of participation and the importance of adhering to the formal order of development.

The introduction of HIV/AIDS prevention initiatives

The introduction of the HIV/Aids prevention initiative illustrates the informal relations between different institutions involved when new project components are included and designed. Autumn 2001 I was hired by DF to classify aspects of the different DF funded development projects, which is required in point 1.7. in the standardised NORAD application form. Point 1.7. is a requirement from the Development Assistance Committee (DAC) of OECD and forwarded through NORAD to Norwegian recipients. DAC is a kind of donor club where the 18 most prominent aid contributors from the western nations, who all have established their own self-contained aid bureaucracy, hold a seat (Hancock, 1989: 45; DAC, 1992). The statistical performance of point 1.7. is regularly referred to as 'DAC classification', and is used by DAC/ OECD to guide and monitor member countries' development portfolio in order to promote policy coherence. DAC members are obliged to "…adapt their current practises against [DAC's] standards" (DAC, 1992: 5). Point 1.7. enables the responding NGO only to give quantitative information in order for DAC to have legible units possible to measure, and requires information about aid form, which DAC-sector the project applies to, as well as a classification of how the project addresses DAC's different target fields. Until 2000, the target fields were research, women and environment. From 2001, human rights and HIV/ AIDS were included. Additionally, NORAD, in accordance with DAC, elsewhere in the application format added questions about HIV/AIDS and Good Governance, which were to be elaborated about in a qualitative manner.42 Hence, all the DF's partner organisations were approached and requested to give input about their existing work on HIV/ AIDS, human rights and good governance. For the IPDP, this resulted in the inclusion of the HIV/ AIDS component,43 while stating that human rights were promoted through the capacity building activity, though not emphasised directly. Good governance and corruption were addressed through institutional building of the partners, which both are governmental organisation.

In MU's final project proposal for 2002 submitted to DF, MU addresses the issue of HIV/ AIDS stating: "The [IPDP] will contribute to the control of HIV/AIDS by involving itself in the local initiatives already undergoing in Aba'ala by the Wereda Health Office and the youth".44 Despite this, DF designed and included the HIV/Aids prevention initiatives component. Due to lack of data it is difficult to state whether the inclusion of the HIV/ AIDS component was the explicit wish of DF or something they felt they needed to do to obtain further funding since DAC and NORAD defined it as a target field. Still, it is important to take into account that combating HIV/ AIDS is prioritised by NORAD (and DAC) and "…all Norwegian NGO activity are in total supposed to build up under the pillars of NORAD's activity", as a representative for the Norwegian Embassy in Ethiopia put it. The director of NORAD states in NORAD's Annual Report for 2001 that "[t]he work against HIV/ AIDS was escalated and emphasised in 2001". This was largely because the Norwegian Minister of Development at that time emphasised and prioritised work on combating HIV/ AIDS. This gave inducements for implementing organisations to focus on these issues, not only because organisations need to build on national policy, but also because prevailing policy largely defines what to be funded. This challenges the use of participation and shows the practical lack of the rhetorically entrusted participatory approach in the planning process.

Integrating women in development?

As participation, 'women in development' (WID) belongs to the development jargon and refers to the strengthening of the position females have in their society. NORAD defines women in development, equality and gender issues as a target area. NORAD's annual report for 2001 states that 'women are an important target group for NORADs development cooperation', and that 'NORAD's support to equality of status regarding male and female are given through special efforts concerning women and by integrating gender issues in general in the cooperation initiatives'. 'Women' is here defined as an object for DAC classification, cf. the application format point 1.7.

During the initial stage of fieldwork, a board member notifies me about Aba'ala and the IPDP. As we walk around in Aba'ala he tells about the project and the local community. He interrupts himself in his elaboration to state that it happens that NGOs come to Afar with a pre-designed project ready to be implemented since '…they find the area interesting as well as rather few NGOs are working in Afar since it is regarded difficult and thus prestigious to work in the area'. Further he says that these NGOs usually come with their own ideas and projects, which they want to implement, and almost exclusively they have a component to address women. 'But that is not what the Afar society needs. They are not ready for those European values yet. The basic needs of the Afars are health for themselves and their livestock, and water for consumption, irrigation and animals', he adds. Hence, he states that some NGOs have been dismissed by the Afar society to work in the area due to the NGOs' intentions of promoting gender issues on the cost of what the Afars themselves identify as their constraints.45 The board member's scepticism to the women component is reflected later on in an interview when he is asked about the profile of the DF:

"I don't have any deep knowledge about our donor, the Development Fund, but one thing is that I think they are interested in gender issues. When you see some of the activities, not only DF, even other donors in general are eager to promote gender issues. It's good, but you can't bring a change on gender issues. Gender issues can be addressed through a process, for example, if you want to empower women, that empowerment should come through a process, not by implementing an activity. Addressing separately gender issues, that is not a good thing, in my understanding. I believe that this is reached through other activities, and doing it separately doesn't provide anything, that's what I think. But we also have to do that, work on gender issues, since DF wants to do it. We want their money, and they want to address gender issues. … I think it is too early to do that to address such issues, it should come on a later stage. Aba'ala needs other development first before it is appropriate to address such issues, it should come at a later stage"

The board member discusses the presence of some IPDP components, and states that there is no need to have an independent women component since women are addressed through other project components. "The water development is such an activity. It secures better access to water, ponds and cisterns and so on, and therefore women don't have to walk six hours each day to fetch water. This is a component that addresses and integrates women. We don't need a separate women in development component". Since gender issues and women are approached through other IPDP components, it is reasonable that the presence of an independent women component serves to make it obvious that the IPDP addresses issues targeted by NORAD. Since NORAD's main approximation to a project is through project documents, these documents and the project's formal order need to be explicit and legible in that NORAD's policy and priority areas are addressed. Representations and simplifications are thus not only means to plan and to establish causality between investments, planned activities and expected results. They also serve to generate documents that through their aesthetic structure feed into a legible form that enables NORAD's executive officers, which are not familiar with the specific project or area, to get an understanding of the project they appraise for possible funding.

Summarising remarks

Various project documents manifest the formal order of the IPDP. The formal order of the IPDP is based on representations of both the field and the practises used to implement the project. This is done to arrange the field and the project as an object consisting of legible units possible to plan on and intervene on. Legible units enable development planners to identify and establish causal relations between the different aspects recognised as important target fields in the project area. As shown in this chapter, there are many gaps between the formal order and what the formal documents set out to describe, both in terms of the socio-cultural aspects, and in terms of the codified practices of project management. These gaps regard both socio-cultural aspects and how the project is actually run. Particularly important is the concept of participatory approach and its implications. Though the IPDP is presented as a community based project based on participation, this chapter shows the difficulties of implementing participatory approaches since they challenge the formal order, and because many guidelines and policies come from the donor and particularly the back-donor.

The next chapter shows how development agents relate to the formal order of the IPDP and illustrates occurrences of interface, which again generate countertendencies. With a focus on actors' agency, the chapter illustrates how local development agents bridge the complexity of local practical knowledge and its formal representations.

Chapter 4. Agency and development discourse

The formal presentation of a development project is necessarily simplistic and reductionistic towards the diversity of the multiple realities it sets out to depict. The project's formal order relies upon representations. The formal project documents stipulate not only what to do, but also the strategies and methods to reach the stipulated objectives. This chapter treats the issue of how development agents relate to the formal structures and order of development promoted by the donor organisation. Focus is put on development agents and how the intersection of expert discursive knowledge and local practical knowledge is communicated by them in various settings. The chapter shows that local informal strategies evolve in relation to the formal structure of a project and the donor-recipient relationship. Whereas the previous chapter showed the representations and formal order of the IPDP, this chapter shows how local development agents cope with these representations practically. Informal practices are means of bridging the gap between project management and the formal codified representations of these practices and knowledge. As will be shown, the viability of the project largely relies upon the informal practices of local development agents even though the practices deviate from the formal order. In this respect, development projects exist on two different but interconnected levels: A formal level articulated in the various documents of the project that are sent between the donor and the recipient, and on an informal local level, which can not be accounted for in the project documents.

Informality within the formal structures

The term informal relates to what is defined as formal. Consequently, what the terms represent do not exist as independent entities. In this thesis, the term informal corresponds to local practices and knowledge that the formal representations of the Integrated Pastoral Development Programme (IPDP) fail to address. The previous chapter shows that development intervention relies upon legible and tangible units. In Seeing Like a State, Scott states that "[d]esigned or planned social order is necessarily schematic; it always ignores essential features of any real, functioning social order" (1998: 6). This chapter describes some of the informal practices that maintain the local social order. These practices are not accounted for in the planned formal order. The IPDP documents not only define the representations and simplifications of the field the project relies upon, but also stipulate how to plan and implement the project. The partnership agreement and donor-recipient relationship largely define how the project is to be run. These formal guidelines on how to operate the project are schematic representations of what is expected to take place and thus the informal is everything that is not included in these formal abstractions. Informal practices relate to and must be understood in relation to what is defined as formal. The informal practices might sustain, support or infringe the formal order.46 Thus, formal order is dependent on the informal practices defined in relation or opposition to the formal. "Formal order, to be more explicit, is always and to some considerable degree parasitic on informal processes, which the formal scheme does not recognize, without which it could not exist, and which it alone cannot create or maintain" (Scott, 1998: 310). The omnipresence and importance of informal processes and strategies are well illustrated by a French form of action called 'work to rule': The employees carry out punctilious what their job descriptions state (which is a formal abstraction or representation of their work) with the result that the productivity comes to a stop. "The workers achieve the practical effect of a walkout while remaining on the job and following their instructions to the letter" (ibid.). This illustrates that formal order depends upon a variety of informal practices and ad hoc improvisations that cannot be codified or formalised (Scott, 1998: 8, 310; Nustad, 2003a: 213). Nustad (ibid.) argues that informal practices can take two forms: Either as the unspecified content within the formal order that contributes to its realisation; or as practices that can sap attempts to realise the goals or objectives stipulated in the formal order. This chapter illustrates the former.

Informalities are defined in relation to formality. Informal social practices and knowledge, cf. what Scott calls metis, are implied necessities of any formal order due to the impossibility of accounting for the multiple realties when codifying the formal representations and order. Development cooperation relies upon a codified and stringent social order of legible units in which the informal is defined in relation to. Project documents define the formal order of a project, while local knowledge and practices that are not included in these formal abstractions of the project and its target field are classified as informal.

Informal knowledge and practice can take the form as specific ad hoc reactions to the formal order or it can be manifested in or articulated through informal strategies. Informality is characterised by its spatial and temporal particularism. Informal practices oppose universality, which can be codified. Hence, informal practices can contribute to bridge the gap between imposed social order and local practical knowledge, as in the case of development intervention such as the IPDP. Identification of informal strategies requires an actor-orientated approach. This illustrates the situations of interface between the formal order of a development discourse and local practical knowledge. Since informal strategies are based on a variety of local practical knowledge, it is not possible a priori to depict the outcome of such interfaces. Thus, it becomes an empirical question.

On the project; off the record

The three cases that are presented below all show how various development agents relate to the formal order of the development project they are associated with. A central aspect of all cases is how the actors act towards the formal order and the system of knowledge articulated in the various project documents.

At a meeting held in Mekelle between the partner network of the Triangular Institutional Cooperation Project,47 the Development Fund, as the donor, informs that the DF partnership model is to be evaluated. Two consultants from Norway are to visit their organisations in Ethiopia and India to evaluate 'the partnership with DF as seen from the South'. DF's partner organisations react and say it is senseless due to the vast amount of money that will be spent on salaries and travels. Rather distended they argue that 'you have the same documents in Norway as we do. We do not know anything more than you. We only do what's stated in the documents and in the partnership agreement. Nothing else. We suggest that you rather spend that money directly on projects to help the poor'. DF agrees about this, but argues that the evaluation of the partnership model is a requirement in their frame agreement with NORAD. DF's partner organisations show great faith in formal planning, which they claim they follow. "We have signed the partnership agreement with DF because we agree about it. Since we agree we are not dissatisfied, and we do not do anything that is not stipulated in the agreement" an IPDP board member and official from MU informs.

Such statements illustrate the emphasis given to formal order in a formal setting. Regular daily practices, however, show a different attitude towards the formal order. The three cases below illustrate informal strategies that evolve in relation to the formal structure or order of development cooperation. These informalities are not reflected in project documents, and illustrate various informal practices in project management. As will be shown, informal practices largely relate to the formal order and contribute to reproduce this formal comprehension of the project through the documents that are consigned to the donor.

A strict plan, flexible implementation: the case with the donor representative

The IPDP was first initiated in 1998 as a pilot project before it culminated into a regular project48 the subsequent year, with a five year phase through a NORAD approved application. 2002 is the last year of the first phase. The project planning stipulated practices, components and activities to achieve the defined objectives. Beginning in 1999, funds and strategies to reach the specified goals were secured and defined, at least in principle. The donor (DF) and their back-donor (NORAD) require annual applications and progress reports from the IPDP. The local project manager often utters frustration in relation to this:

"I don't see why we have to apply each year. A program means long-term funding. I thought that when the application for the five-year phase was accepted we could do our work. Instead, we are forced to write new applications every year, it takes long time, and new objects are put into the project that was not there in the beginning. What's the point of a five year plan if we have to apply each year?".

Several other persons connected to the IPDP address the same problem. The program manager also raises the issue to the visiting evaluation team when he is asked about any constraints in the partnership model with DF. In response, the project manager is told that it is a requirement from the back-donor, from NORAD, and that 'this is how it is, and you have to follow the guidelines, just as DF must with their frame agreement with NORAD'.

In late February 2002, a representative from DF visits the projects it funds through their local partner organisations in Tigray and Afar. The DF representative reserves one day to follow up on the IPDP. The day starts with a breakfast meeting at a local hotel, where the donor and the local project manager discuss the progress of the IPDP and earlier submitted reports. The donor argues that the actual implementation of the project is not up to date with the defined plan and strategy, and that the way the budget is managed and allocated so far has been 'a bit suspicious'. She informs that there is too much money left on the budget from last year. All the money has to be spent in the current year, according to the budget, so 'we show NORAD progress and the need for more funds". The project manager replies that the delay of some of the project's activities is due to two reasons: First, that the car was broken 'so we had to wait for a spare part ordered from Japan', which made communication and project follow-up hard. Secondly, the project manager argues on the economy and that he faces difficulties in controlling the project's funds, because at Mekelle University (who manages the project's accounting) everything is on one account which he does not have access to. This makes it difficult to keep trace with budgetary issues'. Later they discuss how salaries are disbursed. The amount of money given to one of the local field assistants in Aba'ala is questioned. The donor asks "Is this legal?". The project manager replies that "It is approved by the board".

After the general briefing on the IPDP, the three of us leave for Aba'ala for the field visit. The drive takes about an hour and a half, and there is plenty of time for further briefing and discussions on the current and future situation of the IPDP, yet in a more informal manner. The donor expresses understanding towards the project manager:

'I know the difficulties you face, but this phase [year] of the project is very important, and we have to be strict with the implementations because we need to show NORAD that the money passed on to this project is not used carelessly, that we fulfil the application, and that the DF has a good local partner in MU to do this. It is important to secure further funding. So far I've been quite flexible with the project implementation. I know the problems you face; the lack of capacity, difficulties to implement the project because of the people's lack of insight of the project, the budgetary constraints, the importance of a proper car. But DF can't help you with all that, but instead I have been flexible what regards the accomplishment of several of the activities. But now I don't have the opportunity to be flexible any more. We're in the last year of the first phase, which means that what we haven't done, or postponed so far, will have to be implemented by the end of 2002. Or at least before March next year when we submit the project report to NORAD for evaluation. If we want money for a new phase, the least we can do is to show them that we have implemented all planned activities, or show them that the activities has started. But again, I know the local conditions and the difficulties it implies regarding the implementation of the project, and have so far seen it as a necessity to be flexible'.49

The project manager agrees that the local conditions with its shortage of infrastructure and lack of popular awareness demands flexibility in project implementation.

'As you know, they are not used to projects here. They are not used to working in the same way as we demand for a good collaboration and project. The Afar people are rarely exposed to outsiders, and are quite sceptical to change. Therefore, it is important to try to meet the people. To let them know that we do this for them. But it is hard. They don't have the same attitude of working habits, and all the local institutions have a lack of skills and capacity. The whole idea of projects is rather strange for them. The IPDP has a budget of nearly 1 million birr, while the local governmental office has an annual budget of only 40.000 birr. They don't know about project planning, structure and implementation, and even less the requirements of doing things on time as required in the project application and documents. Therefore the flexibility is important for the implementation of this project'.50

With reference to the local context, both representatives emphasise the importance of flexibility in the practical project implementation in relation to the formal stipulated project plan. The DF representative states that 'if it weren't for the flexibility, the project would never have come so far as it has today'. She continues: 'I have always been flexible with the IPDP. I know the conditions here very well. But I also know that the Afar people are among the poorest of the poor and heavily affected by famine and draught, and therefore it is crucial that we manage to get funds for a new phase". Rather apologetic the DF representative says that the implementation pace of the project's activities must be increased in order to implement the previously planned activities. She argues that 'it is the demands from our back-donor that trigger this. The contractual guidelines state that we must write a detailed report to show all the activities undertaken by the project during the first phase.51 This has to be in accordance with the previous applications, as well as the initial stipulated goals'.

The same day of the project visit, the paravet-training of the IPDP takes place in Aba'ala. After a stroll at the nursery site, the project's staff and the visitor from DF approach the paravet class who sit in the shadow outside the veterinary clinic. On the way towards the class, the project manager, who is rather elated by the talk and acceptance of flexibility in project management, expresses that he supports this notion of flexibility. He says that 'the flexibility is what makes the project accepted by the local people', and underlines that there are many formal constraints in the project's bureaucracy that he has to cope with and that it's good that DF knows about this. 'It makes it easier for me to manage the project and to explain the DF about any inconveniences or deviations with the IPDP'. He stresses the differences in culture between highland and lowland Ethiopia, informing that the Afar people are not used to act according to an imposed plan and that they initially 'had to be talked into the project, and that is where flexibility becomes important'.

The instructor of the paravet training welcomes the donor representative. The class consists of 25 people, whereof two women. Some of the participants have walked up to 200 km through semi-arid desert in order to attend the course about basic veterinary issues. In the welcome speech, the instructor portrays the DF representative as the one who 'makes this possible' for the participants. The donor questions the partakers about their experience of the training. Everybody seems to appreciate and acknowledge the expertise they have gained, though they are a bit reluctant to answer the donor. In the final question, the donor asks whether the class have any complaints about the training. Suddenly, the people seem more interested to give answers: 'We never get the chance to have a tea break or to go to the toilet. / We've been here for a long time, even longer than planned, and we have to go back to our livestock, to our families, and to our general occupation. / We were told to stay here for a certain time. It's way overdue. / Some must go home, but the project staff are strict and keep us here'.

The program manager is later confronted with these responses in relation to the notion of flexibility. He argues that 'if we are supposed to be flexible in every matter we would not accomplish anything. Sometimes we have to be strict with the participants and beneficiaries. The delay in the training is caused by lack of personal, economical and medical resources. If we let people leave the course, it will be unsuccessful and halfway for both parts, and since some live far away they would never return to finish the course later'. He draws a parallel to the planning and construction of the river diversion: 'It took us two years to start on the river diversion, because we let them, the beneficiaries, decide themselves where to put it because of the conflict between up and down streamers. If we are supposed to implement all activities defined in the project documents, we also need to be strict. And now it will be a tough time due to the demand from DF and NORAD to implement all the planned activities. It will be tough and hectic'.

The subsequent day of DF's visit to the IPDP project area, a workshop is held in Mekelle on water harvesting issues in relation with the Triangular Institutional Cooperation Project (alias 'the Triangular'). Participants include DF and MU, among others. None of the IPDP staff participate, but some high officials from MU, who also sit in the IPDP steering committee, are present. One of the MU officials monitors and coordinates the overall development related activities that MU is engaged in. In one of the workshop's intermissions, this MU official invites the DF representative and me for coffee.52 This coffee break evolves into an informal meeting, addressing the status of the IPDP and further funding. The donor repeats what she has said earlier to the program manager, yet in a different manner.

'I'm trying to get the project [IPDP] back on track formally. So far, I've been very flexible with the project due to the special social context in Afar, where you have to deal with local informal structures. But now, due to the fact that this is the last year of phase one, I have to get the activities and budget back on track. We are able to reallocate money between different years within one phase, but now we start the initial planning for phase two and therefore we have to show NORAD results and that we are on track. It is approximately 400.000 birr left from last year. I've been flexible, and that has been right, as well as important for the project. But now I'll have to press this through due to the new phase coming up, as well as there is an external evaluation of DF that NORAD requires, and therefore it is important to show the right signals if we wish further funding'.

The donor representative informs that DF will assist the IPDP and MU to design a new application since DF knows what NORAD's priority areas are, which is important to address to secure further funding. It is argued that a component regarding good governance is introduced, and that the work on combating HIV/ AIDS is enhanced, since both areas are prioritised by NORAD. The MU official responds with gratitude and says that the most important is to secure further funding 'because the people in Afar needs it, they are the poorest of the poor, and it is good we get help in the application process. We have to convince NORAD that the project needs money, but the only way to actually do that is for them to come to see themselves. But they never do. So we have to make a good proposal'.

'He knows the NORAD format': the case with the consultant

In May, the project manager prepares a 'feasibility and identification study' to gain information to be used in the IPDP application for the second phase. He takes the idea of participatory approach serious, and makes a lot of thorough and extensive arrangements. He took over the job as program manager right after the first phase was initiated and has never produced such an extensive application before. He applies the formal stipulated guidelines for planning, and "…because this project is community based, I need to know what the beneficiaries regard as important to make the application as valid as possible".

On the day of departure for the fieldtrip, which is planned to last for five days, the project manager arrives late to the meeting point wearing his ordinary suit, which is rather inappropriate for a field trip. He informs that the trip is cancelled and that "[w]e have to wait for a consultant that DF sends from Norway. They don't want us to do it our self, or, they say we need help". The reason given by DF for sending a consultant is that "[h]e knows the NORAD format".53

The project manager is upset about being run over in the decision-making process, and argues that there will not be enough time for the fieldtrip and survey if he is to wait for a consultant to arrive from Norway.

"If we wait with the fieldtrip until the consultant has arrived, we won't have time for it. We are soon expected to submit our application to the DF, and the involvement of another person is very time consuming".

The project manager asks what there really is to know about the NORAD format. He argues that the application form is well arranged with several open-ended questions only to be filled in with the information gained from the beneficiaries. He also recalls the words of the DF representative a few months earlier who informed about the application process. She stated that "[i]n order to make it easier, it is only to copy from the application we made for the first phase. Much of the socio-cultural elements are the same, which would make the application process easier".

After heavy delays, the process around the new application starts as the consultant arrives from Norway. DF has employed him as a consultant and external expert on several earlier occasions. The project manager questions the consultant whether they will manage to meet the application deadline or not. The consultant assures they will make it. He states he has experience and skill with such formats, and that "…everything we shall do and produce is stipulated in the Terms of Reference I've brought from the DF. We'll manage this".

At the first planning meeting, the project manager again expresses his concerns: "The IPDP is a bottom-up project, we need to talk to the community in order to identify their constraints and the activities needed to meet those constraints. It will take us a long time and hard work". The consultant replies that the most important thing is that NORAD is convinced to give further funding. "We need only to write that we have talked to the community, and they believe we have done it. Talking to the whole target group is too exhausting, and I guess that the community opinion about what to do has not changed much since the study prior to phase one". This statement surprises the project manager and he underlines the importance to assess the needs of future beneficiaries, "especially the people in the new areas we're targeting".

The divergence between the project manager and the consultant is settled through choosing a middle course. They decide to go to Aba'ala and talk to the representatives for the community, i.e., the elder council and employees at the governmental offices and other community leaders. The project manager objects, stating that these groups not are the primary beneficiaries, rather people that already have employment and live under relatively good conditions compared to the average of Aba'ala. The consultant argues that it is impossible to apply a literal understanding of participatory approach and bottom-up planning due to the extensive workload it implies. He also refers to his experience in designing projects and applications as a safety valve. When they have agreed to go to Aba'ala for one day to talk with community leaders about the IPDP, the consultant says that 'it would be good to go to Afar. According to my ToR [Terms of Reference] I'm supposed to have discussions with the Department of Agriculture and other institutions in Aba'ala'.

A crucial issue for discussion with the local institutions in Aba'ala concerns how to design the partnership agreement. DF has expressed that they would like MU to withdraw as a formal partner of the project and rather be hired as a local consultant to assist the Department of Agriculture. "This is a wish from DF's side, but I think it is impossible", the project manager states. He argues on DoA's lack of institutional capacity, manpower and knowledge on general project management. On this issue, there is a discrepancy between the consultant, who represents DF's case, and the project manager. The former works according to ToR defined by DF. The ToR states that the consultant shall "[a]ssist in developing the application, in general, based on the inputs given by DF, and specifically assist in developing activities, goals and indicators for the new project phase".54 The project manager states: "Well, you just have to go talk to them [DoA]. I have, and they are perplexed about managing the project all alone. They told me they wouldn't manage it. And they don't have the institutional and technical capacity to implement and supervise the project".55 The consultant answers that it is up to him to make recommendations to the DF, and that he will talk to the DoA about the matter.

In a meeting between the consultant and the project manager, responsibility is distributed about who should write what in the IPDP application. One of the last points in the application form for new projects (as defined by NORAD)56 is "[h]ow is the project to be financed after Norwegian support has ended?" (point 3.5.). The project manager claims that 'it is improbable for the project to be financed by others, and even more impossible that it would be self-sustainable. The IPDP budget is on nearly 1 million birr, while the Department's [of Agriculture] annual budget is only 40.000. There's nobody to finance the project except DF and NORAD'. The consultant replies:

'Well, of course, we cannot say that we are dependent on the funds from an external donor. The application for the first phase did so, but we can't do it once more. Again, we just have to convince them that the project is sustainable, and that the local institutions will take over the responsibility for the introduced infrastructure. We write that through lifting the capacity of the local institutions and creating awareness, the people in Afar will manage the project themselves. I don't think the second-phase application will be accepted if we say that we are dependent on further external funding after the implementation. And we have to write that it is cost-efficient.'

The consultant informs that it is easier to fill out the application if 'you know what they want'. He says that he promotes the inputs given him by DF, who knows the trends and policy of NORAD and the Norwegian Government.

The REST case

Relief Society of Tigray (REST)57 is the biggest local NGO in Tigray, and has been working with DF since 1982. REST is the biggest single partner of DF in terms of funds and donations. Employees at DF often refer to REST as an 'extremely professional and competent NGO on all levels'. REST and DF admire each other mutually, but the head of REST's planning section acknowledges the general structural and economical differences between donor and recipient in terms of donor's policy guidance. He argues that due to the donor-recipient relationship characterising the NGO sector, one cannot talk about an equivalent partnership since the donors always will try to promote their own policy and ideas. "But as a professional NGO, REST manages to cope with this. Even though the relationship is labelled partnership and we're supposed to be equal, we recognise that the donors have certain interests they try to promote". The planning manager further states the importance of raising funds, in which REST's activities rely upon. He states that due to the crucial need for funds, REST sometimes has to undermine its own ideas and policy in order to satisfy the donor, and thus get money:

'REST as an implementing organisation has to raise funds, which is quite crucial for our activities, to sustain them. Doing this we have to take different interests and issues from donor organisations into account. We want the donor's money, and usually they want to have a project, especially here in Tigray and in partnership with REST. There are many reasons for that. For example that REST is the biggest NGO in Tigray, that there is a common international understanding that Ethiopia and especially Tigray is poverty-struck, and that the Ethiopian government encourages organisations to work in Tigray. Also, international donors see it as prestige to have projects in Tigray. Tigray is still known from the Live Aid concerts. Still; we decide ourselves what money to receive, because if there is major differences between the donor and REST we are not interested working with them. But, as long as there are possibilities of coping with the donors' interests we might work with them even though not having the same interests and not pursuing the same objectives. REST has as a principle, that as long as we don't have to move away from our principles, we try to accommodate to the situation. REST won't go against their principles and interests. But if the donor comes with a frame for a funding plan, we operate within that frame to promote our strategy and ideas. We never let the donor decide everything. We agree on the framework, and define much of the content and activities to take place ourselves, within that framework'.

The planning manager expresses that there exist a lot of "global development words" as representations of policy, which the donors usually try to pursue in their partnership agreements with the recipient organisations:

'Participatory approach, women in development, gender issues, PRAs [participatory rural appraisal], poverty alleviation, empowerment, good governance, capacity building etc. These are all among the global development words. As I see it, it is not always in our interest to work on these issues, but requirements and guidance from the donor tell us what to do. A project is never fully drawn up, neither by the target groups or us. The donors are not only suppliers of money, but they also in some way tell us how to use the money'.

He explains that REST relates to these donor-promoted 'global development words' as policy frames which REST, dependent upon the partnership relation with the donor, tries to define the contents of. He acknowledges the donor's power to pursue their interests, but simultaneously REST's ability to independently manipulate and navigate within the donors' proposed frames and policy.

Whereas MU and the IPDP board members underline their independence and the equality in their relation to DF and other donors ("We're partners. Partnership means equality, cooperation and mutual influence and benefit" an IPDP board member states), the general notion of unequal power relations in donor-recipient relationships are acknowledged by all those interviewed at REST. All state that REST is dependent on its donors in order to carry out its development activities and acknowledge that the donors do not merely give the money away unreserved. They recognise the donor's role and that the donor would like to have some degree of control on how the money is spent, both in terms of the activities and projects that are given funds and how the budget and accounts are kept.

In an informal setting, an employee at REST informs that one of REST's projects, which is labelled as an 'urban poverty alleviation and empowerment of women'- project actually addresses rural farmers. The donor organisation's initial wish was to establish and fund a project within Mekelle. The REST employee explains that since REST managed to define the project's target area as the southern zone of Tigray (where Mekelle as the capital of Tigray is situated) instead of Mekelle exclusively, they were able to work with farmers and irrigation schemes outside the urban area. Water harvesting techniques and irrigation schemes among the farmers, and a training course for women in vegetables harvesting techniques became classified as an urban project in the documents, and thus the donor organisation agreed to finance it.58

Understanding off the record informallity

The three cases depicted above show different ways actors relate to imposed structures and guidelines from an external donor, and illustrate situations of interface articulated through various actors. The notion of interface challenges the formal institutional arrangements (Arce and Long, 2000: 11), and emphasises that one needs to identify the formal and informal coalitions that act for or against certain strategic representations of development. Interface studies aim to bring out the discontinuities that exist between different systems of knowledge and how this struggle between, or over, knowledge turns out. Studies of knowledge encounters show the struggles, strategies and interactions that take place. Such studies show how "…actors' goals, perceptions, values, interests and relationship are reinforced or reshaped…" by the process of interface (Arce and Long, 1992: 214). Interface gives focus on the diverse types of interplay and interaction between different knowledge realms. Parkin (in Arce and Long, 2000) presents the notion about 'counterwork'. Counterwork denotes the process that unfolds when different systems of knowledge intersect. Parkin conceptualises counterwork as the rebounding effect of knowledge in its diversity. Partnership relationships and the interactions between donor and recipient involve the interplay of 'hegemonic' and 'non-hegemonic' discourses and values. Thus, Arce and Long also give an account of Wertheim's notion about 'counterpoint', characterised as composed of 'deviant' values that, in some way or another, are institutionally contained. Wertheim's central point is that the "…dynamic processes of change can never be understood if the opposing value systems within society are not taken into full account" (Wertheim cited in Arce & Long, 2000: 11). In studies of interface, one must take the different realms that actors relate to and that shape their perception of everyday life. The analytical amplification of counterpoint values challenges the existing institutional arrangements. Arce and Long call for a combination of counterwork and counterpoint, which they call counter-tendencies. Counter-tendencies evolve in situations of interface. To identify counter-tendencies is a useful methodological approach to get to grips with different and entangled systems of knowledge in situations of interface and what these processes lead to. "Life-worlds exist as specific time, space and experiential configurations …, where some coexist, some clash, some mix, and others separate or retreat into themselves" (Arce and Long, 2000: 13). Thus, the encounters and interactions between the different systems of knowledge are centres to gain empirical data of the process of counter-tendencies.

In the first case, the representative from the donor organisation tries to help the local project manager to cope with the formal structures and requirements forwarded by the back-donor, NORAD. This is done as a means to secure implementation of the different activities. The DF representative states that flexibility is important in the project's implementation and that she tries to mediate between the formal guidelines and constraints of NORAD, and the local knowledge the IPDP project staff relate to and might share with the beneficiaries. The emphasis given to the notion of flexibility by the DF representative is because 'she and DF know the project area and that they are dependent upon the goodwill of the local community for their participation'. The project staff's call for flexibility towards the formal guidelines is a way to mediate between different systems of knowledge in order to promote smooth and contextual project implementation and to prevent local doubt and resistance of the IPDP. As the case shows, the flexible approach of the donor is converted into strict project guidance in the last period of the first phase, which ironically is due to the back-donors formal way of evaluating the first phase. The evaluation measures the IPDP's success and achievements in relation to the planned and stipulated objectives. The evaluation is an important decisive factor when NORAD is to consider whether to fund a second phase or not. In order to get the project 'back on track', and thus obtain a better evaluation, a lot of project activities need to be undertaken before the finalisation of the first phase. In this case, the formal structures and need for progress reports prevail over the flexible planning. Still, the project manager from MU speaks well about the freedom given by DF regarding flexible implementation, even though it involves a crucial change of practice. "Well, now we have to implement a lot the last months, but it is better to bother the local community once in a while than all the time" he says. "If we are supposed to bother them all the time, we would not have that much acceptance or success with the project. The project relies upon the participants' acceptance". The DF representative is apologetic. She has opened for flexibility to cope with the strict project formalism, but now has to make a call for rapid implementation and visible results, which in the end are supposed to be to 'the best of the project regarding the new application towards NORAD'.

In the case of the consultant, a different approach to the development discourse is depicted. The consultant's goal of further funding is the prime motivating factor, thus the MU project manager is rather overwhelmed by the consultant's approach. Because 'he knows the NORAD format' the consultant seems to discern relevant knowledge in order to handle the application process with the 'input given from the Development Fund', and thus he neglects some of the viewpoints of the IPDP's project manager. The situation of interface, or knowledge-encounter, is prominent in this case: Not only between the representations acted upon versus local knowledge, but also between the consultant and the project manager regarding their tasks and how to proceed with the application. The project manager started to work with the application, and prepared a questionnaire to secure a bottom-up approach and to identify the constraints and needs of the Afar people.59 The consultant is more interested 'to assist in developing the application based on the inputs given by DF', and thus to fulfil his mission, which is to make an attractive application for the second phase before the deadline passes. As a result of his purpose, he bypasses the bottom-up approach pursued and promoted by the project manager and the codified guidelines. This example offers an illustration of interface and countertendencies that evolve as the consultant bridges the contradictions within the formal order of development between the much wanted participatory approach, donor's goal of policy coherence and the demand of showing results. The case shows that the person closest connected to DF, which is the final decision maker before submitting the application to NORAD, has most influence over how to make the application. This is communicated through the consultant's repeatedly mentioning of his Terms of Reference defined by DF, as well as DF states that he knows the NORAD format. The project manager's knowledge about the NORAD format is scarce and he has never produced such an extensive application before, thus his knowledge is based on his perceptions of the field and on the formal presentation of the application process. His actions relate more to general development rhetoric than to an established discursive practice. The language of development rhetoric changes fast, and development practice does not mirror its linguistic representations. This is reflected in the discrepancy between the project manager and the consultant. Whereas the former largely relates to development language, the latter's practice diverges from the rhetoric.

The third case, which concerns the Relief Society of Tigray (REST), shows an NGO that is more reflexive towards donors and the development discourse. The case shows that strategic manipulation of the development apparatus relies upon the degree of reflexive knowledge about the development discourse and formal order. As a long-standing NGO, REST is deeper embedded in the development discourse than MU, but their position as recipient organisation and not as representative or bearer of the established development knowledge enables REST to be reflexive towards it. REST acknowledges that the donor-recipient relationship implies differences in power between the two implied organisations, and that the donor's power exceeds the recipient's, even though it is supposed to be an equal partnership. REST argues on the inevitability of bottom-up project planning in the literal and formal sense. It has certain coping mechanisms to adjust to this problem in the selection of partners and in the planning process. Through acceptance about the general frames, which are more or less defined by the donor, REST tries to define the activities and content within these frames. REST's possibilities of manipulation are also shaped by other factors; mainly the structures of power between REST and some of its donors. REST is about 50 times the size of e.g. DF. REST is also generally regarded as a star example of a partner organisation that is highly clever and professional. In many ways, donor organisations are dependent on REST because it is important for a donor to have good recipients. As shown above, the project's representations do not always reflect the project or activities that take place among the beneficiaries in the target area.

'What is there to know?'

When the IPDP project manager gets the information about the arrival of a DF appointed consultancy on the reason that the consultant knows the NORAD format, the project manager almost exclaims "what is there to know". He argues that all the information he needs to produce the application himself is defined and stipulated by NORAD in the application form, ready to be filled in.

The three cases depicted above all show that in project planning and implementation, local project practice differs from its representations as stipulated in project documents. Development interventions formally approach the target area on the basis of project documents, applications and reports, and thus the production of them are among the most important and crucial aspects in the non-facial relation between donor and recipient. The documents constitute the project and are what the (back-)donor relates to regarding project assessments and evaluations. The official project documents stipulate and define what have taken place (reports), what is planned to take place (applications), information about the socio-cultural context and target group (in reports and applications), as well as what, how and who to implement the project (applications). As the three cases show, informality (i.e., the knowledge and practices that fall outside the project's formal order) is necessary for a project's capability to be realised, since the formal order codified in documents are simplistic in their nature towards the complexity of local practical knowledge. In order to manage a project, informalities are of paramount importance, which relate to the undefined practical knowledge that bridges the gap between two different types of knowledge.

'Cultures of formality'

Each of the persons in the three cases refer to different types of formal project documents when they outline the project, or when they legitimise development actions and interventions. Project documents are the core of development projects, whether it is a consultant's Terms of Reference, an application, a strategy paper, project reports, or progress reports. A consultancy report, which offers a third party's 'objective viewpoint' on the project and its related issues, is of cardinal value to the donor. Consultants are widely employed by development organisations to evaluate ongoing projects, or to give recommendations and input to new applications.

In "Cultures of Consultancy", Stirrat argues that the production of consultancy reports is a 'cultural practice' and the documents produced are largely judged on aesthetic criteria. Stirrat writes: "…in practice the pragmatic impact of their work is in many ways irrelevant. How these consultants actually work, what they produce and the way their work is judged is much more a matter of aesthetics" (2000: 31). The production of reports implies to present a particular type of knowledge in a particular way, and thus what characterise consultants' and development agents' activities are that they "…produce a particular output that feeds into a larger whole" (ibid.: 34). Consequently, consultancy reports confirm and reproduce the existing knowledge, which fits into a larger whole, i.e., the development discourse. Though focusing on consultancy reports in particular, Stirrat's general focus is the products of consultancy work when they are hired by donor organisations to produce different types of project documents.

According to Stirrat, the aesthetic of 'good reports' have four characteristics, which generally resemble the IPDP documents. First, he argues on the structure of the report. A report starts with an 'executive summary'; the ambiguity of 'executive' being only too clear.60 Recommendations and paths for the future are given in a report. Everything is sorted out under numbered paragraphs, which outline what to be filled in.61 Acronyms (cf. DF, IPDP, MU, NORAD, REST, etc.) are given prominence. Secondly, the appeal to objectivity is a paramount common denominator of such reports.62 Words like 'perhaps', 'uncertain', 'unclear' and 'some' are as absent as the author is in the text. What is written about is distanced and objectified; doubt is not allowed. Quantitative data are preferred to qualitative information, thus making the field legible, tangible and easier to intervene in.63 The use of consultants is an appeal to objectivity through the involvement of a third and 'neutral' part. Thirdly, Stirrat argues on the appeal to rationality to be classified as a good report: The product of consultancies' work are written in a way that leaves no loose ends, and the analyses presented are based upon the assumption of systematic closure. The world of consultancy reports appears as remarkably tidy: Words like 'contradiction', 'disjuncture', 'fault line' and all other terminology of the contemporary academic language are notable by their absence. This reflects a belief in causality, a teleological approach and faith in planning.64 Fourth is the appeal to legitimacy. The conceptual development apparatus (largely articulated in donors' official policy guidelines) and acronyms present a sense of an official world. Despite the formal translations of the acronyms in the beginning of the report,65 acronyms function to 'officialise' the document while decreasing its transparency for those not familiar with the conceptual development apparatus. In the end of the document is an 'itinerary' that lists the people met and talked to, as a 'proof' to verify that the consultant has done his work.66 Sources cited tend to be other reports, especially those produced by the standard bearers of a development discourse; the World Bank and UN organisations in particular, but also the regional development banks.67 Academic works tend to be neglected. The result is the generation of orthodoxies and 'facts' which are recycled over the years, which produce new versions of old facts over and over (Stirrat, 2000: 40-43).

The products of consultants' work are characterised by representations. The representations are produced on the "…assumption that there is an objectively knowledge about the world which is understandable through the application of rational thought" (Stirrat, 2000: 36). Thus, the rather naïve and traditional but yet so important anthropological question need to be asked: "What on earth is going on here?" (Stirrat, 2000: 32). One needs to explore the representations, what they represent, as well as how they are produced, consumed, maintained and challenged, and most importantly the actors' role in this.

The 'good' documents of the IPDP?

Stirrat argues that:

"What consultants do is cultural activity. Although it is claimed that their work has pragmatic objectives, in practise it is judged in terms of aesthetics, judgement and taste. Consultants are cultural performers, cultural artists, whose product should not be judged in terms of its supposed practical ends" (Stirrat, 2000: 43, emphasis added).

The previously presented cases illustrate that Stirrat is partly wrong and partly right in his assumptions. To neglect the fact that consultancy work and their products do not have any pragmatic objectives is wrong. The case with the consultant shows that the pragmatic objective is to get more funds for the IPDP's second phase. The consultant's knowledge about the format in which the application is to be presented enables him to write an application with this pragmatic objective. The first case with the donor representative shows the interrelatedness between stipulated practical ends and that these had to be fulfilled, though initially being postponed. The three cases nevertheless all show the importance of documents and their presentation form.

The aesthetic value of a document that is to be submitted to a donor is important, and the project design and its aesthetics need to feed into the larger whole of the development discourse and its rhetoric. The IPDP program manager states, with reference to the annual application MU consigns to DF, which later is forwarded to NORAD, that:

"We could send the application directly to NORAD, but it is difficult. We have to apply through DF. This is the way we have to apply; it is not in terms of activities. It is in terms of making the proposal attractive. It is in terms of how to make the format and application attractive for the donor" (emphasis added).

The IPDP project manager writes the application, but DF 'makes it attractive' before submitting it to NORAD. An obvious reason for the discrepancy between the project manager's and DF's work is that the project manager does not have as intimate knowledge about NORAD as DF has. This can also account for the fact that DF sends a consultant to help in the preparation of the application of the second phase.

The aesthetic criteria, or presentation format of an application is important. However, Stirrat's argument that its pragmatic objectives and practical ends are irrelevant is to dismiss the donor-recipient relationship on which his data relies. The three empirical cases presented in this chapter show that what is stated in the project documents and the practices undertaken on project level are interconnected. But local practices and the formal order of project documents do not always coincide, but they relate to each other. Reflexive knowledge about development discourse and the aesthetic criteria enable local actors to manipulate the formal order, or at least produce documents that fulfil donor's aesthetic criteria, and consequently feed into the larger whole of development. Though classified as informal practices, they relate to the formal order and contribute to maintain this order.

Practical interrelatedness of formal structures and informal strategies

One of the implications of donor-recipient relationships is the various encounters of different systems of knowledge which create situations of interface articulated through actors on a local level. The interface produces local counter-tendencies and informal strategies towards the formal order of development. The three cases all show the discrepancy between local practical knowledge and the codified actions, practices and knowledge presented in project documents. This discrepancy is only visible and identified at project level among the actors and not in the project documents. The documents, which through their structure feed into the formal order, thus reproduce the formal order as it is supposed to be in the eye of the donor - and do not reflect what is actually taking place locally. The gap between the formal order and local reality is maintained by those who write, fix and produce the project documents in accordance to the criteria they are submitted to and evaluated by the back-donor. This shows that the project exists on two different levels simultaneously: Whereas the documents that are consigned to the donor and back-donor reproduce the project on a formal level, other local practical strategies happen at project level. The informal strategies of local development agents and consultants maintain the formal representations of the project. The viability of the project, both in terms of project implementation and to generate local acceptance of it, is largely due to informal practices that bridge the gap between formal organisation and local knowledge. As shown, one of the most important informal practices adopted by those who are to submit project documents to a donor is to give the donor what is demanded to get acceptance for whatever is applied for. As a result, the informal strategies confirm the formal order and thus reproduce it and the development discourse.

Summarising remarks

The three cases presented above all illustrate strategies and practices that fall outside of the project's stipulated formal order, and the donor-recipient relationship. These informal strategies contribute to the maintenance of the project they relate to, since one common feature of these practices is to confirm the project towards the donor through documents and thus get acceptance. This inevitably maintains the formal status and order of the project. As the formal order is confirmed towards the donor, the expert discursive knowledge about the project is also reproduced. Consequently, the division between discursive knowledge about the field and local practical knowledge is reproduced.

The project manager in the two first cases show that actors are able to draw upon both formal and informal practices and knowledge dependent upon the situation. Whereas he expresses scepticism towards the consultant who bypasses the formal codified practice, he also underlines the problems attached to a literal understanding of a participatory approach, cf. his statements about the delay of the river diversion and the paravet training, and the general notion about flexibility. This is also exemplified by the representative of DF in the first case where she stresses the importance of flexibility and informality in the implementation process, but later works to get the project 'back on track at least formally'. The cases show that local practical knowledge and expert discursive knowledge about development are not two distinct closed entities, but rather two different systems of knowledge which are imputed to and learnt through experience while they also feed off each other. The cases with the consultant and REST show that the actors degree of reflexivity towards a system of knowledge in many ways are determined upon the actors' embeddedness into that particular knowledge.

In next chapter, an account of planning and development from above is given, which is illustrated through the IPDP and the Ethiopian governmental approach to pastoral people. As argued, the Ethiopian state and NGO initiated development interventions are to some extent interconnected - not only in policy, but also regarding the effects they produce. In this way, the chapter also says something about relationships between the state and NGOs, and the various interests present in state and NGO development policy and activities.

Chapter 5. The IPDP: Seeing like Ethiopia?

This chapter is about relationships between state and NGO planning and activities. General development rhetoric and NGO ideology stress the differences between the state and NGOs. NGOs argue for their comparative advantage in working with beneficiaries and local people. According to general NGO ideology, notions of popular support, local participation and grass root orientation constitute NGOs as separate and different from the state. This regards NGOs in terms of being independent organisations from the state, and because they represent and work with local people based on popular initiatives. By focusing on planning, the effects produced and the relations between different actors involved in the IDPD, and the differences between the state and non-governmental organisations are discussed.

The chapter first presents elements and effects that are identified in state planning. Some of these elements, and most notably the effects produced, are similar regarding state and NGO initiated development activities. This is illustrated in the IPDP. NORAD's policy, guidelines and requirements are briefly accounted for, as they influence the IPDP, and thus show some close connections between state and NGO. Secondly, the Ethiopian government's historical and contemporary approach to pastoral people and area is presented. This shows the context that the IPDP works in. All external funded development projects in Ethiopia must support and cohere to the national development plan. The production of the latter also illustrates that there are no clear-cut distinctions between the state and external actors. Additionally, a different approach is briefly presented midways, which draws on Foucault's notion of governmentality. As this thesis is about development, the concept of developmentality is introduced.

To see like a state

Anthropological studies of the state are scarce. This is mainly due to the state's size and complexity. The state is difficult to grasp with traditional empirical-bound anthropological methods. In Seeing Like a State (1998), Scott studies the state and gives an account on the qualifications of its governance and power from an anthropological perspective. Scott approaches the state by identifying and analysing various areas where the state is articulated and which elements state-power depends on. He tries to grasp "…why so many well-intended schemes to improve the human conditions have gone so tragically awry" (Scott, 1998: 4).

Scott identifies four elements in state intervention, arguing that the combination of these four elements is reflected in tragic episodes of state initiated social engineering. As the book title indicates, focus is on the optical and on how the state's view from above only manages to catch simple and relatively homogeneous patterns. State intervention becomes a question of straight lines and transparency, hence what is local and complex are simplified and standardised (Neumann, 2001a: 566). These legible and standardised units enable the state to establish direct causal relations between intentions, plans, input and expected output.

Michel-Rolph Trouillot (2001) adopts Scott's approach to the state by giving focus to the multiple sites where the state is articulated through its effects. Trouillot validates that governments are not the only actors that 'see like a state'. With reference to Scott, Trouillot argues that "…the state is not reducible to government" (2001: 127) and that "NGOs are only the most obvious cases begging for an ethnography of state effects" (ibid.: 132). Trouillot's foundations for these statements are three related and recognisable propositions about the state: First, that state power has no institutional fixity on neither theoretical nor historical grounds. Secondly; state effects are never obtained solely through national institutions or in governmental cites. Thirdly; these two features have been exacerbated by globalisation. Since the state has no institutional or geographical fixity, Trouillot argues that "… its presence becomes more deceptive that otherwise thought, and we need to theorize the state beyond the empirical obvious" (ibid.: 126). This removal of ethnographic boundaries means that the state becomes more open to ethnographic studies that take the fluidity of borders into account. Ethnographic studies should focus "…on the multiple sites in which state processes and practices are recognisable through their effects" (ibid.), i.e., the sites where the state is identifiable through its self-communication and articulation.

Elements and effects of state intervention

Scott identifies four68 elements that state initiated social engineering depends upon, and argues that the most tragic episodes of state intervention are due to the pernicious combination of these four elements. These elements are recognised in other institutions, such as NGOs, which due to their top-heavy approach in terms of planning and intervention, and their need for legible units, produce state-like effects. This is relevant to the IPDP in terms of the planning and project implementation, which rely upon formal order and representations of the field.

The first of four elements Scott identifies is an 'administrative ordering of nature and society', which denotes a simplification of the field. This is done in order to get an overview, meaning that nature and society are reduced to statistical facts to promote legibility for the viewer. Scott argues that these simplifications have at least five characteristics. 1: State simplifications are observations of only those aspects of social life that are of official interest; they are interested, utilitarian facts. 2: They are nearly always written (verbal or numerical) documentary facts. 3: They are typically static facts. 4: State facts are also aggregate facts. 5: These facts are standardised, that is, regardless of the incommensurability of what is simplified, the facts are collected and treated in the same manner in order to promote comparability and make collective assessments (Scott, 1998: 80). The second element is what Scott calls a high-modernist ideology, conceived as a strong self-confidence in scientific and technical progress. The high-modernistic ideology implies a rational design of social order that commensurate with the scientific understanding of the laws of nature. The state's optical approach to nature and society and its overview and simplifications are done in accordance with high-modernistic ideology. This leads Scott to his third element, which is an authoritarian state able and willing to use its full weight of coercive power to bring the high-modernistic design into being. According to the state's selected representations, or simplifications, the state sets out to form the reality, the nature and socio-cultural aspects within its borders of control (Scott, 1998: 4-5). The fourth element is interrelated with the third, which is a "…prostrate civil society that lacks capacity to resists the state's plans" (ibid.: 5). Scott summarises:

"In sum, the legibility of a society provides the capacity for large-scale social engineering, high-modernistic ideology provides the desire, the authoritarian state provides the determination to act on that desire, and an incapacitated civil society provides the levelled social terrain on which to build" (Scott, 1998: 5).

Scott does not focus on the state in general, rather his concern is about the characteristics of flawed state initiated interventions. According to Scott, the failure of 'certain state initiated schemes to improve the human conditions' is due to the combination of the four identified elements. Scott sees the state as the manifestation of modernistic ideology per se.

Trouillot applies Philip Abrahams who says that "[t]he state is … an ideological project. It is first and foremost an exercise in legitimation" (cited in Trouillot, 2001: 127). Trouillot identifies four state-like effects that are recognised in NGO activities, which in sum largely cohere to and specify Scott's first point. Of these four effects, the first is an isolation effect where the state produces atomised individualised subjects moulded and modelled for governance as part of an undifferentiated but specific public. The second is an identification effect, perceived as a realignment of the atomised subjectivities along collective lines which within individuals recognise themselves as the same and alike, cf. what Scott calls state simplifications and the administrative ordering of nature and society. The production of a conceptual apparatus as theoretical and empirical tools to classify and regulate the collectives is the third effect, which is named a legibility effect. The apparatus that generates and promotes legibility consists of both a language and knowledge for governance. The production and definition of boundaries and jurisdiction, a spatialisation effect, is the fourth effect produced by the state (ibid.). The spatialisation effect implies that the field for intervention is defined as a social sphere detached from other external components that might impoverish the field's legibility.

If anthropologists are to study the state, the defined boundaries that constitute it as an object must be erased. Focus must rather be on the state as a set of processes of which the effects can be observed and identified in the multiple sites they are articulated. Other organisations and institutions than the state also produce these effects. NGOs generate state-like effects through their processes. This underlines that the state and its civil society are intertwined and that the state is not merely reducible to governments.69 NGOs, among other institutions and organisations, rely upon similar elements and thus produce the same effects as states do through their interventions. The lessons from the ethnography of the state, the elements identified in tragic state interventions and the effects produced, are adaptable to NGOs to account for and illustrate the general processes of development interventions.

The IPDP: seeing and acting like a state?

The representations of the IPDP, illustrated in chapter three, are examples of the simplifications of the field made by planners to have legible units to intervene on. The IPDP planners rely upon a readable and accessible field. These simplifications are generated in accordance with the high-modernistic ideology, i.e., a rational, scientific and technocratic knowledge. Though in many respects similar, high-modernism is not completely identifiable with modernism. Whereas modernism has political connotations, the high-modernism is an apolitical attribute the state relies upon for its governance, planning and intervention. Referring to villagisation in Tanzania, the city planning of Brasilia, German forestry planning and The 1917 Revolution in Russia, Scott states that "[h]igh modernism [is] politically polymorphous; it could appear in any political disguise, even an anarchistic one" (1998: 164). High-modernism refers to the ideology reflected in planning and governance among both states and NGOs.

In the case of the IPDP, nature and society are ordered in a general and simplistic manner. Legible units enable the planners to arrange a chain of causality between the input of resources and the expected output, as well as they make planning, administration and intervention more uncomplicated. A central member of the IPDP board states: "IPDP was initiated to assist the pastoral people, the Afars in the area. The nomadic Afar people have always been neglected in development projects. Tigreans have always had projects and development. The IPDP addresses these nomadic Afars that are remote to development". On question related to any possible constraints regarding the project, he responds:

"The lack of data, statistics, something to give the overview. Since the project manager doesn't live in Aba'ala, we need good data samples and statistics to plan for the IPDP. But Afar have since long time been neglected by the government, so we collect data and samples ourselves. When we talk about capacity building, this also includes research and collecting data. The DHG [Dryland Husbandry Group] and Mekelle University helps IPDP in this work".

The IPDP is deduced from information gathered by project staff, DHG and MU. The previous program manager tells from the initiation of the IPDP that the Development Fund wanted a new project outside Tigray, and "DF asked me to come up with ideas, because when you collect data, you come up with recommendations for the users, the beneficiaries". A consultant assigned by DF came to give input in the making of the initial pilot phase. The consultant and the first project manager "…went to field for, let's say maximum 15 days, we done [sic.] the survey … and there was some very similarities [between] what I had been putting on the paper and what the survey told her to do". Although one of the project objectives is "[t]o strengthen the integration and relations between the Afar and Tigray people", the consequence of how the project was planned is that it formally and intentionally either approaches the pastoral and agro-pastoral modes of production, which are attributes ascribed to Afars, or farming, which is associated with Tigreans. The integration of these two groups in the various project components is largely absent. The representations produced of the field initially neglected the Tigreans in the project area and their agricultural mode of production. Despite this, and generally speaking, Tigreans are those who benefit most from the project's activities. Reasons for this are found in the representations made in the initial planning phase, which produced static simplifications of the multifaceted field. This illustrates that the IPDP produced legible units and representations.

The effects of isolation and identification, as described by Trouillot, are generally interrelated, as also is illustrated regarding the IPDP. In the initial planning process, the people of the target area were codified as undifferentiated individuals, in the sense that they are detached from each other and seen as 'atomised individualised subjects', i.e., an isolation effect. These subjects are then realigned in accordance with the planners' stipulated collective lines and groups in which people are seen as the same, that is, an identification effect. Individuals are not seen as independent actors with their own practices, but rather ascribed to and recognised as one of the two homogeneous ethnic groups in the area; Afars or Tigreans. The latter group is more or less neglected in the project, despite the planners' 'acknowledgment of the area's heterogeneity'. "Just listen to the name of the project. It is a project for the Afars both in terms of people and region, and not the Tigreans", a board member states. The formal representations of the project undermine that the project area consists of various aspects that crosscut the notion of homogeneous ethnic groups, e.g. that the Afars are either pastoralists or agro-pastoralists, and the Tigreans are farmers. Additionally, the idea that Afar is inhabited by Afars underscores the identification effect.70 Tigreans are not only a majority in Aba'ala town, but also in other project areas, such as the tabias of Hidmo and Wuhdet.71 Many beneficiaries are approached as practicing pastoral or agro-pastoral modes of production although they define themselves as farmers and Tigreans. This illustrates the realignment of rather complex and heterogeneous groups into atomised subjectivities ordered along collective lines to generate legible and tangible units of the people and socio-cultural aspects with the purpose of intervention. People are not seen as individuals, but as similar and static objects living within a defined area. Despite this, the project tries to bridge the relations between Afars and Tigreans. The IPDP activities are mainly designed to and directed towards Afars, largely based on the assumption that only Afars live in Afar. Nevertheless, many project activities in practice address almost exclusively Tigreans. This is due to planning in general, since the interventions and activities necessarily need to take place somewhere. As the nomadic Afars move around, it is easier to approach the settled Tigrean population. The planners' perception of Aba'ala wereda and the legible units and representations produced in order to intervene illustrate the identification effect. The generation of legible units of the target area and group enable the planners to establish causal relations between input, effort and output.

Isolation and identification effects are distinctive with regard to the different activities of the project, especially those activities concerned with training or capacity building. These activities also underscore the modernist aspects of planning: "10 Afar animal health technicians will be trained outside Aba'ala in both modern veterinary medicine and local animal health knowledge for six months"; "A short-term training in office administration will be designed for 20 local Afar leaders … about governance, accountability, transparency, planning, budgeting, monitoring and evaluation". On small business management, 40 Afar women will be trained to "…improve the credit and savings programme at a grassroots level in the pastoral areas of Afar" in order to "strengthen the role and capacity of women in running effective and successful businesses…".72 The high-modernistic ideology promotes a strong faith in rational planning, and the idea that a specific input determines the output.

The development apparatus' theoretical and empirical tools to classify, regulate and stipulate collectivity and intervention, that is, the production of representations, is what Trouillot names the legibility effect. Legibility is a means to establish a formal order and is gained by generating simplistic and general models where individualism, complexity and ethnographic particularism are reduced to a legible pattern or scheme characterised by representations. Scott states that "[d]esigned or planned social order is necessarily schematic; it always ignores essential features of any real, functioning social order" (1998: 6). This is illustrated in the IPDP regarding the representations stipulated in various project documents. The spatialisation effect is signified with the definition of target areas and beneficiaries. This implies a construction of boundaries, which segregate those within from the outside, and confines the area for interventions for the developers, who treat the people and the area as segregated from its surroundings and external factors. As a result, the problems to be solved are explained with reference to internal conditions and not to factors external to the defined boundaries. This is not only an effect of development work, but also a means to legitimate and organise intervention. It is illustrated in the IPDP by the definition of target groups and target area.

Two elements contested

As illustrated above, the four state-like effects stipulated by Trouillot are identified in the IPDP. The effects arise and are recognisable as a result of the processes and practices involved in the planning and general implementation of the IPDP. Trouillot shows the relevance of adopting Scott's notion of the elements found in tragic state initiated engineering on NGOs. Two of the four elements identified by Scott are present in the IPDP: The administrative ordering of nature and society based on simplifications and the high-modernistic ideology. The two remaining elements (i.e., an authoritarian state willing to use its coercive power to implement its high-modernistic ideas and a powerless civil society that lacks the ability to oppose these plans) are not identified in the ethnography of the IPDP. Donor-recipient relationship is formally supposed to be an equal relation, cf. the notion about partnership. In practice, there are certain inequalities. The imbalance is articulated in the encounter between donor and recipient which underlines that power are relational and do not solely exist as latent forces in the formal structures. Therefore, the possible coercive power of the donor towards the recipient must be analysed in the multiple sites it is recognisable through its processes. Scott's main concern is about power relations between planners and local civil society. Applied to the IPDP, this would be the relationship between the IPDP and the beneficiaries. Nevertheless, is it applicable in the donor-recipient relationship between DF and MU/ DoA as the recipient organisations not only are seen as representing the target group, but also because they are the beneficiaries of some of the activities, cf. the IPDP objectives of strengthening the local capacity and the institutional building. As shown in the previous chapters, local strategies and counter-tendencies occur in order to cope with formal structures imposed by the donor. The formal order, which largely is defined according to the donor, is not necessarily reflected in practice. There are additional reasons to why Scott's two last elements do not accord to the IPDP.

First of all, a donor and recipient are mutually dependent on each other. Not only is the MU and DoA dependent on DF's transfer of resources, but also the donor is dependent on having a good and accountable partner and recipient organisation in order to validate its role and actions towards NORAD. The involvement of the consultant who knows the NORAD format is a means to make the IPDP and the implied organisations appear good. Moreover, the general role of DF regarding the annual applications sheds light on this. The IPDP project manager argues that MU could consign their applications directly to NORAD, but admits that they do not have any special knowledge about NORAD, and thus DF helps to make the application 'more attractive'. Secondly, Scott refers to the tragic results of states' interventions. Trouillot validates the application of Scott's elements to NGOs, but the IPDP and its accomplishments are not tragic episodes of social engineering in Scott's sense. Thirdly, the three cases depicted in the previous chapter falsify the notion that DF applies an authoritative coercive power. The cases show that the disequilibrium of power in donor-recipient relationship is possible to manipulate and that also the donor NGO can apply such practises. The REST case and the statements from the IPDP board members also emphasise that recipients are entitled to say no and reject a project proposal. This challenges the notion of the donor as an authoritative actor willing to apply its potential coercive power. Fourthly, the notion of whether the IPDP's beneficiaries are a 'prostrate civil society' is ambiguous, because the IPDP's intentions are to empower the civil society by its capacity-building components, awareness processes and bottom-up planning. Despite the falsification regarding the application of bottom-up planning, cf. e.g. the case with the consultant in previous chapter, the beneficiaries are not necessarily victimised by the planners' use of authoritative power. The reason that participatory community planning is neglected is due to the mode used to plan, which relies on a legible ordering of nature and society.


In the following paragraph a somewhat different approach to donor-recipient relationships is given to illuminate about a contradiction in the formal order of development. Formally and intentionally, the IPDP and DF seek to empower the recipients in order for them to plan, manage and implement their own project. The formal design of the IPDP coheres to Chambers' notion of participation and empowerment, i.e. to "…enable people to do their own analysis, to take command, to gain in confidence and to make their own decisions" (1995: 30; cf. chapter three). The decisions they are suppose to make, however, shall resemble the policy of donors, and just about all donor agencies aim at policy coherence. Participatory approaches, empowerment and policy coherence are alike what Foucault calls governmentality; a phenomenon relating to governing and control. The concept of governmentality was advanced by Foucault as a tool to understand precisely what is special about the use of power in modern societies. Foucault developed the concept primary to understand state power over individuals. It is also adaptable to other power relations aligned with institutions that seek to embrace and control people, such as development interventions by NGOs. Governmentality is composed of two words; government and mentality, and describes that state power is maintained by individuals since they have adopted the state's mentality. The concept denotes a relation of power that is related to how individuals govern themselves; that the individuals are controlled and governed by their own truth and mentality (Neumann, 2002a; Foucault, 2002). Through the state's transfer of its knowledge and mentality, the individuals are supposed to govern themselves in accordance to the state's mentality. People get a sense of freedom and individualism, imposed by the governor. Thus, individuals control themselves while being forced to freedom. Since this freedom and individualism are based on the governor, the governor's power and mindset are maintained. Hence, governmentality is about indirect control and rule. The individuals control themselves according to the governor's interests. Governmentality indicates that the citizens internalise the state's values. The intention of empowerment and participatory approaches within the development sector resemble this, i.e. that the beneficiaries shall internalise the developers' ideas and mentality and thus manage their own project under the donor's supervision and guidance, and in accordance to the development agency's requirements. What is seemingly liberal (empowerment and participatory approach where the beneficiaries make the decisions) is actual a means to transfer developmentality as an indirect mechanism of control and management over the target group. Developmentality is about making the beneficiaries internalise the developers' values and policy in order to 'empower' them to govern themselves and sanction their own norm-violation of the development discourse. Developmentality signifies that the beneficiaries should develop and manage their own project by adopting the donors's (developers's) mentality. The mentality that is transferred is found in the development apparatus's policy. The notions of participation and policy coherence oppose each other, and illustrate a formal contradiction within the formal order of development and its apparatus.

This contradiction is found in the development apparatus's formal order and in general development rhetoric. Donors in general, as the Development Fund and NORAD, embrace and emphasise the ideas of empowerment, participatory approaches, bottom-up planning and community-based projects. Simultaneously they aim at policy coherence with other donor organisations and countries (cf. DAC's role) and with the Ethiopian government's national development plan. The idea of participation, however, is to generate grass root orientated projects to secure popular awareness and to design more contextual development projects since it is the beneficiaries themselves that best know about their own problems. Participatory approaches are also a means to make the beneficiaries responsible and accountable for the project. The potential success of the processes of empowerment relies on whether the beneficiaries internalise the knowledge and mentality of the development apparatus or not. Participation and processes of empowerment imply that the developers try to make the recipients internalise their system of knowledge. The beneficiaries are not only supposed to plan, design and manage their own project, they are also supposed to monitor and control it according to the ideas and policy guidelines of their donor. Participation and empowerment are about giving the beneficiaries a sense of ownership to the project. Developmentality and the ambiguity of conditionality, or policy coherence, and participation (to generate a feeling of responsibility and ownership) are illustrated by the words of one donor informant cited in The Reality of Aid: "[O]wnership exists when they do what we want them to do but they do so voluntarily" (Randel, German and Ewing, 2002: 8). The abovementioned is alike the concept of governmentality, but as it is about transferring development knowledge, I find it plausible to denote this phenomenon as developmentality.

The concept of developmentality is interesting not only when discussing the ideas of policy coherence, participatory approaches and bottom-up planning, but also when studying partnership relations between development organisations. Donor-recipient relationships imply transfer of knowledge and structures. The guidelines of such partnerships and the donors' power to define and influence the recipients illustrate the phenomenon of developmentality. Recipient organisations are responsible and accountable to their donors, which is a requirement NORAD consigns to their recipients. Through Norwegian NGOs this requirement is transferred to other recipient organisations. Imposed duties of accounting, reporting and evaluation are means for the 'governor' to transfer its 'mentality' to others. This relates to the relationship not only between NORAD and DF, but also between DF and MU.

There are inequalities between donor and recipient in terms of power that are latent in the partnership relation, which the recipients are aware of:

"It would be stupid of us as recipient not to acknowledge the power to define the projects by the donors. Of course we have to consider the policy of the donor. We are obliged to, in order to get funds. Though an independent NGO, REST has to follow several guidelines from different donors, as well as national and international policies on development. That's the rule of the game, and we must play along. It is an everlasting struggle for funds, to make the ends meet. If we don't manage to do this, do our job, the poor, rural and less-developed peasants will suffer".73

Whether the donor employs its formally potential power or not, there is a discontinuity in power between donor and recipient articulated in their encounter. This potential power is consolidated in partnership agreements and in the guidelines between NORAD, Norwegian NGOs and their partners abroad. Therefore, these relations need to be elaborated.

NORAD guidelines

The cooperation agreement between NORAD and their grant recipients (Norwegian NGOs) states that the recipients must submit annual reports on each project that receives funds from NORAD. Regarding the IPDP, DF is accountable and responsible to NORAD, and DF's recipients are responsible towards DF. This means that MU and DoA generally need to cope with the same guidelines towards DF as DF must towards NORAD. Hence, annual applications, progress reports, auditing-reports, evaluations and budgets have to be submitted to the grant-recipients' respective donor. At the end of a project phase, the recipient is obligated to submit an extended report. This "…report shall as far as possible be founded on evaluations that have been conducted".74 Through this model of partnership, the recipient, whether it is a Norwegian NGO or a foreign recipient, has obligations towards the donor and back-donor. Among the compulsions are that the recipient is "…obliged to ensure that the activities under this agreement are carried out in compliance with the politically adopted guidelines for Norwegian development policy, including the Norwegian Government's Action Plan for Combating Poverty in the South towards 2015, … Report No. 19 to the Storting (1995-96) and annual provisions stated in Proposition No. 1 to the Storting".75

The grant recipient, DF, "...has the administrative and financial responsibility for the grant" and NORAD and the Office of the Auditor General of Norway "...reserve the right to carry out control at any time to confirm that the grant is being used according to intentions".76 An official from the Norwegian embassy in Addis Ababa, who explains that the role of the NORAD run embassy is to supervise the Norwegian funded development projects in Ethiopia, underscores this. The role of the embassy is to control whether the money is used according to planned intentions, evaluate incoming applications and give acceptance to NORAD funded development projects. The representative from the embassy states that all NORAD funded development projects in Ethiopia have to be in coherence to the pillars of Norwegian development aid, and recipient organisations need to fulfil the requirements put forward by NORAD.

Løngreen (2001) argues that the guidelines towards recipient countries and organisations also contribute to stigmatise and maintain the division between 'us' an 'them', based upon representations. This is reflected in 'the development gaze', i.e., how the donors see the recipients, which produces certain images of the recipients.

Representations and the 'development gaze'

Løngreen scrutinises several of DANIDA's77 guidelines and recipient country profiles with the approach: "What images were produced of development and the 'Others' by the uses of visual representation in the material?" (2001: 221). She argues that 'the development gaze', which refers to the donor's optical view generated by the development discourse, presents the poor recipient countries as static, unmodern, homogeneous and in the need of 'our' interventions and aid. The development gaze can thus both be defined "…as social practice as well as a specific way of representing this very practice through visual representations" (ibid.: 227). Løngreen's main concern is that the representations created by the donor reproduce the static view of 'the Others', and that it shows a clear aspect of the power connected to donor and recipient relationships:

"[T]here is an obvious power-relation between 'Us Here' and 'Them Out There'. In development projects we ('Us Here') are always those who give support to the 'Others'. This has consequences for the ways in which we look at each other. Even though we are talking about participatory projects there is always an element of control. To legitimate the use of funds in development projects it is necessary to legitimize the use of funds to persons who have given them such as private sponsors and taxpayers. The development relation therefore becomes bureaucratic and accordingly some sort of administration always springs from development projects" (ibid.: 227).

Løngreen and NORAD's requirements, articulated through the contractual agreement and by the representative from the Norwegian embassy, both underscore the relations of power in donor-recipient relationship. Much of the power lies in the processes and practices used to establish and define development projects and the relationship between donor and recipient which generate certain state-like effects on how developers approach the target area. The formal guidelines of NORAD and DF challenge their own representations, which state that they apply a participatory approach and bottom-up planning for their community based projects. Instead, generalised assumptions are produced of the target area and communicated in a formal manner through project documents. Development planning thus shapes how we perceive 'the others'.

The representations produced of the IPDP project area, Aba'ala wereda, are that the area is inhabited by a mixture of pastorals, agro-pastorals, peasants, Afars and Tigreans.78 This variety is acknowledged in the project documents but not taken into practical account in project planning and implementation. Referring to the project document of 2002, the IPDP is supposed to cover "…important aspects of people's lives. This means focusing on the pastoral mode of production, acknowledging the fact that Aba'la is a heterogeneous physical and human environment".79 First, this quotation explicitly states the simplifications done in order to intervene. Secondly, as I have argued in chapter three, it is the sedentary people mainly in and around Aba'ala town who get most benefits from the IPDP. Thirdly, it is easier to approach sedentary people than nomads, because settlements are more legible. Fourth; development projects (i.e., the construction of dams, ponds, irrigation sites, among others) have to take place somewhere, which implies that in order to get the benefits, the most advantageous is to live nearby the project.

In the forthcoming section, the Ethiopian government's historical and contemporary representations of and approach to pastoral people are depicted. Not only is this interesting regarding the IPDP, since it visualises reasons for the inherent scepticism towards external actors among the Afar people, but also is it interesting because it resembles Scott's elements - found in tragic state initiated social engineering to a larger extent than the IPDP does. It makes the distinction between the large-scale approach of the Ethiopian government and the micro-scale approach of the IPDP clear, despite the latter also produces state-like effects. The last section of this chapter elaborates around contemporary governmental policy towards pastoral people. This is relevant to the IPDP, since all development interventions in Ethiopia are supposed to cohere to the Ethiopian governmental development plan on general policy issues.

Pastoralists on the government's scene

The general view on pastoralists has largely been hi-jacked either by romantics (often anthropologists) who idealise the pastoral way of life, or by pessimists (mainly range scientists, ecologists and economists) who talk about overgrazing, range degradation and desertification as the inevitable consequences of a pastoral way of life (Hogg, 1997b: 1). Regarding Ethiopia's policy towards pastoral communities, the latter view has been adopted. Pastoralism is seen as a threat to the common resources of the country: "The overall approach and tone of government range management interventions was dictated by the dominant range ecology paradigm…" (ibid.: 15). This view has dominated the governmental and indirectly the NGOs' approach to pastoral communities. 80 The Ethiopian government tries to sedentarise pastoral communities with the objective to prevent range degradation. This is also the case with pastoralists in Afar, who to a great extent have been exposed to government-initiated sedentarisation and resettlement programmes (Getachew, 2001; Said, 1998).

The Ethiopian state's encounter with the pastoralists; a historical account

Historically, the Christian highlanders have scarcely been interested in the Afars of the Ethiopian lowlands. Afar-Abyssinian relations passed through various phases up until the late 19th century; sometimes quite hostile, but generally cordial. This was due to several reasons: The strategic location of the Afars along the Red Sea coast; the existence of trade routes to the hinterland; the location of saline lakes which were the source of salt used as a medium of exchange for the Abyssinian kingdom until the early 20th century. These issues coupled with the fact that the Afar land was the entrance point for external aggressors made Abyssinia's rulers wary of antagonising the Afar, and the regime's relations with its peripheries were for a long time at an impasse (Clapham, 2002). It was first with Menilek's (1889-1913)81 expansionist drive that the Afars were encapsulated into the Ethiopian empire. The relationship between Menilek and the Afars were cordial since Menilek's only concern was of an economic matter. The Afars' problems in the encounter with the highlanders arose first during the reign of Haile Selassie (1916-74), 82 and especially in the 1960s when Haile Selassie intensified his modernisation efforts. He adopted a policy of assimilation of the Afars into the culture of the ruling Amharas with the objective of Christianisation and a dissemination of Amhara language and culture. The crucial turn between Afar and Amhara came in 1960, when the Ethiopian emperor ordered an economic development programme along the Awash River Valley.83 This introduced large-scale commercial cotton farms, and the subsequent establishment of the Awash Valley Authority (AVA). AVA was a governmental agency entrusted with the agricultural development in the fertile Awash River basin. The effect of this was that the tracts of Afar dry-season grazing land were lost to commercial irrigation schemes. Under the management of AVA, development in the Awash Valley took the form of large-scale mechanised commercial enterprises mostly managed by foreign companies in joint venture with the state (Said, 1997; Said 1998). Since then, Afars have always had a problem with and been sceptical to external actors, i.e., the state in particular and development authorities in general. There have been many encounters between Afars and external actors in course of history. These relations and encounters need further elaboration in order to understand the general scepticism among the Afars towards outsiders and thus give an account on why 'they needed to be talked into the project', as was the case with the IPDP. Governmental development schemes towards pastoral people have by and large pursued to sedentarise the pastoral communities.

Land reforms, producers cooperatives and sedentarisation

Shortly after what was to be known as the Derg84 came to power in 1974, the military junta that took power committed itself to radical national reforms, which would have serious social, political and economic implications. The radical policy measures that were adopted were primarily designed to change the agricultural production system and transform the traditional power structures of rural societies (Viezzoli, 1992: 163). The most significant reform was the land reform proclamation of 1975. It abolished age-old feudal production relations and consequently all land was nationalised without compensation to the previous users and landholders. Use-rights to land were allocated among local inhabitants by newly created and Derg appointed peasant associations (Clapham, 2002). Nationalisation of all rural land not only led to the expropriation of the Afar sultan, but also deprived the pastoralists of their large tracts of dry-season grazing land, which through the large governmental irrigation schemes were turned into state farms. All commercial farms were nationalised. Since 1974, there has been considerable expansion of state run irrigation schemes, particularly in the Awash Valley. The establishment of the Awash National Park in 1966, covering an area of 830 km2, and the construction of large dams, e.g. the Koka Dam in the Upper Valley, generated major challenges in nature resource use among the people living in the Awash Valley. The expropriation of grazing land and changes in the Awash flood regime have, in combination, seriously affected the lives of the Afar pastoralists (Said, 1997: 124).

Another objective of the Derg's policy was to establish large-scale state run producers cooperatives, which was to be achieved through sedentarisation, or villagisation, and resettlement programmes. By nationalising all land, huge state owned cooperatives were established. The aims were to increase the productivity and rationalise the use of arable land, and thus to increase the economic profit. As argued below, the nationalising of land became also a means for the government to control its population. Sedentarisation and resettlement programmes contributed to the implementation of the Ethiopian government's economic policy and had the objective to restore the production equilibrium between the various regions.

"Resettlement involved the long-distance movement of people from 'overpopulated areas into planned modern villages elsewhere. … Villagization, on the other hand, involved the concentration of scattered homesteads into designated villages by peasants who continued to farm the same land as before" (Clapham, 2002: 19).

The Ethiopian government's resettlement and villagisation policy aimed particularly at three distinct target groups: The urban unemployed, the populations regularly affected by natural disasters, and nomadic populations. The objectives of the government's regional framework policy were to create income generating activities and commercial networks among the beneficiaries by settling the people in areas where virgin land was converted to agricultural plots (Viezzoli, 1992). This not only diminished the grazing area that the nomadic population and its livestock relied upon, but through utilising its coercive power the state (re-)settled enormous groups of people. The resources were also drawn from the Ethiopian periphery into the centre of the Amhara highland "…in order to meet the growing state bureaucracy and expanding military. To the Afar this always meant social injustice" (Said, 1997: 138)

In 1974 the Derg established the peasant associations as a means to administer the land reform. Initially, they were genuine representations of peasant interests, through elected groups set up to administer land distribution in a traditional system of collective responsibility, including local courts to arbitrate in disputes and adjudicate minor offences. Later, the Derg realised they had consigned too much control over resources and decision-making (Pausewang, Tronvoll and Aalen, 2002a). From 1977, "…the peasant associations were reorganised radically and centralised as instruments of control over the peasants in the hands of the military government and, after 1985, the Workers' Party of Ethiopia (WPE)" (ibid.: 5). Through what has been called the project of encadrement, the Derg appointed local party officials to control and govern local activities. The processes of villagisation and resettlement "…were greatly accelerated by the eagerness of local party officials to vie for ventral approval" (Clapham, 2002: 21).

Clapham argues that the pastoral people of Ethiopia were only partially incorporated into the new vision of the central government. The pastoral people represented insuperable obstacles to the Derg's social vision of development in the peripheries: The pastoral people's mode of production and living were hard to transform in order to settle and collectivise them. In addition, the pastoral people, as Muslims, represented a security threat. This was approached by continuing the imperial regime's resettlement and cash-crop cultivation in the pastorals' dry-season grazing areas, and by supporting the internal rivalry between different Afar clans (ibid.: 22). Though never completely successful in terms of sedentarisation, the Derg's policy towards the Afars heavily affected their way of living and the possibilities of continuing their pastoral life and mode of production.

Sedentarisation policy in contemporary Ethiopia

The new government that came to power in 1991 sustained certain features of the Derg regime's policy regarding development issues. Some are implemented through contemporary policy. Hogg states that it was "…implicit in the new Government's economic statement of policy in 1991, after the collapse of the Derg, that sedentary life in nomadic areas was considered to be the ultimate goal of good conservation and economic policy" (1997: 15, my emphasis). If implicit in 1991, it is today explicit: An official publication dated July 2002 from the Ethiopian government, Ethiopia: Sustainable Development and Poverty Reduction Program (SDPRP),85 states, under point 7.3. about 'Pastoral Development', that: "There is a need for developing technology packages, which help to strengthen the useful traditional practices and do away with harmful ones" (point: 7.3.2.). Later in the document, it becomes clear what the government recognises as harmful traditions to be approached through development projects

"…designed to deal with problems of the pastoral communities, provisions of the necessary infrastructure for sustainable development and radical transformation of their livelihood … Such a development agenda could be well effected only if the people can somehow be settled. Selective settlement programs are believed to be the only viable option in the long run. …Settlement programs within these areas should be directly linked with the development of irrigation projects. Otherwise, they are bound to fail. …the change goes beyond a change of location and alters their cultural life. … It would also require a significant cultural transformation. The objective is to settle pastoral population" (ibid., my emphasis).

The concepts 'producers cooperatives' and 'resettlement programs', highly associated with the Derg, are no longer used, despite many of the same ideas are preserved. Whereas the Derg used military power to enforce their policy, the EPRDF86 uses its power through state ownership of all land. Local party cadres, appointed by central authorities, control and distribute land on local level. Point 7.1.3. in the SDPRP declares: "The land policy states that land belongs to the people and the Federal and regional governments administer it on behalf of the people". Decentralised governmental offices are scarce in Ethiopia, where 85% of the people live in rural areas, and the control over land is a means to control the population. The surveillance and distribution of land is executed by party cadres, which are appointed by the central government, on a local level (Pausewang, 2000: 4). The management of land and people are, in the Ethiopian context, directly interconnected. The Afars are also affected by this: The good grazing land close to rivers have been occupied and converted into irrigation areas and hillsides are enclosed, both within the notion or paradigm of 'correct' natural resource management. The result is an increase in population and animal density due to less accessible land, and the pastoral people are thus facing the dilemma of whether to move to more remote and marginalised areas or settle and cultivate the land. This makes it hard for pastoral traditions to survive, due to lack of water and good arable land in the more remote areas.

The control of land and people is among the government's means to implement its general policy and to maintain its power to do so. The Afars, which by the government are perceived as a remote and backward people (Getachew, 2001), have been victims of state aspirations to prescribe and deal with 'the problems of the pastoral communities'.

Prescribing Ethiopia

The coherence between the elements James Scott (1998) identifies in tragic state initiated social engineering programs and Ethiopia's (re-)settlement programs is salient. First, the pastoral people are targeted by the state's aspiration to control them, to administer their society. The state creates the pastoral groups as an object of knowledge possible to intervene on by generating simplifications and legible units. Local and individual variations are neglected and consequently the pastoral people of the lowlands in Ethiopia are seen as one homogeneous group. This is done in accordance to the sweeping vision of rational central planning, and a high modernistic ideology applied by the state. Scott states in a general manner, though relevant in the case of Ethiopia, that:

"If, as we have seen, the simplified, utilitarian descriptions of state officials had a tendency, through the exercise of state power, to bring the facts into line with their representations, then one might say that the high-modern state began with extensive prescriptions for a new society, and it intended to impose them" (1998: 90).

The Ethiopian government's prescriptions of the pastoral communities are given and imposed, as it has for the last 40 years in various degrees, which underline the high-modernist ideology Ethiopian state interventions relate to and rely upon. High-modernism does not refer to the state's policy, but to the way a state necessary needs to see in order to implement its plans. In retrospect, this high-modernistic optical view is reflected in the last three Ethiopian reigns approach towards pastoral people: The large scale irrigation development plans of emperor Haile Selassie in Awash Valley, the resettlement schemes and the state-run producer cooperatives of the Derg, and the settlement programs of the post-Derg government. All these reflect a high-modernist ideology applied by the various governments regardless of the incommensurability of the different governments' politics or their fundamental ideological values. Scott argues that "…high-modernism was not the exclusive property of any political tendency; it had both right- and left-wing variants…" (ibid.: 88), and high-modernistic ideology "…tends to devalue or banish politics" (ibid.: 94).

The Sustainable Development and Poverty Reduction Program (SDPRP) makes it obvious that the Ethiopian government wants to reshape society and especially the remote pastoral areas. Though not referring to Ethiopia, Anna Lowenhaut Tsing underlines the state's difficulties aligned with the approach to remote pastoral areas. She refers to the Indonesian state's attempt to approach the nomadic Meratus hill people of Kalimantan. Tsing states that the Meratus live in an area that "so far, has eluded the clarity and visibility required for model development schemes" (cited in Scott, 1998: 187). The various Ethiopian governments have faced the same problems in their approach to pastoral communities. Nomads will always challenge the state, because they move around. Consequently, the state has difficulties in 'seeing' them. Ethiopia's settlement politics are just one example of general state conduct there are many variants of. Whereas legibility is a necessity to model development schemes it is also a condition of manipulation. "Any substantial state intervention in society … requires the invention of units that are visible" (ibid.: 183). To see the issue of sedentarisation of pastoral people in Ethiopia merely as the state's intention of development or preventing range degradation would conceal important aspects. Among the state's needs are the requirements for defence and state conscription (Said, 1997, 1998), surveillance and political control (Pausewang, Tronvoll and Aalen, 2002a, 20002b, 2002c), labour discipline, taxation and legibility (Scott, 1998).

The settlement programs of Ethiopia's various governments are attempts to redesign rural life and production from above, to reshape rural space into state space, and create legible units of taxation, labour discipline, surveillance and political control. Whatever units are produced, redefined or manipulated by the state, they must be organised in a manner that permits them to be "…identified, observed, recorded, counted, aggregated and monitored" (Scott, 1998: 183). Infrastructure in rural Ethiopia is scarce, and the government's paramount means to control the population is to control access to the most important resource for the people: state owned land. State surveillance, monitoring, (re-)distribution of land in rural areas are done by 'party cadres'. The party cadres, which are appointed by central authorities, operate on local level as the central party's advocates. The cadres are part of a party structure below the surface "…that keeps tight control at all levels and makes sure that no one can use [the] democratic institutions efficiently to challenge its power" (Pausewang, Tronvoll and Aalen, 2002c: 230).87 In line with the constitution, local administration has received a formal structure supposed to facilitate democratic participation of the local people and accountability of leaders to their people, but the local practice diverge from the formal order. "The kebele council is elected and kebele leaders are in theory an executive committee of the people's directly elected representatives. In practice, however, the kebele are controlled by cadres of the governing party, who make sure that kebele chairman and leaders follow closely the EPRDF line" (ibid.: 232). The kebele's distribution of land rights is a political process: Those not following the governing party's line, i.e., EPRDF, are in best case overlooked in the redistribution process. Through the local cadres the governing party manages to control its area and people, and in order for the state to control the pastoral people the state tries to settle the nomadic groups. Sedentarisation programmes are an important means for the government to gain control over this part of its population.

The issues of sedentarisation are reflected in Ethiopia's recently ratified Poverty Reduction Strategy Paper (PRSP), i.e. SDPRP. Though the PRSP is supposed to be an autonomous product of Ethiopia's government, it comprises the policies and interests of several actors which have participated in making the Ethiopian SDPRP.

SDPRP: a cacophony of interests

The general approach to all contemporary development activities in Ethiopia is stipulated in the Sustainable Development and Poverty Reduction Program (SDPRP). Formally, the SDPRP is a genuine product of the Ethiopian government, but a multitude of actors were involved in the processes that in July 2002 finalised the SDPRP. The SDPRP is an IMF- required Poverty Reduction Strategy Paper (PRSP). To be able to get debt relief, loans or funds all countries that collaborates with either IMF or the World Bank need to have an IMF- approved PRSP. In interaction and cooperation with multilateral development agencies, it is of primary importance for the recipient country to have an approved PRSP, and the different donors, particularly the IMF, have great influence over the final product.

Though intended to be an autonomous product of the Ethiopian government, the SDPRP comprises several actors' interests, which in the end are formally ascribed to the Ethiopian government. All donors (governmental, NGOs, international) of Ethiopian recipients had the opportunity to state their interests in the process of defining the SDPRP, which in the end was to be approved by IMF.88 Some Norwegian NGOs stated their own case. In general, however, both the Norwegian government's and Norwegian NGOs' interests were communicated through the NORAD run embassy in Addis Ababa. A representative from this embassy stated that '…it is hard to get through among all these interests and agencies that pull in all directions'. All development organisations from all of Ethiopia's donor countries were included in this process, as well as the multilateral agencies. Though the SDPRP is ascribed to the local government, it is also made up of several external actors' interests. The various donors' interests and chance to promote these are strong, due to their means of sanction, particularly IMF, which in the end approves the document.

The final output of the process came in July 2002, as the Sustainable Development and Poverty Reduction Program (SDPRP) which stipulates Ethiopia's general policy on development issues. As the SDPRP states, it is not a blueprint on how to become developed, or even how to do development. Rather, it is a policy statement that stipulates which issues to be addressed. The Ethiopian government has, in the SDPRP, "…formulated policies and strategies to guide over all development with focus on rural and agricultural development" (SDPRP, 2002: i). As shown above, pastoral communities are, in the SDPRP, defined as a problem that need radical transformations to secure sustainable development, which is to be achieved by settling the pastoral populations.

Diverging results and effects

All development actors operating in Ethiopia need to work in accordance to the SDPRP. Norwegian NGOs must build upon the pillars of NORAD, and they are both committed to adhere to the development policy of Ethiopia, as all actors in Ethiopia are obliged to. Because of the need for formal coherence between the Ethiopian government's and an NGO's approach toward pastoral people, one might assume that the elements the interventions rely upon, and the effects and results produced, would be similar. However, they are not. In their approaches, both actors generate the same effects as Trouillot (2001) describes, but do not rely on the same elements as stipulated by Scott (1998). This is due to differences in size, and thus the complexity of the field the two actors set out to intervene in. Whereas the IPDP focuses on a rather small group in Aba'ala wereda, the government approaches all pastoral communities in Ethiopia. One cannot assume that the effects or the inevitable results of 'tragic state initiated social engineering' are the same with the two different cases, due to the enormous gap between the respective actors' target fields. Additionally, local practical knowledge (metis) and actors' agency are transformative elements between the provided input and expected output. Planning relies on simplifications and the production of legible units. The larger and more ambitious the plan is, the more reductionistic and simplistic planning becomes. Thus, the larger a plan is, the more of the ethnographic particularism it tries to grasp is missed in its representations. The divergence between formal order and local reality increases, just as the divergence between the map and the terrain does when scale increases.

Summarising remarks

Planned development interventions are schematic, and rely upon simplified representations of the target area and group, as well as of the strategies and practices that are to be used to achieve the stipulated goals. Representations are means for planners and implementers to promote legibility and establish causality between input and results, and between the various project components. The need for legibility is as important for the state as for NGOs. Planning from above is a feature of both state initiated social engineering and NGO funded development projects. Both actors' fields for intervention are complex, and both actors reduce the complexity in producing representations. The similarities in planning and intervention make the state and NGOs produce similar effects, which illustrate the parallels between state and NGO initiated work. The IPDP also faces problems similar to the state's in approaching pastoral people, since pastoral people regularly move around. Since the IPDP does not have the ability or power to impose settlement plans, it ends up mainly addressing people already settled in the area.

As shown above, the IPDP produces similar effects as the state, but as two out of three partners in the IPDP are governmental organisations (GO) one might not find it special that the IDPD creates state-like effects. Nevertheless, DF is the main architect of the IPDP. Despite including GOs, the IPDP is denoted as a community based NGO project that operates separate and independent from the state and its interests in Aba'ala by DF, MU and DoA. The state funded initiatives in Aba'ala are scarce, and state institutions rather small. The IPDP's budget is almost 30 times the size of DoA's. In many ways the IPDP almost becomes the state in Aba'ala, since the IPDP is the provider of expertise and funds to DoA. But whereas DF is dependent on MU and DoA as implementers and to get access in Afar, MU and DoA are dependent on DF for funds, which show that GOs and an NGO can be largely interdependent.

The next chapter gives an account of relationships between global and local knowledge, exploring the consequence of encounters between a global development discourse and local practical knowledge. The notion of a global hegemonic worldview is contested and prospects of interface situations are provided.

Chapter 6. Globalisation and development: homogenisation or hybridisation?

Development, whether perceived through the optics of modernisation, dependency, or post-development theory, refers to intentional external intervention in local or national social, economic, political, cultural and demographic processes. Whereas development represents itself in terms of beneficial intentions, post-developers, who position themselves outside development's institutional structure, have a less flattering opinion about development. Post-developers see institutional development

"…as a bureaucratic force with global reach and an explicitly pro-capitalist agenda, operating as a tool of regimes that seek to perpetuate relations of inequality and dependence between the West and the rest and, through their representation, to perpetuate the construction of others as post-colonial subjects" (Green, 2003: 124, emphasis added).

This view proposes that development represents a hegemonic and monolithic discourse that overrides cultural variations wherever it is communicated and implemented by its advocates or development agents. As so far empirically shown, what happens locally is not solely determined by an imposed global development discourse, and despite the apparently standardisation of development projects "…those involved in planning development are well aware of the limitations of what they are trying to achieve…" (ibid.).

This chapter takes as its primary concern the processes involved with and aligned to development and globalisation. The encounter between development experts' knowledge and local practical knowledge, as articulated by local development agents, represents the encounter between a global development discourse and local knowledge. Situations of interface generate local counter-tendencies and informal coping strategies towards the imposed ideas. Consequently, one can neither assume that what is imposed will be the unbiased outcome, nor that the various places and actors will generate the same counter-tendencies. Responses are co-determined by local context and practical knowledge. Aligned to issues of development and globalisation, this chapter raises questions about the relations between what is perceived as global and local. Global processes and ideas are articulated locally, and the articulator affects what is and how it is communicated. In order to get a grip on what happens in various localities, one needs to consider relevant external and global issues that contribute to the constitution of the local. The local and the global are interconnected, and the local-global dichotomy is solely of analytical value. There are no clear-cut boundaries between what is conceived as local and global since what appears to be local can be included in the global and the global gets local expressions as it's transformed and articulated. Anthropologists do not study villages, but what happens in villages. "What happens in one locality influences what happens in the others, whether that is intended and can be foreseen or not" (Hannerz, 2003: 21).

This chapter is twofold. It explores local-global relations in the context of development and it discusses what the processes of development entail. Does the interface and encounter between different realms or discourses lead to increased standardisation or does it stimulate to local creativity and entrepreneurship? Another interesting question is whether a global development discourse and donor's power necessarily prevail over the recipients and the local practical knowledge. Acknowledging the multitude and range of definitions on globalisation and other aligned words, Eriksen's conception that "[g]lobalisation consists in all processes that make distant irrelevant"89 is adopted. These processes "… take place within nations but also transcends them, such that attention limited to local processes … and units of analysis yields incomplete understanding" (Kearney, 1995: 548). Globalisation also denotes "…the goal of a process of building transnational political and cultural ties … through [amongst others] NGOs" (Tsing, 2002: 457). In questioning what such processes of globalisation do in the world, this chapter shows that one can "…investigate globalist projects and dreams without assuming that they remake the world as they want" (ibid.: 456). The local-global dichotomy has no empirical validity and is merely of an analytical matter.

Development: cultural flattening or local creativity

The departure point is that cultural globalisation takes place and, according to many scholars concerned with globalisation, it has for a while (see e.g. Appadurai, 2002; Tsing, 2002; Schuurman, 2001; Albrow, 2001; Eriksen, 2003). When globalisation started is not of that much importance for the present argument. Processes of globalisation have come about for a long time, and during the last 50 years there has been a significant increase in the processes' velocity and outreach in space. Contemporary globalisation distinguishes itself from previously due to the "…intensified and accelerated contact across geographical boundaries, mediated by information technology…" (Eriksen, 2003: 225), or what Harvey calls time-space-compression (Harvey, 1989).90 Due to the rapid increase in globalisation's velocity and volume, an interesting question is what the effects of these processes labelled globalisation are.

De Ruijter (2001) depicts three possible future perspectives on culture resulting from globalisation processes: Cultural convergence, divergence and hybridisation. Cultural convergence refers to a process of growing sameness. This perspective "…represents the classic vision of modernisation as a steamroller denying and eliminating the cultural differences in its path" (ibid.: 35), also known as "MacDonaldisation' (ibid.). The convergence thesis proposes that globalisation will lead to cultural standardisation and uniformity in accordance with the most hegemonic worldview due to growing global interdependence and interconnection. Whereas the first viewpoint refers to a process of homogenisation, the second perspective, divergence, accentuates precisely that which is ignored in the convergence thesis; the differences. In the encounter between two separate realms, the differences and distinctions are emphasised. This leads to a strengthening of the cultural differences between the two encountering systems of knowledge. This view holds that cultural differences are immutable and inevitably generate rivalry and conflict. The third perspective states that globalisation results in cultural mixing or hybridisation. "Hybridization emphasises the idea that the global powers are - and will always be - quite vulnerable to very small-scale and local resistances" (de Ruijter, 2001: 37). Hybridisation acknowledges that "…communities are always in flux, divided, contested; people are perpetually escaping them as well as mobilizing to enforce them" (Kalb, cited in de Ruijter, 2001: 37). Eriksen (forthcoming, 2004) states that "[h]ybridity directs attention towards individuals or cultural forms which are reflexively - self-consciously - mixed, that is syntheses of cultural forms or fragments of diverse origins" (13-14).

The question of whether the processes of globalisation "…lead to increased creativity or to a general 'flattening'" (Eriksen, 2003: 225) is also relevant to ask in relation to development: Does development and the encounters between various systems of knowledge that it entails result in local creativity and hybridisation, or does it denote a homogenising process which lead to cultural flattening? The genealogy and contemporary manifestations of the development discourse validate to denote it as a global discourse. The panoptic development apparatus, with its standardised development problems and solutions, can be seen as one system of knowledge that tries to penetrate and prevail over other systems of knowledge. The development discourse is in various ways adopted, reproduced and diffused by development actors and agencies all over the world seeking to deploy it in various settings. In the deployment area, global development discourse sometimes becomes adopted, contested, transformed or rejected.

The donor-recipient relationship binds global development discourse together with a locality and entails the interface between different systems of knowledge. Before further elaborations around occurring counter-tendencies in situations of interface to shed some light upon processes of globalisation, an account is given on the standardisation of development projects. Standardisation is prevalent in the formal order and discursive formation of development. The apparent standardisation is due to the methods and practices applied in planning, implementation and identification of policy and objectives. As argued earlier, the formal is not necessarily what is going on locally. I show two different results of development and globalisation which are dependent on where focus for the analysis is put. This has implications for how the formative power of structures and discourse are perceived, and consequently also for the understanding of globalisation processes.

Development discourse and standardisation

Referring to Arce (2000) and Ferguson (1994), Green argues that

"…while the standardization of development globally is partly a function of the standardization of development problems and solutions, …it is also a consequence of the kinds of practices used to plan and implement development as a process of transforming policy visions into manageable realities through the social constitution of 'projects' subject to specific techniques of audit, organization and control" (2003: 123-4).

The standardisation of development is mainly due to the process of planning and the stipulated means and objectives. Green's quotation resembles elements of the Integrated Pastoral Development Programme (IPDP). The IPDP planners rely upon representations or simplifications of the field in order to operate with legible units possible to plan on and intervene in. In regard to the policy and choices made concerning project objectives and components, the IPDP must cohere to NORAD's, as the back-donor, policy and wishes. Planning and choice of project components are interrelated. The formal presentation of the IPDP states that the project is community based and that it applies participatory planning in order to empower the beneficiaries to develop and design the project themselves. The case with the consultant (cf. chapter three) shows that in planning for a second phase, the target group was put on the sideline in the planning process, because the consultant both knew the format and had received policy input from the Development Fund (DF) to direct the process and the output.

The actual lack of bottom-up planning and the accentuation of policy coherence contribute to the standardisation of development problems and solutions. Donor countries' agencies (e.g. NORAD), multilateral agencies (e.g. IMF) and donor-NGOs (e.g. DF) develop the policy and coordinate with each other in order to promote policy coherence toward the recipient parties. How developers work, plan, and the objectives they pursue "…constitute the profile of modernity, or the shape of what, in the context of the West, has often been called the Modern Project" (Albrow, 2001: 22).

The modern project of standardised problems and solutions

Development is a 'Modern Project', and both capitalist and state socialist countries ascribe to it. The profile of modernity is recognised in policy-makers', planners' and developers' yearn to gain control over their defined target field. Emphasis is put on technological advance, the increasing power of productive processes, the importance of reliable administration, the need for a trained and disciplined workforce, the desirability of growth and visible results, and the faith in planning for it (ibid.; cf. chapter five; Scott, 1998). In the West, modernity is normally equated with the modern project, but "…modernity does not simply have a project. In the West modernity is the project … [because] it is self-activated, conceived, and possessed by human agents who are in control of themselves and it" (Albrow, 2001: 22). Albrow notes that modernisation is supposed to happen to the target group when developers intervene. It is an ambiguity between modernity, which is supposed to be self-activated, and the notion that it is imposed by developers through participatory approaches. Development policy reflects more the ideology of western donor agencies than the actual wishes of the recipients.

The development discourse is maintained by the increased and standardised flows of development policy globally. A central element in this respect is the Development Assistance Committee (DAC) of the Organisation for Economic Co-operation and Development (OECD). The IPDP relates to DAC through DF and NORAD. The point of DAC classification in the NORAD application form underlines DAC's role and demonstrates the standardisation of development problems and solutions. Whereas the IPDP has to reflect and underpin NORAD's policy, NORAD's policy shall cohere to DAC's, in which NORAD constitutes one of the core members. NORAD's and DAC's policy are interrelated, but DAC's policy largely prevails over NORAD's, as NORAD is merely one of the members that contribute to deduce DAC's guidelines. According to DAC itself, DAC's role is to monitor and coordinate the development assistance provided by its member agencies, "…developing a series of principles in the key area of aid management" (DAC, 1992: 5). The "…Member aid agencies, the World Bank, the IMF and the UNDP … bring together the results of the work done under DAC auspices on essential aspects of the aid planning process" (ibid.). DAC notes that

"DAC Members see the principles incorporated in the manual not only as guidance for the orientation of their own aid policies, but also as a basis for dialogue and co-operation with developing countries in order to improve the effectiveness of the totality of resources devoted to the development effort" (ibid.).

DAC, being the fellowship of western governmental aid and donor agencies, stipulates principles and guidelines for development cooperation in order to secure consistent, coherent and homogeneous development assistance.

DAC plays an important role in general policy-making and policy coherence. Regarding the IPDP it is illustrated in chapter three with the introduction of 'HIV/ AIDS Prevention Initiatives' and 'Good Governance'. In the NORAD application form, all recipient organisations are asked to statistically classify the project's relevance in relation to some key sectors targeted and emphasised by DAC. Through the application, the data is submitted to NORAD. NORAD forwards the data to DAC which monitors all different projects funded by its member agencies. For the 2001 application new elements for DAC classification in NORAD's application form were added: 'HIV/ AIDS Prevention Initiatives' and 'Human Rights' were added as objects for DAC classification. In addition 'Good Governance' was included as a separate point in the application form (point 3.6.). Subsequently, these components were included in the IPDP, as well as in many other DF funded projects. In the case of the IPDP, good governance was ascribed to the capacity building component and the IPDP's auditing and evaluation system, which, as a requirement from the donor, intentionally are to reveal any economic irregularity and misconduct by the officials in the local implementing organisations. The HIV/ AIDS awareness component was added, even though it overlaps an already existing governmental HIV/ AIDS programme in the area. In retrospect, an IPDP board member states that "The whole process was strange. We were just told to do so. We had to expand the project, but the budget remained the same. And what's the reason with two HIV/AIDS programmes in Aba'ala".

The inclusion of good governance and the HIV/ AIDS component illustrate the ambiguity between bottom-up planning, policy coherence and donor-imposed conditionality. It thus shows an internal contradiction within the development apparatus between, on the one hand, the general objective of policy coherence, and on the other hand the goal of empowerment and participatory, bottom-up approaches. The implementation of the latter would entail that projects' and documents' legibility is reduced as the complexity and ethnographic particularism increases. Who to ascribe the inclusion of the new components is difficult to identify. Nevertheless, this illustrates the lack of bottom-up planning. The inclusion of HIV/AIDS prevention initiatives and good governance reflect the trend that took place within the development discourse at the time. Development policy and projects are characterised by rather short-term planning due to the continuous change of buzz themes or dogmas (cf. Tvedt, 1990; Tvedt, 2003; Nustad, 2003a). The IPDP project manager's concern about yearly applications, despite the IPDP has acceptance for a five years phase, can be accounted for as a controlling mechanism from NORAD. This also enables NORAD to give new input and guidelines annually instead of every fifth year.

The inclusions of the new DAC emphasised target fields show how the standardisation of development globally is a function of the standardisation of development problems and solutions. By defining the problems, developers and policy makers also come up with solutions. Referring to the sudden inclusion of the HIV/ AIDS component an IPDP board member says that the HIV/ AIDS problem gets disproportionately much attention despite the extent of the problem in Aba'ala. He emphasises the importance of generating awareness around the threat of HIV/ AIDS, but since the issue is already approach by the local governmental offices he states that it would be better if the IPDP approached general health issues or the problem of e.g. malaria, which in certain areas tend to be hyper epidemic.91 This is the same board member that previously expressed that the gender issues and women in development component of the IPDP reflect more the policy of the donor that the interest of the beneficiaries. Women and gender issues are also key target areas of DAC. To ascribe the inclusion of new IPDP project elements to either DAC or any other organisation or institution would be simplistic. However, the cases show prevailing power structures and policy interests at stake, but most notably the lack of bottom-up planning, which the IPDP states it applies. The complex global network of relations of power in terms of funding and ability to define policy are important aspects of how development is defined and implemented. Despite the difficulties in identifying which organisation(s) or policy maker(s) to ascribe alterations in development policy, it nevertheless illustrates the general lack of engaging participants.

The standardisation of problems and solutions show the incorrectness of classifying the IPDP as a bottom-up project. The standardisation of development and the lack of participatory approaches are not only due to the defined paradigms of development problems and solutions, but are also consequences of the practices and methods used to plan and implement projects (cf. Green, 2003).

Standardised planning and implementation

By intention, development planning involves mainly beneficiaries. In practice, various stakeholders are engaged, from back-donor (NORAD), donor (DF), implementing organisations (MU and DoA) and the beneficiaries, which all are attributed with unequal relations of power. Despite the beneficiaries, through an empowerment process, are to have control over the IPDP's design, the actual case is that the general policy and the components are largely a priori defined by the DF and the guidelines given in NORAD's standardised application form.

The way the IPDP is planned and designed is articulated in various project documents, which rely on representations and simplifications of the field. Planning is characterised by great faith in linear, scientific and technical progress. The construction of the field as a manageable object enables the planners to generate causal relations between the provided inputs and expected results. The main purpose of project documents is to indicate and validate these causal relations established in the planning process. Project documents are

"…vital in constituting the project as a slice of manageable reality where project inputs can be seen to relate to outputs and, largely by inference, to effects on the ground. Impacts are in fact inferred from the outputs having been achieved through what are known as 'output-to-purpose' reviews, that is when at the end of a specified time period assessors try to determine the extent to which the achievement of the outputs stated in the original plan has contributed to the 'purpose' of the project" (Green, 2003: 129-130).

Project documents serve to create mono causal linkages between inputs and outputs in project planning to legitimate the spending of funds towards the (back-)donor. This is due to what Green calls the "…amorphousness of social reality and the virtual impossibility of determining with any certainty the actual relation between what a project does and other social processes…" (ibid.: 129).

Whereas chapter three below show the representations the IPDP relies upon, chapter five show the range of use of these simplifications. To measure the effects, the IPDP application stipulates indicators to evaluate to what extent the objectives have been achieved. The indicators are largely quantitative factors as e.g. percentages, numbers, frequencies, and durability. All are legible units possible to monitor, measure and compare. The means to measure the IPDP's success also reflect both how the planning is made in advance and how the project is implemented since the IPDP is managed by objectives. The planning of the second phase, cf. the case with the consultant, illustrates how the participatory approach is bypassed. The consultant, who works according to the guidelines of DF, rather relies on the established representations of the field when making the new application. The representations are as a consequence recycled.

The means applied to measure and monitor the IPDP are generated in the planning and design of the project. The legible units in the project documents enable the donor to see whether the stipulated objectives have been achieved or not. The documents are designed92 so that the donors whom sit detached from the project area can read and control the project. Funding agencies are "…concerned with the relation between investments and outputs over a specific time-frame" (Green, 2003: 129). In addition is the IPDP accountable93 towards DF. The IPDP's 'checks and balances systems'94 regulate and control its financial resources and the quality of development work. MU conducts annual internal audits of the IPDP, which is a NORAD requirement forwarded by DF. Nustad argues for audits to be seen as "…instrument of power … closely related to what Foucault [2002] described as governmentality, the self-discipline achieved through creating knowledges and subject positions in such a way as to ensure that the objects of knowledge monitor themselves, without the state needing to keep an eye on day-to-day operations" (2003b: 130).95

The way the IPDP and development projects in general are planned, designed and implemented, as well as how the development industry contributes to define development problems and solutions show that it is plausible to talk about an apparent standardisation of development in general which fits into and sustain the development discourse. A strict implementation of the formal order and firm application of the stipulated guidelines would probably resemble a process of convergence, standardisation or homogenisation. However, as development projects imply the encounter between different systems of knowledge which generate counter-tendencies, one cannot assume a priori the outcome of the processes involved. Whereas a standardisation of development concerning the IPDP is articulated and observed with reference to the various formal documents produced and exchanged between DF and MU, a less rigid and hegemonic result is identified when local situations of interface are taken into account. Trouillot, writing about frontiers and borders in anthropological theory, states that "[n]o discursive field is fully 'ours' or 'theirs'" (1992: 25). Discourses can be learnt and shared through texts, experience and practices. Reflexivity towards a system of knowledge largely depends upon the actor's degree of embeddedness into that particular discourse. Local development agents' notion that they not only are a suppressed part of but also interconnected with an external development apparatus and discourse enable them to be reflexive. This also underlines a distinction between what one could call two different forms of globalisations. The first is objective globalisation, which is as old as the hills (e.g. cultural meetings due to travel or migration). The other is reflexive globalisation, which relies on actors' recognition that they are interconnected across space. The latter underscores that actors know that they necessarily must relate to the international system of development, the development discourse and its apparatus. Reflexive globalisation is 'newer than the hill', and its size and significance increase continuously.

Actors' interaction with a global development discourse denotes a situation of interface that can produce counter-tendencies. The local and the global represent two analytical realms, and globalisation processes have to be seen in relation to the various sites where these processes are articulated and communicated since "…once the world is acknowledged, one must deal with 'local response'" (Trouillot, 1992: 34).

Agency and hybridisation

Do globalisation and development lead to a general cultural flattening or cultural creativity? Post-development scholars' approach to development proposes a development discourse that does not reflect, but rather constructs the realm developers relate to, and thus their practices. Consequently, the realm development represents are regarded as a hegemonic, monolithic and homogenising discourse (cf. Kiely, 1999) which effect is a cultural westernisation and standardisation. The apparent standardisation of development identified by post-development scholars is due to the macro-orientated approach and the disproportionately large focus on formal order without taking local actors' relation to the development discourse and the application and appropriateness of it into account. The formal order and organisation of development are not irrelevant to what is happening, but it is important to underscore that the formal order is not the only thing that is happening. If focus is pitched from a structuralist to an actor-orientated approach, other empirical data are identified that do not directly reflect the formal order. This also lower the formative power ascribed to discourses and structures. Local actors' knowledge of the discourse they encounter enable reflexivity and the generation of informal coping mechanisms because they have the notion they are interconnected with DF, rather than being the suppressed part in an unilateral partnership. Thus, the deployment of development discourse does not necessarily represent an uncritical acceptance of a formal order among recipient organisations, but signifies that the imposed discourse gains local expression through a process of transformations and contextualisation.

The cases depicted in chapter four show how different actors engaged in the IPDP act and relate to the development discourse. The cases also show that knowledge about the imposed guidelines enables local development agents to navigate within or on the borderline of the formal order of the 'partnership' relation between DF and MU/DoA. The first case shows that the DF representative acknowledges some flexibility and informalities in the general IPDP project management. Informalities evolve and are applied as means for the local project manager to bridge between the development discourse and the local reception of the IPDP. The DF representative states that she knows the difficulties in working with pastoral people, whereupon she acknowledges the importance of informalities in project management, despite the possibility that they might diverge from the formal order. She argues that the flexibility applied in project management is necessary to secure contextual development and to get acceptance from the local community. The relative success of the IPDP is due to informal practices and deviations from the formal order. Nevertheless, the DF representative stresses that it is the last year of the current phase and to secure further financial support, previously postponed or delayed activities must be accomplished. The DF representative states that DF so far has been flexible, but now DF needs to get the project back on track, primarily due to the forthcoming evaluation and the application process for phase two. The donor is not only flexible in handling the formal order towards NORAD, but also flexible with the guidelines DF assigns to MU. This illustrates not only that informal practices are important for the fulfilment of the formal order, but also that the donor side acknowledges the constraints and difficulties with strict implementation in a difficult area.

The second case illustrates that the consultant's knowledge about NORAD's policy, guidelines and the general NORAD format enables him to be eclectic in his use of the development discourse and the tasks he is to perform. The consultant works according to an agenda stipulated by DF, which states the expected outputs of his work. The consultant's desire to achieve these defined outputs results in bypassing the participatory approach. The IPDP project manager is concerned with how the planning is carried out since his arranged household survey was abandoned because he had to wait for the consultant to arrive. When the consultant arrived, he took over the role of the project manager in preparing the application for the second phase. The consultant's melange of hypocrisy and cleverness due to his knowledge of the NORAD format and the development discourse enable him to produce an adequate application despite essential features of the formal planning process are evaded.

The third case with REST shows how their knowledge about the development discourse enables them to be reflexive towards the knowledge transfer and guidelines from their many donors. The case shows that the knowledge of the development discourse can be learned through interacting with it and consequently be manipulated at local level. REST has gained lessons about donor-recipient relationships, donor's policy and conditionality, and the development discourse in general through their longstanding position as an NGO. Through the acknowledgement of the implications of and the inequalities in power between donor and recipient, REST manages to transform certain features of the development discourse and donor's policy to increase the local appropriateness.

The inclusion of the components of HIV/ AIDS prevention initiatives and good governance also denote different creative processes where certain features of development's formal order are bypassed. The HIV/ AIDS component is merely added to the existing components without changing the defined objectives of the IPDP. None of the IPDP's four main goals says anything about reducing or preventing the threat of HIV/ AIDS in the area. The inclusion of good governance shows a different strategy applied. Already existing activities of the IPDP are redefined and renamed to address the new themes approached and questioned by NORAD. The IPDP's work to promote good governance is ascribed to the checks and balances system, the capacity building and training of the governmental administration in Aba'ala and the general auditing. DF incorporates these two new elements into the IPDP in order to fulfil the implicit guidelines from NORAD. DF and the IPDP are not compelled to contain these components, but the inclusion of and emphasise put on them in NORAD's policy and the standardised application form made DF embrace them. The IPDP's absorption of the issues that are given precedence by NORAD increases the possibility of further economical support. This case shows the use of informal strategies in the design of the IPDP among DF's staff in Norway. DF's intermediary position between NORAD and MU enables them to contextualise NORAD's policy and guidelines before forwarding them to MU, while simultaneously helping MU in making the applications more attractive before consigning them to NORAD. The creativity applied in DF's relation to NORAD shows not only a means to make the IPDP formally cohere with NORAD's policy in order to secure funds, but also illustrates how the development discourse is reproduced, since the outcome of the creative practices feeds into the larger whole of the development discourse.

The HIV/ AIDS prevention initiatives component consist of radio broadcasting, education and poster production. There is no radio transmitter in Aba'ala, and transmitting signals to Aba'ala would probably be very difficult due to its geographical position in the bottom of the escarpment area. Additionally only a limited number of people have access to radio. What regards the poster production, I never observed any posters in Aba'ala that addressed HIV/ AIDS issues. Regarding education, the local schoolteacher often asked why the IPDP does not support the local school. This is also an issue among many IPDP board members. They want to support the local education programme, which corresponds to the wishes of the local people. The household survey shows that a great majority see lack of education as a crucial constraint, and the respondents state that they would like the IPDP to include an education component. This has not been done. A board member explains that it is because DF says that an education component will infringe with the IPDP's stipulated objectives.

However, as the IPDP had money left on its budget (because of the delay and lack of implementation of some activities they already had received funding for) the board decided to give money to repair the local school. The school requested the IPDP for money, and "[a]fter discussing the request by the Steering Committee members, 60% (the other 40% was contributed by DHP) support was given".96 This was done without conferring with DF. Later, this was accepted by DF. Their acceptance can be understood in two ways: The disbursement to the school can be legitimised as being part of the HIV/ AIDS component's education activity (despite the money was given to rebuild the school). Additionally, DF would not have any problems in legitimising this towards NORAD, as the Norwegian development policy at that time emphasised investments in education programmes. As the inclusion of the HIV/ AIDS component to the IPDP must be seen in relation to the previous Minister of Development's emphasis of this issue, DF's acceptance of the grant given to the school can be seen in relation to the present minister who emphasises and promotes education initiatives.

A common denominator of the cases referred to above is that they all render information about how development agents and recipients relate to their respective donor's requirements and development policy, and the development discourse in general. The cases also show the counter-tendencies that occur due to the many encounters between various systems of knowledge, which become identifiable with the application of an actor-orientated approach. Development agents are able to translate, contextualise and utilise the development discourse through their reflexive knowledge about the donor's policy and wishes. The development discourse does not denote a separate type of knowledge to local development agents' practical knowledge. These two systems of knowledge draw on each other. They can change as they are articulated. The development discourse is not a hegemonic knowledge that has a uniform impact wherever it is communicated. Actors will always "…attempt to come to grips emotionally, cognitively and institutionally with the various 'externalities' they confront" (Long, 2001: 220). The encounters between various systems of knowledge where so-called 'weaker' or 'subordinate' actors create space for themselves are characterised by different local modes of creativity and mixing.

The development discourse represents only one among many systems of knowledge that development actors relate to. Another common denominator of the cases referred to above is that informal practices applied as coping mechanisms towards the formal order also contribute to reproduce the development discourse. The informal practices largely contribute to confirm the project and the formal order in the eye of the donor, despite these strategies diverge from the expected and codified practices. Informalities are made possible exactly because they are not reflected and reported in formal documents.97 This illustrates that the formal order is parasitic on informal practices (cf. Scott, 1998) and that there is no direct equilibrium between the formal representations or order and what actually happens, despite the inter-relatedness between the two. Actors' knowledge about what is perceived as formal render possible to reproduce the formal order with informal practices. This shows the interconnection and that, to re-quote Barth, "…formal organisation is [not] irrelevant to what is happening - only that formal organisation is not what is happening" (Barth, 1993: 157).

The first case shows that the DF representative needs to get the project 'back on track' after the endorsement of the flexible project implementation in order to re-establish the IPDP's formal order. The case with the consultant shows that informal practices are applied to produce a new application. Despite the informal practices diverge from the formal order, the generated product confirms and reproduces the development discourse. The case with REST shows that REST's acknowledgement of development discourse enables them to be reflexive. Acceptance for a project was obtained because the donor's wishes were acknowledged in the project documents. The inclusion of the HIV/AIDS and good governance components to the IPDP shows that also the IPDP's donor, DF, is strategic towards NORAD. The case with the money given to the rebuilding of the school also illustrates this. Despite that the IPDP board infringed the formal order, this was later turned into something positive towards NORAD since the result of it cohered to NORAD's prevailing policy about education. By confirming NORAD's policy, they feed into the development discourse. The informal strategies and practices are made possible precisely because they relate to the formal order.

Reflexive interface and reproduction of the formal order

What type of mixing is at stake in the encounter between a monolithic development discourse and local practical knowledge as articulated through actors? Eriksen (forthcoming, 2004) tries to get to grips with different forms of mixing. The melange of knowledge is best approached through identifying "…the ambiguous grey zones, which can be located to the space between categories and boundaries…" (ibid.: 2), that is, in situations of interface. The encounter between donor and recipient in the development sector represents such an ambiguous grey zone where knowledge is continuously challenged, contested or strengthened. The sites where transnational relations are articulated are privileged when studying the interplay between discourse and actor, and the global and local. The development discourse represents a global, universal and uniform knowledge, which is challenged with the unique particularity of the various localities it encounters. The IPDP is such a site where the juncture between what is conceived as global and local transpires.

The cases depicted above all show that there is an interaction and creative mixing between the development discourse and local knowledge, articulated by the actors involved. The self-awareness of development agents enable them to be reflexive to and aware of the principles forwarded by the donor, being it DF or NORAD, and the differences in power this entail. Referring to hybridisation, Eriksen writes that "[s]ometimes people are acutely aware of changes taking place in their immediate environment, and take measure to stop it, to enhance it, or to channel it in their preferred direction" (ibid.: 13). Each of the presented cases shows how different forms of creativity are activated because of the encounter between different systems of knowledge. The first case illustrates how both the donor and recipient organisations become prisoners of the requirements of their back-donor, NORAD. The choices made are also forms for creativity with the objective to secure further funds for a second phase. The consultant in the second case illustrates a mix of personal cleverness, cynical manipulation and a hypocritical approach to his expected output though having a reflexive awareness to his tasks when he states that the most important is to secure further funding. The third case, with REST, shows an organisation's ingenuity in relation to external organisations and donors. DF's role in the inclusion of the HIV/AIDS and good governance components illustrates a creative and pragmatic conduct towards NORAD and that counter-tendencies might take place among the donor in its interface with NORAD. The case with the money given to the school shows that the IPDP board exceeded its formal position. Both the board (towards DF) and DF (towards NORAD) had to legitimise this action.

The cases all demonstrate different kinds of local small-scale creativity towards something greater, which lies in the demarcation between something locally wanted and something externally imposed. Each case shows an informal strategy; in common the cases show how something from the outside, a global development discourse, is transformed. The cases also illustrate how the development discourse and the formal order, which the actors relate to, are reproduced. Through the counter-tendencies and local informal strategies, the global development discourse is reproduced. Not as a monolithic and homogenising system of knowledge that penetrates the project and directly affects the actors, but as a discourse or realm development planners and donors relate to. The translations of that which comes with the flow from the donor takes place at the local level and is not reflected in the documents that are consigned to the donor. What is being transferred back to the donor (project documents, audits, reports, etc.) reflects, reproduces and confirms the development discourse. A strategy is to give the donor what they want and expect (in terms of formal documents) regardless of the practices applied to do so diverge from the formal order. The informal practices not only bridge, but also sustain the gap between the formal codified order and the complex and local multiple particularities. This is not contradictious. It is made possible because the informal practices and the output produced by these practices are articulated at two different levels. In this way, development discourse is reproduced. But as the discourse is reproduced, the possibilities for local actors to continue to apply these informal strategies are maintained. As informal strategies contribute to the realisation of the formal order, the formal order continues to be parasitic on informal processes (cf. Scott, 1998: 310).

Creativity and hybridisation

The presented informal strategies and processes reflect a form of mixing that Eriksen denotes as hybridisation. Eriksen states that "[h]ybridity directs attention towards individuals or cultural forms which are reflexively - self-consciously - mixed, that is syntheses of cultural forms or fragments of diverse origins" (forthcoming, 2004: 13-14). Hybridisation opposes creolisation. The latter is a form of mixing that suggests the "…presence of a standardised, relatively stable cultural idiom…" (ibid.: 14) as the result from an enduring social encounter between at least two parties with mutual influence. The former emphasises actors' creativity in their encounter with an alien realm or discourse through applying eclectic and creative strategies. To get to grips with the concept of creativity, Eriksen quotes Salman Rushdie, who states that "[a] bit of this and a bit of that; that is how newness enters the world" (2003: 223; see also Eriksen, 1999: 9). Whether something new 'enters the world' through the IPDP's development agents' informal strategies is questionable, but surely by utilising 'bits of this and that' they can be characterised as bricoleurs, or entrepreneurs which Eriksen defines as "…an individual who bridges formerly separate realms, thereby creating something new" (Eriksen, 2003: 224).

These processes of mixing are characterised as hybridisation. Hybridisation denotes reflexive awareness towards imposed and introduced elements. This is reflected among the local project staff associated with the IPDP. This says something about how one perceives local-global relations. Nustad (2003b) states that there is no local-global dualism. He argues that "[t]he global and the local is better understood as two perspectives that are applied to the same objects" (ibid.: 125). The global and local represent no dualism, and what are perceived as local can appear in a global setting and vice versa. This echoes the relationship between local practical knowledge and the development discourse. The two systems of knowledge are interconnected and interrelated. Global development discourse will always get local variations as it is articulated. The local-global dualism is merely an analytical approach that serves as a tool to arrange and understand the world. Hence, this dualism is no empirical truth.

Dualism, irrelevant distance and hyperspace

The local-global dualism is widely adapted by scholars engaged in processes of development and change. Often, there are inconsistencies in the widespread use of this dualism, since the same scholars simultaneously argue for the possibility of intrusion of the local into the global, meaning that processes of globalisation not only imply flows from the global to the local but also from the local to the global. Nustad, stating that empirical material does not reflect the local-global dualism, argues on the interrelatedness between what is perceived as global and local. He argues that globality is an aspect of the local, and "…that what appears to be local is 'essentially included within the global'" (Nustad, 2003b: 123).98 As implied throughout this thesis, the local and global are interconnected, which illustrates that there is no dualism between the two. The terms serve, however, as analytical concepts to distinguish between two factors that empirically feed off each other reciprocally and are interrelated.

Nustad argues on the epistemological and ontological value of the local-global dichotomy, stating that "…the global and the local are creations that inhabit our theories, and that the distinction finds no reflection in reality" (2003b: 125). Though interesting, this view is hardly novel: It is rather concepts that scholars utilise in order to manage their data, as e.g. the development discourse is. Also relevant is the distinction between analytical etic and emic terms, i.e., the terms and meaning as used by the researcher's informants. Some of the agents associated with the IPDP use the distinction local and global, while other hardly know what is outside the project area and even fewer have a perception about something global. However, the local-global distinction is of analytical value and helps understand the IPDP initiated processes.

The donor-recipient relationship, what flows between these actors, and the processes this relation involves imply processes of globalisation. The distance between DF and MU/DoA is irrelevant for the processes taking place. The IPDP initiated processes take place regardless of the funding agency's presence. This is because the project is codified in documents to which the implementing organisations relate. The transnational relationship between DF and MU/ DoA and the documents it entails constitute the IPDP. The transnational flows between donor and recipient are two-way. What is sent from the recipient organisations to their donor (i.e., applications, reports, etc.) largely cohere to donor's requirements and expectations in terms of content and structure, which accord to donor's policy and guidelines (i.e., the policy statements, guidelines, etc.). This coherence is probably more prevalent in the eyes of NORAD who is the final decision maker, than in the eyes of DF which 'helps MU in making the applications more attractive'. In designing the IPDP, many actors in various places are involved, and some prevail over others regarding their influence on the final product.

Green states that "[l]ocations and local knowledge are simply not relevant to the construction of chains of causality and indicators of assessment that development constituted as project entails" (2003: 138). She argues that space, and thus distance, is largely irrelevant for the construction of development projects and that these spaces resemble non-places, or hyperspace. In many ways, project areas are also perceived as such spaces when approached through the project's formal order. Non-spaces or hyperspace

"…are places through which people pass in their capacity as individuals detached from social networks and obligations. …constituted through text instructing the user how the space should be used. … Both non-places and hyperspace are real places, in that they are territorially situated" (ibid.: 139).

Development spaces, as articulated in the formal order of development, are non-places that are constituted through social relations, texts and deterritorialisation. Distance is irrelevant for the processes involved and where the planned activities are supposed to take place. Local implementation relates to the documents regardless of where the documents are produced since they constitute the project formally. As distance is irrelevant, the processes involved can be denoted as globalisation.

However, and as pinpointed earlier, there are no direct causal relations between the formal order and local practice. Processes of globalisation get local expressions as they are articulated. As it is impossible to identify global processes on a global level, one are only able to identify the global processes as they are articulated in or through a site feasible to study. The notion of development discourse is similar. Both global processes and development discourse must be studied in the various sites they are articulated.

The two approaches to the IPDP, i.e., discourse and practice, do not illustrate two different phenomena, but rather two different perspectives to and comprehensions of the same element. "So one could argue for the local and global as well: they are not two processes in the world, but two perspectives on the same point" (Nustad, 2003b: 127). Approaching the IPDP on a formal ideational level, its connection to a global development discourse becomes prominent, and one could probably argue for processes of homogenisation. When including a study of the actors involved and their practices, a different prospect about the future and the processes of globalisation are presented. This is not only a matter of where the analytical focus is put, but also a matter of how the formative power ascribed to development discourse and the processes of globalisation in relation to actors and their agency is perceived. It also shows that the focus for the analysis feeds off and is interconnected with the perception of the formative power of discourse and global processes. This is an empirical question. This chapter, or the whole thesis for that matter, has shown various strategies and reflexivity among actors in relation to a global development discourse, which budge on post-development scholars' comprehension of the development discourse as a standardised, hegemonic and homogenising system of knowledge. Processes of globalisation get local expression as they are articulated. Accordingly, the results of these processes are not a priori given but rather need to be studied empirically. It illustrates that there "…is no standpoint from which a phenomenon can be grasped in its entirety" (ibid.). The analyst must make a call for methodological and theoretical triangulation (Denzin, 1989). In studying localities one must also include external factors that alongside internal ones make up the study object. Regarding the IPDP, one must both study what happens locally as well as have an understanding of the external factors that are articulated in and affect the locality. This underscores that what is perceived as global and local are interconnected and that there are 'several sites in one', as Hannerz (2003) states, because "…one cannot take the local field as 'given' any more" (ibid.: 21). Simone Abram states that

"[w]e could not 'explain' the local situation without recourse to a broader field which follows the effects of decisions in one place through their many transformations into decisions at another place. It is for this reason that locating fieldwork in the mental space of a policy, rather than the geographical space of settlement, makes accessible processes of globalisation, flows of concepts, and the networks that span the local and the global" (Abram, 2003: 146).

One cannot merely study something as a local product, but also needs to take into account the several sites, flow and external policy that shape the object where it is articulated. The global and the local are intertwined, and in a particular local setting several external elements and global aspects might be articulated and communicated. How the global is translated and articulated locally is dependent upon the context.

Local agency reproducing global discourse

The local informal strategies, which is the counter-tendencies arising from situations of interface, support the notion that something perceived as global are reproduced by local actions and initiatives. In the case of the IPDP, the transnational flow of global ideas from donor to recipient is not a process characterised by local resistance, rejection, acceptance or something imposed by external actors, but rather a process of translation. The translation, which could be seen as mediation between development discourse and local knowledge, helps to contextualise the project towards the beneficiaries. The translations are what make the project viable and accepted by the beneficiaries and the back-donor. Regarding the IPDP, the local agents solve the crux between different systems of knowledge through informal strategies towards the formal order. This renders possible the generation of an informal project space, which is identified around and outside the formal abstraction of the IPDP. This project space relates to both the beneficiaries and the formal order. The activities that infringe with the formal understanding of the project take place in this informal project space. Since the informal strategies take place locally and are not reflected in the formal representations of the project they do not intrude the formal order but rather contributes to the reproduction of the development discourse. Not only is the development discourse maintained as the formal order is confirmed through documents consigned to the donor, but the actors' opportunities for future informal strategies are also reproduced.

Summarising remarks

The answer to the question of whether development represents a process of general cultural flattening, or cultural creativity is not only a matter of where the focus is put and which theoretical approach that is applied. It is also a matter of the degree of formative power one ascribes to the development discourse and processes of globalisation. These two matters are interrelated. As shown above, the project exists on two levels, i.e. a formal level in documents and an informal local level. These do not necessarily reflect each other, but they relate to each other. Post-development scholars' would probably argue that development represents a homogenising process that consequently leads to a cultural flattening despite the developers' good intentions. This is mainly due to their general approach of the development discourse as a monolithic and hegemonic system of knowledge and their lack of considering actors' agency. If focus is put on where the development discourse is deployed and how local development agents relate to this system of knowledge, not only is a different prospect of the future depicted, but also a different view and understanding of the structures' and discourse's formative power is offered. The general comprehension of development as such becomes more nuanced. The informal strategies illustrate local creativity and hybridisation. Situations of interface stimulate to local counter-tendencies. Watching ideas travel from donor to recipient and focusing on the point of intersection and encounter "…[w]e observe a process of translation -not one of reception, rejection, resistance or acceptance" (Latour, cited in Czarniawska and Joerges, 1996: 19).

Local development agents' reflexivity towards the development discourse enables them to generate informal strategies to cope with the formal order. These informalities are not reflected in the various project documents to which the donor, who is detached from the project area, relate. The distance between local knowledge and practice, and the formal abstractions of them, makes the informal strategies possible. Project documents try to bridge this distance but as argued, the practises are not mirrored in the project documents. The practices do however relate (but not reflect) to the formal order. Consequently, the practises contribute to maintain the formal order to which they are counter-tendencies. This illustrates the interconnectedness between formal and informal practices and the interconnectedness between what is perceived as global and local.

Chapter 7. Concluding remarks

In this thesis, I have given an outline of some implications of a donor-recipient relationship within the development sector as they are articulated in the Integrated Pastoral Development Programme (IPDP). Development 'partnership' is more than an equal relationship between donor and recipient organisations. It also implies the encounter between different systems of knowledge. Regarding the IPDP, this is illustrated by the partnership between Development Fund (DF), Mekelle University (MU) and the Department of Agriculture (DoA). In this concluding chapter, I briefly reiterate the main arguments.

Issues and processes related to development are a complex field. In this thesis, I have illustrated various knowledge encounters, or situations of interface, that take place in development cooperation between a Norwegian donor NGO and its recipient organisations in Ethiopia. Knowledge is articulated and transferred in two ways; either through language (oral or written) or through actions and practices. I have discussed the knowledge encounter between a formal system of development knowledge and local practical knowledge. Whereas the former is palpable as it is codified and defined in various project documents, the latter denotes knowledge that can only come from practical experience, and is thus difficult to codify and record. As donor and recipient are detached, project documents are a vital mode of transferring knowledge between them. Local practical knowledge is reduced, and the informal practices are not reflected in these documents. The planners, when establishing the formal order of development, try to invoke this type of knowledge in documents when designing a project. However: It is impossible to formalise and codify local practical knowledge unambiguously, not only due to its complexity, but also because of its continuous change. Consequently, attempts to make this knowledge tangible become characterised by simplifications and representations. Despite the inevitable divergence between a project's formal order and the multiple local complexities, the project documents, which define the project's formal order, manifest the project, and represent a system of knowledge that development agents must relate to in implementing the project.

There is a necessary reduction of complexity in every translation from territory to map. The point is that several different maps can be constructed from identical territories. The IPDP constructs one map of the project's target area and group. The local project staff not only constructs their maps of the area, but also relate to already codified and established maps of development issues. Post-development theoreticians depict development agents' map-construction of the formal order of development. In this thesis, I have shown how these maps interrelate and affect each other, while challenging them with my field experience and the anthropological map generated through this experience.

Post-structural development critique approaches development as a western constructed and embedded discourse that has evolved and become manifested in 'the era of development', i.e., since WW2. Their approach largely rest on an intake to development through the formal, ideational, defined and codified order of development as generated, presented and seen from what is characterised as the west or the developed world, comprising the majority of donor-countries and policy makers. Despite my critiques of post-development theory, it is an effective approach to understand the formal order of development and thus the system of knowledge that is communicated to recipient organisations through development cooperation. As it highlights the formal order, a discursive approach to development is an important way of understanding the formal processes of intended development. However, such an approach only depicts one of the systems of knowledge to which actors relate. By including an actor-oriented approach, it is possible to see how development discourse is perceived, challenged, translated, adopted or rejected locally as it encounters various local practices and knowledge. An actor-oriented approach makes it possible to identify situations of interface, and counter-tendencies that arise from these knowledge encounters, in detail and from an insider's perspective.

The study of the IPDP shows the counter-tendencies that arise in the encounter between development discourse and local practical knowledge. Actors generate informal strategies in their encounter with the formal structures of development. An important point is that actors can be reflexive towards externally imposed knowledge and thus be aware and reflexive towards various systems of knowledge that together comprise their entangled life worlds. This enables them to be eclectic and strategic in their choices and preferences. The inclusion of an actor-oriented approach does not only meet the critique regarding post-development theoreticians' lack of empirical foundation and focus on agency. It also illustrates that the formative power ascribed to structures and discourse is less than generally assumed by post-development theoreticians. Despite the constraints identified with a discursive approach, the focus on actors' agency also illustrates the value of seeing development as a discourse - and as one of the systems of knowledge development actors relate to.

The IPDP exists on two levels, i.e. one formal level, which is codified in various project documents, and one local, practical and informal level. To get to grips with the two levels, two different approaches are required. Both the two levels and the two approaches are interconnected. The formal order affects local practices. This is illustrated by applying theoretical and methodological triangulation. The two approaches spin off and affect each other. A discursive approach illustrates the frames of the formal order, and an actor-orientated approach illustrates how actors relate and act in relation to these frames. Thus, development discourse is not irrelevant for what is taking place, only that it is not what takes place. However, what is taking place locally relates to the formal order. Despite practices diverge from the formal order, they can also contribute to fulfil or confirm the formal order. Consequently, there is neither dualism between formal and informal, nor between discourse and practice. These dichotomies are interrelated. As the development discourse has global connotations, I showed the interrelation between what is perceived as global and local. As local practices reproduce global development discourse, the analytical local-global dichotomy is also reproduced.

Some summarising empirical remarks and exemplifications

The various strategies of actors related to the IPDP illustrate the counter-tendencies that evolve due to the situations of interface between development discourse and local practical knowledge. The various cases I have presented illustrate counter-tendencies and how various strategies appear. A common denominator of these strategies on the informal level is that they all relate to the formal order, and in many ways confirm and thereby reproduce the development discourse on the formal level. This is possible because the IPDP exists on two separate, but also interplaying levels, i.e., a formal and an informal level. In many ways these levels oppose each other simultaneously as they are interrelated. As the formal order is reproduced, the possibilities of continuous informal strategies remain.

The thesis illustrates that actors who are not embedded in or seen as the main bearers of the development discourse are able to be reflexive towards this system of knowledge. Consequently, they are able to develop strategies as coping mechanisms towards the formal order to bridge the gap between development discourse and local practical knowledge. This contributes to contextualise the project and to make it viable, since the implementation of the IPDP largely depends upon the beneficiaries' acceptance and goodwill. The development agents' informal strategies enable this. These strategies can take various forms by various agents. The IPDP manager is somewhat caught in between the beneficiaries and DF, and has problems in being a direct link between the two. He faces difficulties when trying to implement the formal order. The project is dependent on good relations with the beneficiaries, which is gained by being flexible and contextual in the implementation of the formal order. The project manager must also provide and show DF results. While seeing the necessity of flexible implementation, the project manager is simultaneously troubled by the need of producing explicit results to show the donor. However, the DF representative acknowledges the necessity of not being too strict, and not only supports but also encourages the manager to be flexible. But as it is the last year of the IPDP's first phase and an extensive report has to be produced and consigned to NORAD, the DF representative expresses the need to hurry the implementation of previously postponed activities. This illustrates that also the donor tries to be flexible. Whereas the IPDP project manager is caught between the beneficiaries and the project's formal order, the DF representative is caught between the wish to be flexible towards the IPDP staff and, on the other hand, to fulfil DF's obligations to NORAD. The case with the consultant, who is contracted by DF, shows how he, by knowing NORAD's application form and their policy, is strategic, clever and manipulative in his approach. Despite bypassing the formal codified approach and mode of generating an application, he produces an application that 'feeds into the larger whole' of NORAD- documents. The REST case illustrates actors' cleverness applied in making applications, that they are able to be manipulative towards bearers of development discourse, and that documents do not necessarily reflect what is actually going on locally. Compiled, these cases illustrate that the formal order, which is manifested in documents, is of high importance, but also that the formal order is not necessarily what takes place. The cases show that development discourse is interrelated with local practice, and that there are interplays between what is perceived as formal and informal.

The IPDP is manifested in documents. The cases show that various strategies, which do not cohere to the formal order, are applied to produce these documents. A general trend is that the highly acknowledged notion of participatory approach is bypassed. Those evaluating the documents, i.e., the donor and the back-donor, do not have the possibility to check whether participatory approaches are applied or not, since they are detached from the process. Therefore, the documents serve as the back-donor's only intake about the project, and for the applicants it is crucial to know how to address this donor. The project's practices are apparently highly standardised if approached through the formal documents. This is not only due to the stipulated guidelines and format they are supposed to be presented in. It is also due to the standardisation of development problems and solutions, and the general methods used when planning. Contents and structure of project documents are largely characterised by simplified representations and legible units.

Legibility is not only a result of the methods used to plan and design the project. It is also a means to establish causality between provided inputs and expected outputs. Additionally, this design of project documents is largely due to the standard form in which the application (and other project documents for that matter) is to be submitted, which is defined by NORAD. Legibility is not only a means for the planners to intervene and design the project. It is also a means for the funding agency to have legible units of the project to make it easy to measure, monitor and control it. Legibility is also a way to get an understanding of the project, through its documents, and its relevance according to the overall stipulated policy. NORAD's need for legible units in project documents is illustrated in the standardised project application form. The standardisation makes it easy for NORAD's executive officers to comprehend and compare the project with other projects that compete over the same funds. This is also exemplified with the 'DAC- classification'. This not only illustrates how project documents are ordered to be legible, but also illustrates the formal standardisation of development problems. Projects do not exist independently. Intentionally they are supposed to be designed by using participatory approaches. However, projects must cohere not only to the funding agency's policy, but also to general common trends and prioritised areas of western governmental donor agencies. In addition, all external funded development activities in Ethiopia must cohere to the Ethiopian government's national development strategy. This underscores a contradiction in the formal order and general development rhetoric between participatory approaches and policy coherence.

DF is aware of the importance of good documents and that the project's goals and components shall cohere to NORAD's policy. Not only is this illustrated in how DF helps MU in making the application more attractive, by feeding their recipients information about NORAD's target areas and how to address them. It is also illustrated by the inclusion of the consultant in generating a new application for the second phase of the IPDP. He has knowledge about the format, the guidelines, and the problems and policy emphasised by NORAD. Thus, he knows how to address NORAD and how to make the application attractive. He bases his work on the input provided by DF. Actors who are not fully embedded or seen as the main bearers of development discourse can be reflexive, using their knowledge eclectically to pursue their goal of further funding for the IPDP. However, as the strategies involve the production of documents that confirm the formal order, development discourse is reproduced on the formal level.

It would be impossible to plan and design a project that formally ascribes both to participatory approaches and to general policy coherence with external actors and international agreements. This ambiguity is reflected in the IPDP. The IPDP is portrayed as a community based project. How the inclusion of various new project components took place illustrates the lack of participation and bottom-up planning, while challenging the notion of the IPDP as a community based project that applies participatory approaches. Recipient development agents are not only situated passively in the juncture of this ambiguity and formal contradiction. In various ways, they also serve to bridge both the contradiction between top-heavy policy and bottom-up planning, and the encounter between formal discursive development order and local practical knowledge. Local development agents also give legitimacy to the formal inconsistency by representing and mediating between both the target groups and the donor organisation. A crucial point to make any formal codified and designed order viable is individuals and their agency. Agency brings flexibility and local contextual knowledge to rigid structures and systems, and contributes to mediate between and make the two incommensurable entities of participation and policy coherence approach each other. The formal codified order's viability relies on human agency. Formal order is schematic, and its maintenance relies on informal practices and processes. Actors are able to be reflexive towards externally imposed knowledge.

Highlighting some theoretical remarks

The cases referred to above illustrate crucial features of the donor-recipient relationship that would have been invisible with merely a discursive approach to development, and underline the importance of including agency in the analysis in order to get a more complete picture of what is taking place. In depicting development discourse, post-development theoreticians give primary attention to the formal and ideational level of development, as articulated by various donors and policy makers. They construct development as a hegemonic, monolithic and homogenising discourse that is seen and criticised as a western construction and tool in bringing about western modernity. By moving my focus from discourses to the situations where these meet, I show a more nuanced picture of development discourse. This becomes not only a critique of post-development theory, but also a strengthening of its relevance when studying the knowledge encounters of donor-recipient relationships.

Both the anthropological actor orientated and post-development theoretical approaches to development celebrate cultural complexity and ethnographic particularism, but they do it differently. An actor-orientated approach seeks to explore the multitude of complexity and local variations among the target group as an insight to understand the processes involved in knowledge encounters. Post-development theoreticians criticise bearers of the development discourse for not considering cultural variations in their construction of projects and target groups. However, and as I have shown in this thesis, post-development theoreticians advertise for including ethnographic particularism among development interventionists and policy makers, but fail to do it themselves in their portrayal of the development discourse as a hegemonic and homogenising system of knowledge. In combining the discursive and actor-orientated approaches, the portrayal of development discourse becomes more nuanced while illustrating the complexity and variations of the discourse as it is deployed. In including an ethnographic and actor-orientated approach, the shortages of post-development theory are largely met. In addressing and meeting the traditional critique of post-development theory, its relevance is also strengthened.

My study implies a critique against a notion of development as a hegemonic and homogenising entity that shapes the area and actors' practices wherever it is articulated. Additionally, this thesis is a critique of a strong belief in discourse's and structure's formative power when imposed others. This is largely a matter of where analytical focus is put, but also how one perceives the formative power of structures and discourses. These two elements are interrelated, and they show that power is relational. The empirical analysis of the various actors relation to the IPDP shows that development discourse is not as hegemonic and homogenising that post-development scholars tend to believe. The agency of development agents are interrelated and entangled in various systems of knowledge that shape their actions and choices. The empirical actor-orientated approach thus spins off on how one perceives the formative power of discourses and structures. How the arising counter-tendencies turn out is an empirical question.

I also briefly introduced the concept of developmentality. The term draws on Foucault's notion of governmentality. Developmentality is about the donor's transfer of development knowledge to the recipients, who through processes of empowerment, emancipation and participation are supposed to design, take control over and manage their own project in accordance to the donor's wishes, policy and guidelines. Developmentality denotes indirect structures of power and mechanisms of control. The donor tries to make the beneficiaries internalise their development knowledge (or probably the development discourse). In this way, the beneficiaries get a sense of ownership. What is seemingly liberal (empowerment and participatory approach where the beneficiaries make the decisions) is actual a means to transfer developmentality as an indirect mechanism of control and management over the target group, since the decisions they make ultimately must support policy coherence. I hope and aim to further develop, elaborate more and investigate around the notion of developmentality in future work.

I have argued for actors' possibility to be reflexive towards external knowledge, how they relate to it, and what they are doing. Now I would like to be reflexive. Development planning relies upon a reductionist ordering of nature and society in order to intervene. This is a general feature of all planned state and NGO interventions. The whole idea of planning can perhaps be described as extremely un-anthropological. Whereas planning is necessarily schematic and based on simplifications, the anthropological discipline is characterised by studying complex configurations that make up societies and takes as its primary focus locally bound ethnographic particularities. As planning is about construction, anthropology is (despite its ethnographic constructions) largely about deconstructing established 'facts' by identifying the multiple realities and ethnographic particularism. Consequently, anthropological knowledge will always challenge any established formal order. Recently I was told a story from a development workshop on participatory approaches. Introductorily the chairman asked whether there were any anthropologists present. He did so to know how to proceed and manage to workshop, because his experience was that anthropologists inevitably make incessant complaints and quarrels in such meetings where development policy is to be sketched and outlined.

When anthropologists criticise development's formal order, this is done according to a different knowledge and discourse than the knowledge and worldview the formal order of development is based on. Hence, anthropological discourse can probably be seen as largely incommensurable with development discourse, and the bearers and articulators of these two various discourses operate on separate levels. Bearers of a development discourse, as bearers of an anthropological discourse, are characterised by a low degree of inter discursivety. Consequently, the discourses of anthropology and development rarely meet to challenge and affect each other.

This is, for example, illustrated in the case of Nustad's presentation of his recent book, Gavens Makt (2003a), which generally offers a post-development theoretical approach to the discursive formation of Norwegian development assistance over the past 50 years. The presentation was attended by representatives for NORAD and the Norwegian Ministry of Foreign Affairs (MFA), who both were invited to comment on the book. Before their contributions, both underlined that they were there as 'themselves and not as representatives for either NORAD or MFA', and that their arguments was for 'their own account and not for their respective institution'. Consequently, representatives of development discourse and an anthropological discourse did not encounter each other formally. The representatives for the development discourse in that particular context withdrew from their official positions. Such relationships are largely mutual. As development workers neglect anthropological and post-structural critique of development, anthropologists and post-development theoreticians neglect the critique raised against them from representatives of the development sector. Influential critique and possible change largely come from within the prevailing discourse.

The latter introspective argument is not meant to dismiss my own project. I never intended to evaluate or criticise either the Development Fund, Mekelle University, the Department of Agriculture or anything else related to the IPDP. I have illustrated the processes involved in planning, implementation and management of a project and the role of both donor and recipient organisations in this work. My informants in Ethiopia often asked me about any recommendations I might have of how to improve the IPDP. I always stated I had none. I still don't have any. My study never intended to give an assessment of the IPDP, but rather to get an understanding of the processes involved. It is up to my informants to extract information (in the sense as a difference that makes a difference) from this thesis, if possible, to the benefit and improvement of the IPDP. However; I would presume that a conventional assessment of the project probably would conclude that the IPDP lags behind regarding the implementations of some of the planned activities. Nevertheless, my last question in the household survey addressed whether those interviewed knew about the IPDP, and if so what their judgement of it would be. All interviewees stated they were either pleased or very pleased with the general work conducted by the IPDP in Aba'ala.

This tells us that what regards an assessment you get different answers dependent on how the evaluation is conducted and how the project is measured. The general trend in evaluations of development projects, as also the two IPDP evaluations referred to earlier show, is that they merely measure the implementation of project components according to the application and the codified formal order. Evaluations measure the fulfilment of the formal order, and thus relate to the representations and legible units, which are necessities and tools not only for planners and implementers, but also for evaluators. An estimation of the beneficiaries' comprehension of the project and its practical effects would be too time-consuming and complex. It would probably not generate legible units. However, the apparently satisfied beneficiaries of the IPDP only have insight towards what goes on, and not of the formal order and the planned activities that are postponed or cancelled. As I have shown, local development agents and the informal strategies they apply are of paramount importance as they contribute to mediate between the project's formal order, the beneficiaries, and local practical knowledge. Thus, local agency also contributes to the feasibility of the project. The strategies applied by the development agents are of paramount importance for the IPDP and its viability, both in relation to the beneficiaries, and in relation to the donor and back-donor.


Abram, Simone. 2003. Anthropologies in Politicies, Anthropologies in Places: Reflections on Fieldwork 'in' Documents and Policies. In Globalisation. Studies in Anthropology, ed. Thomas Hylland Eriksen, 138-157. London: Pluto Press.

Albrow, Martin. 2001. Globalization after Modernization: a New Paradigm for Development Studies. In Globalization and Development Studies. Challenges for the 21st Century, ed. Frans J. Schuurman. London: SAGE Publications.

Alemu, Diress T., Farah, Kassim O., and Mbuvi, David M. 1998. Impact of Land Use on Woody Vegetation in the Semi-Arid Area of Abala District, North Afar, Ethiopia. Pastoral Information Network Programme (PINEP), Working Paper Series No. 16 May 1998 The African Pastoral Forum. Kabete, Kenya.

Amin, Samir. 1976. Unequal development: as essay on the social formations of peripheral capitalism. Hassocks: Harvester Press.

Appadurai, Arjun. 2002. Disjuncture and Differences in the Global Economy. In The Anthropology of Globalisation, eds. J. Inda and R. Rosaldo: Blackwell Publishers Ltd.

Arce, Alberto, and Long, Norman. 1992. The dynamics of knowledge. Interfaces between bureaucrats and peasants. In Battlefields of knowledge. The interlocking of theory and practice in social research and development, eds. Norman Long and Ann Long, 211-246. London: Routledge.

Arce, Alberto, and Long, Norman. 2000. Reconfiguring Modernity and Development from an Anthropological Perspective. In Anthropology, Development and the Post-Modern Challenge, eds. Alberto Arce and Norman Long, 1-31. London: Routledge.

Archetti, Eduardo P. 1994. Introduction. In Exploring the Written. Anthropology and the Multiplicity of Writing, ed. Eduardo P. Arcetti. Oslo: Scandinavian University Press.

Barth, Fredrik. 1993. A General Framework for Analyzing the Meaning of Acts. In Balinese Worlds, ed. Fredrik Barth, 157-174. Chicago: University of Chicago Press.

Bateson, Gregory. 1972. Steps to an Ecology of Mind. New York: Ballantine Books.

Berger, Peter L., and Luckmann, Thomas. 1992 [1967]. Den samfundsskabte virkelighed: en videnssociologisk afhandling på dansk ved. København: Lindhardt og Ringhof.

Borchgrevink, Axel. 1989. Economy, Morality and the Revolution in a Nicaraguan Peasant Village, Department of Social Anthropology, University of Oslo: Magister Artium Degree.

Brigg, Morgan. 2002. Post-development, Foucault and the colonialisation metaphor. Third World Quarterly 23:421-436.

Chambers, Robert. 1995. Paradigm shifts and the practice of participatory research and development. In Power and Participatory Development, eds. Nici Nelson and Susan Wright, 30-42. London: Intermediate Technology Publications.

Clapham, Christopher. 2002. Controlling Space in Ethiopia. In Remapping Ethiopia. Socialism & After, eds. Wendy James, Donald L. Donham, Eisei Kurimoto and Alessandro Triulzi, 9-32. Oxford; Athens, OH; Addis Ababa, Ethiopia: James Curry, Ohio University Press, Addis Ababa University Press.

Cowen, Michael P., and Shenton, Robert W. 1996. Doctrines of Develpoment. London: Routledge.

Czarniawska, Barbara, and Joerges, Bernard. 1996. Travel of Ideas. In Organizational Change of Translation, ed. de Gruyter, 13-48. Berlin: Wissenschaftzentrum Berlin für Sozialforschung.

DAC. 1992. Development Assistance Manual: DAC Principles for Effective Aid, 142. Paris: Development Assistance Committee (DAC); Organisation for Economic Co-operation and Development (OECD).

Denzin, Norman. 1989. Strategies of Multiple Triangulation. In The Research Act. A Theoretical Introduction to Sociological Methods, ed. Norman Denzin, 234-247. Englewood Cliffs, N.J: Prentice-Hall, Inc.

de Ruijter, Arie. 2001. Globalisation: a Challenge to the Social Sciences. In Globalization and development studies: Challenges for the 21st century, ed. Frans J. Schuurman, 31-43. London: SAGE.

Eriksen, Thomas Hylland. 1989a. Innledning: Hvor mange sykehussenger? In Hvor mange hvite elefanter? Kulturdimensjoen i bistandsarbeidet, ed. Thomas Hylland Eriksen, 9-18. Oslo: Ad Notam Forlag AS.

Eriksen, Thomas Hylland ed. 1989b. Hvor mange hvite elefanter? : kulturdimensjonen i bistandsarbeidet. Oslo: Ad notam.

Eriksen, Thomas Hylland. 1999. Hybrid kreativitet. In Ambivalsens og fundamentalisme: Seks essay om kulturens globalisering, eds. Thomas Hylland Eriksen and Oscar Hemer, 9-21. Oslo: Spartacus Forlag.

Eriksen, Thomas Hylland. 2003. Creolization and Creativity. Global Networks 3:223-237.

Eriksen, Thomas Hylland. forthcoming, 2004. Mixing and Boundaries. Steps towards a more accurate understanding of creolization. Beta version 0.9. in press:21pp.

Escobar, Arturo. 1991. Anthropology and the Development Encounter: The Making and Marketing of Development Anthropology. American Ethnologist: A journal of the American Anthropological Association 4:658-682.

Escobar, Arturo. 1995. Encountering development: the making and unmaking of the Third World. Princeton: Princeton University Press.

Esteva, Gustavo. 1992. Development. In The Development Dictionary. A Guide to Knowledge as Power, ed. Wolfgang Sachs, 6-25. London and New Jersey: Zed Books Ltd.

Ferguson, James. 1994. The Anti-Politics Machine: "Development", Depoliticization, and Bureaucratic Power in Lesotho. Minnesota: Univeristy of Minnesota Press.

Foucault, Michel. 1999. Diskursens orden. Tiltredelsesforelesning holdt ved Collège de France 2. desember 1970. Oslo: Spartacus Forlag A/S.

Foucault, Michel. 2002. Forelesninger om regjering og styringskunst [1978; 1979]: Cappelens upopulære skrifter. Oslo: J.W. Cappelens Forlag a.s.

Frank, Andrè G. 1967. Capitalism and underdevelopment in Latin America. New York: Monthly Review Press.

Gardner, Katy, and Lewis, David. 1996. Anthropology, Development and the Crisis of Modernity. In Anthropology, Development and the Post-modern Challenge, eds. Katy Gardner and David Lewis, 1-25: Pluto Press.

Getachew, Kassa Negussie. 2001. Among the Pastoral Afar in Ethiopia. Tradition, Continuity and Socio-Economic Change. Utrecht, The Netherlands and Addis Ababa, Ethiopia: International Books. In association with Organisation for Social Science Research in Eastern and Southern Africa (OSSREA).

Green, Maia. 2003. Globalizing Development in Tanzania. Policy Franchising through Participatory Project Management. Critique of Anthropology 23:123-143.

Gudeman, Stephen, and Rivera, Alberto. 1990. Conversations in Columbia. The domestic economy in life and text. Cambridge: Cambridge University Press.

Hammer, Svein. 2001. Diskursbegrepet i sosiologisk relieff. Sosiologi i dag. Diskursanalyse (4):7-24.

Hancock, Graham. 1989. Lords of Poverty. London: Macmillan London Ltd.

Hannerz, Ulf. 2003. Several Sites in One. In Globalisation. Studies in Anthropology. ed. Thomas Hylland Eriksen, 18-38. London: Pluto Press.

Harvey, David. 1989. The Condition of Post-Modernity. An inquiry into the origins of cultural change. Oxford: Basil Blackwell.

Hendry, Joy, and Watson, C. W. 2001. Introduction. In An Anthropology of Indirect Communication, eds. Joy Hendry and C. W. Watson, 1-15. London and New York: Routledge.

Hobart, Mark. 1993. Introduction: The Growth of Ignorance? In An Anthropological Critique of Development: The Growth of Ignorance, ed. Mark Hobart, 1-30. London: Routledge.

Hogg, Richard. (ed.) 1997a. Pastoralists, Ethnicity and the State in Ethiopia. London: HAAN Publishing.

Hogg, Richard. 1997b. Introduction. In Pastoralists, Ethicity and the State in Ethiopia, ed. Richard Hogg, 1-21. London: HAAN Publishing.

Jørgensen, Marianne Winther, and Phillips, Louise. 1999. Diskursanalyse som teori og metode.vol. 1. Roskilde: Roskilde Universitetsforlag.

Kearney, Michael. 1995. The Local and the Global: The Anthropology of Globalization and Transnationalism. Annual Review of Anthropology. 24: 547-565.

Kelemework, Tafere. 2000. MA Thesis: Indigenous Institutions of Conflict Resolution among the Aba'la Afar of North-eastern Ethiopia., College of Social Science, Addis Ababa University: MA.

Kiely, Ray. 1999. The Last Refuge of the Noble Savage? A Critical Assessment of Post-Development Theory. The European Journal of Development Research 11:30-55.

Kårhus, Randi. 1992. Diskurs som analytisk begrep. Norsk Antropologisk Tidsskrift:105-117.

Kårhus, Randi. 2001. En Foucault-inspirert diskursanalyse. Sosiologi i dag: Diskursanalyse. (4):25-46.

Lewis, David, Bebbington, Anthony J., Batterbury, Simon P. J., Shah, Alpa, Olson, Elizabeth, Siddiqi, M. Shameem, and Duvall, Sandra. 2003. Practice, Power and Meaning: Framework for Studying Organizational Culture in Multi-Agency Rural Development Projects. Journal of International Development 15:541-557.

Liland, Frode, and Kjerland, Kirsten Alsaker. 2003. Norge, utviklingsland og internasjonalt samarbeid. In Norsk utviklingshjelps historie 3 1989-2002: På bred front. Oslo: Fagbokforlaget.

Long, Norman. 1989. Encounters at the interface : a perspective on social discontinuities in rural development. Wageningen: Agricultural University.

Long, Norman. 1992a. From Paradigm Lost to Paradigm Regained: The Case of an Actor-orientated Sociology of Development. In Battlefields of knowledge. The interlocking of theory and practice in social research and development, eds. Norman Long and Ann Long, 16-43. London: Routledge.

Long, Norman. 1992b. Conclusion. In Battlefields of Knowledge. The Interlocking of Theory and Practice in Social Research and Development, eds. Norman Long and Ann Long, 268-277. London: Routledge.

Long, Norman. 1992c. Introduction. In Battlefields of Knowledge. The Interlocking of Theory and Practice in Social Research and Development, eds. Norman Long and Ann Long, 3-15. London: Routledge.

Long, Norman. 2001. Development Sociology: Actor perspectives. London: Routledge.

Long, Norman, and (eds), Ann Long. 1992. Battlefields of knowledge. The interlocking of theory and practice in social research and development. London: Routledge.

Løngreen, Hanne. 2001. The Development Gaze. Visual Representation of Development in Information Material from Danida. In Encounter Images in the Meetings between Africa and Europe, ed. Mai Palmberg, 221-232. Uppsala, Sweden: The Nordic Africa Institute.

Manzo, Kate. 1991. Modernist discourse and the crisis of development theory. Studies in Comparative International Development 26:3-36.

Melhuus, Marit, and Klausen, Arne Martin eds. 1982. Sosialantropologers rolle i bistandsarbeid: rapport fra et seminar, Sanner 1982. vol. 7. Oslo: Oslo occasional papers in social anthropology.

Neumann, Iver B. 2001a. Scott om staten. Bokspalte: James C. Scott: Seeing Like a State. Internasjonal Politikk 56:565-570.

Neumann, Iver B. 2001b. Mening, makt og materialitet. En innføring i diskursanalyse. Bergen: Fagbokforlaget.

Neumann, Iver B. 2002a. Forord. In Forelsesninger om regjering og styringskunst, ed. Michel Foucault, 5-38. Oslo: Cappelen Akademisk Forlag.

Neumann, Iver B. 2002b. Returning Practice to the Linguistic Turn: The Case of Diplomacy. Millenium 32:627-652.

NORAD. 1999. NORAD investerer i fremtida. NORADs strategi mot år 2005. Oslo: NORAD.

NORAD. 2001. Grant Schemes for Humanitarian Assistance and Development Cooperation by Norwegian and International Voluntary Actors. Guidelines. Accessible at www.norad.no/norsk/files/GUIDLINES%20english%20revised %2030%20oct%202001.doc

NORAD. 2002. Annual Report 2001. Oslo: NORAD.

NORAD. 2003a. Annual Report 2002. Oslo: NORAD.

NORAD. 2003b. SWAPs and Civil Society. The Roles of Civil Society Organisations in Sector Programmes, 50. Oslo: Submitted to NORAD by Centre for Health and Social Development (HeSo).

Nustad, Knut G. 1998. Community leadership and development administration in a Durban squatter settlement, Department of Social Anthropology, University of Cambridge: ph.D.

Nustad, Knut G. 2000. Formell utvikling og uformelle strategier: et post-apartheid utviklingsprosjekt. Norsk Antropologisk Tidsskrift 11:212-226.

Nustad, Knut G. 2001a. Development: the devil we know? Third World Quarterly 22:479-489.

Nustad, Knut G. 2001b. Utvikling som diskursivt rom: makt, fattigdom og aktører. Sosiologi i dag. Diskursanalyse. (4):73-88.

Nustad, Knut G. 2003a. Gavens makt. Norsk utviklingshjelp som formynderskap. Oslo: Pax Forlag A/S.

Nustad, Knut G. 2003b. Considering Global/Local Relations: Beyond Dualism. In Globalisation. Studies in Anthropology, ed. Thomas Hylland Eriksen, 122-137. London: Pluto Press.

Pausewang, Siegfried. 2000. Eritrea - Etiopia: Krig mellom to av verdens fattigste land. In Hvor hender det?.2000.2001 (2). NUPI, Oslo.

Pausewang, Siegfried, Tronvoll, Kjetil, and Aalen, Lovise. 2002a. Democratisation in Ethiopia. Some Notes on Theory and Methodology. In Ethiopia Since the Derg. A Decade of Democratic Pretension and Performance, eds. Siegfried Pausewang, Kjetil Tronvoll and Lovise Aalen, 1-25. London and New York: Zed Books.

Pausewang, Siegfried, Tronvoll, Kjetil, and Aalen, Lovise. 2002b. A Process of Democratisation or Controll? The Historical and Political Context. In Ethiopia Since the Derg. A Decade of Democratic Pretension and Performance, eds. Siegfried Pausewang, Kjetil Tronvoll and Lovise Aalen, 26-44. London and New York: Zed Books.

Pausewang, Siegfried, Tronvoll, Kjetil, and Aalen, Lovise. 2002c. Conclusion: Democracy Unfulfilled? In Ethiopia Since the Derg. A Decade of Democratic Pretension and Performance, eds. Siegfried Pausewang, Kjetil Tronvoll and Lovise Aalen, 230-244. London and New York: Zed Books.

Pieterse, Jan Nederveen. 2000. After post-development. Third World Quarterly 21:175-191.

Porter, Douglas J. 1995. Scenes from childhood: The homesickness of development discourses. In Power of Development, ed. Jonathan Crush. London: Routledge.

Randel, Judith, German, Tone, and Ewing, Deborah. 2002. The Reality of Aid, 2002. An Independent Review of Poverty Reduction and International Development Assistance. IBON Books, IBON Foundation, Inc., Manila, The Philippines.

Rostow, Walt. 1961. The stages of economic growth: A non-communist manifesto. Cambridge: Cambridge University Press.

Sachs, Wolfgang. 1992a. Introduction. In The Development Dictionary. A Guide to Knowledge as Power, ed. Wolfgang Sachs, 1-5. London and New Jersey: Zed Books Ltd.

Sachs, Wolfgang ed. 1992b. The Development Dictionary. A Guide to Knowledge as Power. London and New Jersey: Zed Books Ltd.

Said, Ali. 1997. Resource Use Conflict in the Middle Awash Valley of Ethiopia: The Crisis of Afar Pastoralism. In Pastoralists, Ethicity and the State in Ethiopia, ed. Richard Hogg, 123-141. London: HAAN Publishing.

Said, Ali. 1998. Afar Ethicity in Ethiopian Politics. In Ethnicity and the State in Eastern Africa, eds. M. A. Mohamed Salih and John Markakis, 108-115. Uppsala, Sweden: The Nordic Africa Institute.

Saugestad, Sissel. 2001. The Inconvenient Indigenous. Remote Area Development in Botswana, Donor Assistance, and the First People of Kalahari. Uppsala: The Nordic Africa Institute.

Schuurman, Frans J. 2000. Paradigms lost, paradigms regained? Development studies in the twenty-first century. Third World Quarterly 21:7-20.

Schuurman, Frans J. 2001. Globalization and Development Studies: Introducing the Challenges. In Globalization and development studies: Challenges for the 21st century, ed. Frans J. Schuurman, 3-14. London: SAGE.

Schaanning, Espen. 1995. Forord. In Michel Foucault: Seksualitetens Historie I. Viljen til viten, 7-12. Gjøvik: EXIL.

Scott, James C. 1998. Seeing Like a State. How Certain Schemes to Improve the Human Condition Have Failed. New Haven and London: Yale University Press.

SDPRP. 2002. Sustainable Development and Poverty Reduction Program. The Ethiopian PRSP of the Federal Democratic Republic of Ethiopia (FDRE). July 2002. Accessible at http://www.imf.org/external/np/prsp/2002/eth/01/ 

Stirrat, R. L. 2000. Cultures of Consultancy. Critique of Anthropology 20:31-46.

Svarstad, Hanne. 2001. Samfunnsvitenskapelige diskursanalyser. Sosiologi i dag. Diskursanalyse 4:3-6.

Trouillot, Michel-Rolph. 1992. The Carribean Region: An Open Frontier in Anthropological Theory. Annual Review of Anthropology 21:19-42.

Trouillot, Michel-Rolph. 2001. The Anthropology of the State in the Age of Globalisation. Current Anthropology 42:125-138.

Tsing, Anna. 2002. Conclusion: The Global Situation. In The Anthropology of Globalisation. A Reader, eds. Jonathan Xavier Inda and Renato Rosaldo, 453-485. Oxford: Blackwell Publishers Ltd.

Tvedt, Terje. 1990. Bilder an "de andre". Om utviklingslandene i bistandsepoken. Oslo: Universitetsforlaget.

Tvedt, Terje. 2003. Utviklingshjelp, utenrikspolitikk og makt. Den norske modellen: Makt- og demokratiutredningen 1998-2003. Oslo: Gyldendal Akademisk.

Viezzoli, Claudio. 1992. Scenarios of economic development in the Beles Valley. In Resettlement and rural development in Ethiopia : social and economic research, training, and technical assistance in the Beles Valley, eds. Paolo Dieci and Claudio Viezzoli, 161-185. Milano, Italy: Angeli.

Wadel, Cato. 1991. Feltarbeid i egen kultur: en innføring i kvalitativt orientert samfunnsforskning. Flekkefjord: SEEK.

Wallerstein, Immanuel. 1974. The modern world-system. New York: Academic Press.

Unpublished Sources

AIPDP. 2000. Programme Review: Afar Integrated Pastoral Development Programme (AIPDP) - Review of Progress 1998-2000, 1-20. Oslo: The Development Fund, Oslo, Norway.

Development Fund. 2002. Terms of Reference for the consultancy work of [the consultant] for the project follow-up of the Afar Integrated Pastoral Development Project, June 2002, for the Development Fund. Dated: 6 June 2002

Development Fund. 2002. Innledning. (Accompanying letter for Development Fund's applications according to their framework agreement with NORAD. Consigned to NORAD).

Dioli, Maurizio, and Gebre-Mariam, Ayele. 2001. Evaluation of DF-funded Afar Integrated Pastoral Development Programme. Aba'la Wereda, Afar Region, Ethiopia, 44: External evaluation of IPDP, Develoment Fund.

IPDP. 2000. Annual Report for 1999.

IPDP. 2000. Application for 2001 (approved).

IPDP. 2001. Annual Report for 2000.

IPDP. 2001. Application for 2002 (approved).

IPDP. 2002. Annual Report for 2001.

IPDP. 2002. Project Proposal Phase II (2003-2007).

NORAD. 2002. NORAD - Department for Non-Governmental Organizations. Application for Support to New Project. (NORAD's standardised application format for NGOs. Old version).

NORAD. 2003. NORAD, Department for Civil Society and Private Sector Development. Application form for various recipients of financial assistance (New version). Accessible through http://www.norad.no/default.asp?V_DOC_ID=46 


1 Despite the IPDP formally being named a programme, it is throughout this thesis described as a project, since the latter corresponds to the emic denotation of the IPDP. Additionally, various project documents name the project differently. The name alternates between IPDP, AIPDP (Afar Integrated Pastoral Development Programme), AIDP (Afar Integrated Development Programme), or IPD (Integrated Development Programme). I choose to use IPDP, which is the denotation most commonly used in written sources and exclusively used orally among my informants. 

2 Wereda was formerly, until the formation of the Federal Government in 1995 and the definition of new regional borders that led to the establishment of the Afar region, called sub-district. Wereda is the highest governmental administrative district under the regional level. A region is comprised of zones (Afar region of five, Tigray region of four zones) which do not have any authoritative status. The administrative level beneath wereda is tabia (one wereda comprises an average of 15 tabias). Under tabias are kushets, i.e. a village. In my area, Aba'ala wereda, kushets are similar to tabias as the lowest administrative level. In terms of delivering services, the wereda is the most important administrative unit.

3 Formerly Aba'ala was called Shehet. There are also differences in the spelling of Aba'ala. I've come across: Ab'ala, Aba'la, Abala and Aba'ala. I use the latter mainly due to one of my informants who said it was most proper and that they try to get consensus of one spelling since it is difficult to translate Tigrean and the Arabic words into English.

4 Derg is Amharic for committee.

5 Ras means 'prince'. Ras Mengesha Seyoum was a prince under Emperor Haile Selassie.

6 Taken from http://www.norad.no/default.asp?V_DOC_ID=1402. Accessed 3rd November 2003. My translation.

7 http://www.utviklingsfondet.no/sw599.asp. Accessed 3rd November 2003.

8 As REST, WAT was formerly formally connected to TPLF.

9 This information is compiled from the IPDP application for 2002.

10 The board or steering committee is comprised of 12 members (number of representatives in brackets): Afar Zone 2 Department of Agriculture (2), Zone 2 Administrative council (1), Zone 2 Women's Affairs Office (1), Wereda Administrative Council (1), Wereda Economic Development Office (1), representatives from the elders (2) and Mekelle University (4).

11 IPDP application for 2002, point 3.4.

12 IPDP application 2002, point 3.4.

13 PRSP is an abbreviation for Poverty Reduction Strategy Plan. An approved PRPS is a requirement from the International Monetary Fund (IMF) to establish and guide national development activities funded by IMF.

14 In general development-speak 'those who are to do the development' refera to the target-groups and beneficiaries, who through participatory approaches and processes of empowerment are supposed to plan and design the project, and thus develop themselves. However, I focus on the development agents in the sense of the employees and representatives of the various donor organisations, and not the recipients or beneficiaries.

15 Morgan Brigg argues that development emerged as a result of tendencies within the social sciences in the post-war era (2002).

16 Due to its strategic location at the Horn of Africa, Ethiopia was driven between the interests of the US and the Soviet Union. The US supported Ethiopia heavily up until 1974, when Haile Selassie was overthrown in a military coup d'état led by Menghistu Haile Mariam, who had support from the Soviet Union. After this revolution, the US withdrew its aid from Ethiopia, whereas the USSR increased its economic allocations to Ethiopia. Though those victorious in the revolution of 1991 also adhered to Marxist ideology, they were supported by the US, since it meant a battle against the interests of the USSR. Though no direct causal relation, the fall of the Dergue and Menghistu Haile Mariam in 1991 largely coincided with the fall of the USSR.

17 Truman's inauguration speech consisted of four major points: (1) About NATO cooperation; (2) About the Marshall plan; (3) Rearmament; (4) Development Assistance.

18 Brigg argues that Truman's Point Four merely was an "…idea taken up as a public relations exercise…" (424), as an afterthought in Truman's overall scheme and not something thoroughly considered, intentionally and planned. Still, from my point of view this does not dismiss the effects that emerged from Truman's doctrine, it only stresses the fact that even though unintentional, the power connected to the spoken words might have huge (side-)effects.

19 Actually, Esteva states that Truman was not the first to use the word "underdeveloped". Occasionally during WW2 the American administration used the term in writings about the economic basis for peace and to refer to the gap between rich and poor. "But the expression found no further echo, neither with the public nor with the experts. …it only acquired relevance when Truman presented it as the emblem of his own policy" (Esteva, 1992: 8).

20 Development assistance was not merely initiated on the basis of good intentions. In 1964, the American Senator McGovern said that "The people we assist today will become our customers tomorrow. … An enormous marked for American produce of all kinds will come into being if India can achieve half the productivity of Canada" (cited in Hancock, 1989: 70). The system of development was also established as a tool to limit the influence of communism over the anti-colonial movement and to underscore what the Americans comprehended as an archaic European colonialism (Tvedt, 2003).

21 In 'The Stages of Economic Growth: a non-communist manifesto' (1961), Walt Rostow stipulates five stages or phases the process of development consists of on the way to contemporary, western modernity: The first stage is the traditional society, or pre-Newtonian stage as Rostow calls it. Secondly is the 'transitional stage' where the conditions for "take-off" are developed. Thirdly and most crucial is the "take-off- stage". This stage is short in time (about two decades) and is compared to the English revolution; characterised by rapid growth and expansion, being a period in which modernism is victorious over traditionalism. Stage four is 'drive to maturity' which no longer has industrialisation as a goal, but administrative and technological potential to produce everything they choose and want to produce. Fifth is the 'stage of mass-consumption', which, according to Rostow, had its break-through in the US in 1913-14 with Fordism. This stage was reached in Europe and Japan after WW2, through the Marshall help initiative.

22 Taken from a lecture given by Keith Hart in Oslo, September 9th, 2001. See also Nustad (2003a). Hart also stated that contemporary modernity, the 'super modern society', is based upon or composed of human rights, gender equality, environmental sensibility, etc.

23 Kate Manzo's word for the '-ism' created by modernisation theoreticians, those who pursue development (Manzo, 1991).

24 Kate Manzo's word for some of the "founding fathers" of the dependency theory, a school that emerged in Mexico lead by the Mexican socio-economist André Gunder Frank (Manzo, 1991).

25 These four premises are: 1: The United States along with other industrialised nations was on top of the social evolutionary scale, but today this premise of superiority has been shattered by the ecological predicament, i.e., that the ecological system harmed by industrialisation prevents others from industrialising. 2: Truman's vision of development which would engage loyalty of the decolonised countries towards the US due to the rising influence of the Soviet Union, the first country to industrialise outside capitalism. For over 40 years, development was a weapon in the competition between political systems, but today the East-West confrontation is over, due to the fall of the USSR. 3: Development has changed the face of the world, but not as intended, as illustrated by the increasing division between rich and poor countries. 4: Suspicion towards the project of development grows and some see it as a misconceived idea and fears the success of development where everyone will become alike (Sachs, 1992: 2-3).

26 The relation between modernity and post-modernity is widely questioned, see e.g. Schaanning (1992), and this debate is not issue for this thesis. Still, like post-development scholars, I argue that development is a modernistic project per se.

27 Brigg exemplifies this, e.g. discourses, institutions, architectural forms, regulatory decisions, laws, administrative measures, scientific statements, philosophical, moral and philanthropic propositions (2002: 427).

28 With reference to its intentionality, knowledge is often divided between normative or descriptive knowledge. I do not tend to adopt this dichotomy, since both systems of knowledge are information and can do something - regardless of normative or descriptive information. Bateson defines information as a difference that makes a difference (Bateson, 1972), a definition that makes a distinction between normative and descriptive knowledge meaningless.

29 This distinction also prevents misguided descriptions where "we" are theoreticians and "they" are practitioners.

30 AIPDP is the same project as the IPDP. The 'A' denotes Afar, i.e. AIPDP means Afar Integrated Pastoral Development Programme.

31 From www.u-fondet.no, accessed 6 November 2002.

32 If the households that combine agriculture with other income-generating activities such as salaried employment and trade are included, this figure increases to 18 households (31%).

33 The household survey was conducted in May and June 2002, that is, the period of the year just before the rainy season, thus a lot of nomadic people living in the areas might be on the move in search of water and grazing areas for their livestock. Still: it is the sons in the households who serve as herders, and the rest of the families, including the head of households (HH) were in the area. Further, a lot of HH-heads coming from remote areas were interviewed when they came to Aba'ala City, especially for the weekly Thursday marked.

34 Though not being a complete list, examples of such words are: participation, grass roots, community based, empowerment, women in development, water harvesting, introduce, capacity building, good governance, awareness, comparative advantages etc. Tvedt denotes such words as "words of honour" (honnørord) (1990).

35 The DHP is a research-orientated project funded by Sida (the Swedish equivalent to NORAD) and feeds the IPDP with its research results. The DHP and the IPDP have the same steering committee. The DHP started three years prior to the IPDP.

36 The indicators that will be used is, according to the project application: "- Number of tree and fruit seedlings produced, planted and survived; - Number of people getting employment opportunities at the nursery site; - Number of women involved in vegetable production."

37 Amongst the species planted is the eucalyptus tree. A Belgian researcher in Mekelle challenges the widespread use of the eucalyptus tree because it damages the soil and actually accentuates soil degradation since the roots do not bind to each other, it attracts mercury that causes toxication of the soil, and it uses an enormous amount of water, thus is parasitic on other species. This is acknowledged by the IPDP manager. Nevertheless, he argues that the eucalyptus is good since it grows fast and therefore is good as fuel-wood and building material. He argues that it is a question of what to choose from two necessary evils.

38 Due to lack of data, I can not verify or falsify this statement, though it visualises different comprehensions or insights as to the project's activities.

39 The Development Fund's application for the AIPD 2002 submitted to and accepted by NORAD. AIPD is a new acronym used in this application. The project's full name is Afar Integrated Pastoral Development Programme. The project is always orally referred to as IPDP, an abbreviation I use.

40 Translated from Norwegian, taken from the Introduction in the Development Fund's application to NORAD for the project year of 2002, under point 1.1: "Sentrale kjennetegn ved organisasjonen".

41 Chambers (1995) states that this implies putting humans rather than infrastructure as the focal point of development, where top-down becomes bottom-up, planning becomes participation, the standardised becomes diverse, centralised planning turns into decentralised participation, etc. (see Chambers, 1995: 32 for an complete list of resembling dichotomies).

42 These questions appear as point 3.5 and 3.6 in the application format. The text for 3.5 is: "HIV/AIDS: In what ways is the organization dealing with HIV/AIDS situation? If the organization is not working with HIV/AIDS NORAD would like argument for/ against including for e.g. preventive measures against HIV/AIDS as project activities". The text for 3.6 is: "Good governance/corruption: Give an account of how the organization is working with good governance and corruption".

43 In MU's final project proposal for 2002 submitted to DF, MU addresses the issue of HIV/Aids like this: "The [IPDP] will contribute to the control of HIV/AIDS by involving itself in the local initiatives already undergoing in Aba'ala by the Wereda Health Office and the youth".

44 My emphasis.

45 This does not refer to DF, MU or the IPDP. Though the local community initially were a bit reluctant to the IPDP, and 'had to be talked into it', both the project staff, the board members and the beneficiaries state that DF and MU as external actors have gained the local community's trust.

46 Informal practical knowledge is similar to what Scott denotes as metis, that is 'knowledge that can come only from practical experience' (1998: 6). Informal practical knowledge resembles Scott conception of metis as "…practical skills, variously called know-how, … common sense, experience, a knack…" (ibid.: 311).

47 Development Fund funds this network, "the Triangular", where both academic institutions and implementing NGOs from Norway, Ethiopia and India participate. The project emphasises south-south relations, which the donor facilitates in order for the partners in India and Ethiopia to exchange know-how about development issues. Mekelle University (MU) and Relief Society of Tigray are the Ethiopian participants.

48 According to the project documents, the IPDP is a program, and not a project. Both DF representatives and local staff and board members talk about it as a project, thus project is applied to talk about the IPDP.

49 This quotation is not literal, rather a renarration of what was said. It is based on notes taken during the conversation between the donor and local project manager.

50 This quotation is not literal, rather a renarration of what was said. It is based on notes taken during the conversation between the donor and local project manager.

51 From the partnership agreement between NORAD and Development Fund; an agreement all NGO's receiving funding from NORAD has to sign: "Cooperation Agreement between The Norwegian Agency for Development Cooperation (NORAD) and 'grant recipient. [Development Fund]". Accessed from www.norad.no January 26, 2003.

52 Obviously, I was invited merely because I already was talking to the DF representative.

53 This is the second time the fieldtrip is delayed, but for different reasons. The first time, the reason given was that nobody was around to accept and sign the necessary papers for a small withdrawal from MU and the project's bank account to cover the expenses of the field study, set to approximately 2000 birr (2000 NOK). At that time, the project manager expressed frustration over not being allowed or able to manage and control funds on the project he supervises, instead "I have to wait for a week for a signature".

54 Cited from the Terms of References for the consultant, produced by DF, and signed by both parts.

55 In a meeting between DoA and the evaluation team investigating the partnership model, DoA expressed concern over the plan to transfer the whole responsibility to them and that they lack the overall infrastructure, capacity, knowledge and technology to manage the project alone.

56 The form is called "NORAD -Department for Non-Governmental Organizations. Application for Support to New Project, Year: 2003". From autumn 2002 a new application format substitutse the old ones, as the one used in this case.

57 REST arose in the mid 1980s as the humanitarian wing of the Tigray People's Liberation Front (TPLF) to coordinate aid assistance in Tigray. Today TPLF is part of the governing Ethiopian party Ethiopian People's Revolutionary Democratic Front (EPRDF), and REST is formally independent NGO. REST has several external donors.

58 This information is from one of my informants that I accidentally met at a tea-house in Mekelle. I never observed this project. The project referred to is not funded by a Norwegian donor organisation. Also important is that, as far as I know, this is not a general policy of REST, merely one example on a coping mechanism in relation to the constraints latent in the donor-recipient relationship.

59 The project managers' approach does however diverge from the intentional notion of participatory approaches, that is, to empower the beneficiaries to design, take control over and manage their own project (cf. Chambers, 1995).

60 The ambiguity of executive is whether the term refers to whether it has to do with administration and management, i.e. that the summary executes something, or to someone's (the summary's?) authority and power to carry out decisions.

61 Cf. the NORAD application form and also that the project manager states that this form is well arranged with headlines, numbers and questions only to be filled in. Introductorily in the NORAD application format, 'general information' is required, i.e. name (acronyms), area, duration, partner's name etc. before a 'brief project description' to give an outline of the document's structure.

62 Cf. the consultant whose work is to represent an objective and external third person view.

63 The priority given to quantitative data is reflected both in the planning and design of the project and in the methods used to measure the different components' successes. Cf. the different representations and e.g. the nursery component on how the success is measured: Number of seedlings distributed, and not the practical use of them.

64 The IPDP application is tidily arranged, with a direct causal relation between input (funds) and output (objectives). The goals/ objectives are to be reached through different components that consist of different activities.

65 This is not the case with the IPDP. Rather the acronyms and the translation are given continuously throughout the document, but once the acronym is presented in brackets and related to something, they are used all through the document.

66 Cf. "Annex I" of the evaluation (Dioli and Gebre-Mariam, 2001) and "Annex 2: Ininerary [sic]" of IPDP Project Proposal Phase II (2003-2007), submitted by MU to DF. Annex 2 is produced by the consultant referred to in the second case above.

67 In the NORAD application format for existing and new projects, as for the NORAD project report format, one of the last points (under point 4.2.) is "Detailed documentation is to be found in the following documents". These are: the IPDP progress and annual reports from 1998 and 1999, IPDP mid-term and annual report 2000, IPDP Programme Review 2000, two documents produced by the Dryland Husbandry Project (DHP) in Aba'ala, one FAO (Food and Agriculture Organisation of the United Nations) report from 1988, and two academic works, whereas one is produced by an IPDP board member and the other an MA-thesis.

68 Scott depicts these four elements in Seeing Like a State's 'Introduction' (page 4-6). Though, on page 88 these four elements are reduced to three. The only difference is that in the latter case the first element is a compilation of the first and second element presented in the 'Introduction'.

69 The notion that the state was equivalent to the government was re-thought by, among others, Foucault after the student-riots in Paris in 1968-69. According to previous theories, the state would collapse due to such disturbances, but it didn't. The reason given by Foucault was that state effects were produced by others than just the government, and that public institutions and organisations relied upon the state apparatus, as well as embedded and decentralised legislation supports the state's/ government's ideology. Another fact was that, as Scott puts it, many institutions actually did "see like a state". In respect to this, Foucault introduced the notion of governmentality (Neumann, 2002a). Below I elaborate further around governmentality.

70 The census referred to in chapter three states that in Aba'ala town 2.765 people are Tigreans and 405 are Afar of a total of 3.300 (the rest are different other groups, e.g. Amhara, Oromo). Also an idea that Afar region is inhabited by Afar people is contested by the fact that Afar region prior to the formation of the Federal Government in 1995 were split between Tigray and Welo region. Afar as a region was established in 1995.

71 Data obtained from the household survey. All ten households interviewed in Hidmo are Tigreans. Of the 20 samples from Wuhdet, 16 are Tigrean- and four are Afar-households. The project manager verifies the plausibility of these data.

72 These quotations are taken from the IPDP project application for 2002, point 3.1: "What are the project's anticipated results in the year for which application is being made?"

73 Stated by a project manager at REST.

74 Point 5: "Reporting" in NORAD's "Cooperation agreement between The Norwegian Agency for Development Cooperation (NORAD) and (grant recipient's name) (Grant recipient)". www.norad.no accessed May 29th, 2002.

75 Ibid.: point 1.2.1.

76 Ibid., point 6.1. My emphasis.

77 DANIDA is Denmark's equivalent to NORAD.

78 The 1994 census states that also some Oromos and Amharas live in the area.

79 For the whole quotation from the projects document, see chapter three.

80Foreign NGOs operating in Ethiopia have to be approved by a governmental office which secures that the NGOs operate in accordance with the Ethiopian National Development Plan.

81 Menilek is perceived as one of the greatest emperors of Ethiopia, for several reasons: He lead the army and defeated the invading Italian troops in the Battle of Adwa. Under his reign, Ethiopia expanded to its present size, and his expansionistic drive contributed to ethnic consolidation. Menilek also started the project of modernising Ethiopia, through improving city infrastructure and building a railway to Djibouti in the East,

82 In 1916, Ras Tafari Mekonnen (1892-1975; ras is amharic for prince, i.e. the name denotes Mekonnen, the prince of the Tafari people) was appointed as regent and prime minister under Queen Zadwitu. He was crowned as emperor in 1928. In 1930 he took the title negus nagest (king of kings) and the name Haile Selassie

83 Awash River is in the southern part of what's today the Afar-region.

84 The military junta that came to power as a result of the coup d'etat in 1974. The coup d'etat was in 1974, but the area of Mengistu Haile Mariam was from 1977-1991: The popular movement, which ended the regime of Haile Selassie in 1974, was slowly sidelined, and Mengistu himself took power after the execution of General Tafari Bente on February 5 1977. Thus, the common understanding that the area of the Derg and Mengistu are coherent are wrong (Pausewang et al., 2002: 26, 44).

85 This is Ethiopia's Poverty Reduction Strategy Paper (PRSP), and is a requirement for collaboration with the International Monetary Fund (IMF), the World Bank and a host of other donors. This PRSP is 202 pages long and states in the introduction that it 'formulates policies and strategies to guide over all development' in Ethiopia.

86 An abbreviation of Ethiopian People's Revolutionary Democratic Front, the governing party in Ethiopia. The EPRDF was established in 1989 by TPLF (Tigrean People's Liberation Front) with TPLF-affiliated parties outside Tigray, in order to have a broader and united front against the Derg government in Addis Ababa, which was overthrown in 1991. After a transitional period, EPRDF took the governmental office, which they have held ever since.

87 Pausewang, Tronvoll and Aalen have not conducted fieldwork in the Afar- region. Nevertheless the issue of party-cadres are also present in central areas of Afar, i.e. the areas close to the highland, where there are e.g. settlements, governmental offices or arable land. Aba'ala town was initially settled approximately 35 years ago, and was at that time part of the Tigray region. I was told by an IPDP board member that some people moved to Aba'ala to get access to arable land (along the river), while others were forced to move there by the government to build roads to the saline lakes and to the Red Sea. Party cadres came along with these settlements.

88 Two out of three times I arrived at the Norwegian Embassy in Addis Ababa for an appointment only to be told that the official I was supposed to interview had to rush into a meeting with the Ethiopian bureaucracy regarding the PRSP, and thus I was redirected into another official's office.

89 Personal communication, 2003.

90 Time-space compression denotes a trend in capitalist political economy and implies acceleration and shortening in the average turnover time between investments and the taking of profit. Harvey adopts the term and use time-space compression to signify the current trends in cultural change due to the rapidly increase in encounters between different cultures and systems of knowledge (cf. Kearney, 1995).

91 This was later confirmed by an Ethiopian medical doctor who stayed in Aba'ala for some weeks. He travelled around in the area and reported on a hyper-epidemic caused by malaria in areas outside Aba'ala town. He expressed concern on the lack of anti-malaria initiatives from NGOs and the Ethiopian government.

92 Cf. chapter four and Stirrat (2000) on the structure, language and aesthetics of project documents.

93 This is the main reason to the engagement of Mekelle University, besides they have the technical skills in general project management, in the IPDP. Intentionally DF wants only to have a partnership with Department of Agriculture (DoA) in Aba'ala, but MU is included due to DoA lack of authorized accountants.

94 I.e. the steering committee, cf. IPDP application for 2002 point 3.6 about good governance.

95 This is largely analogous to what I denoted as developmentality in chapter five.

96 Annual report, 2001, point 2.2.9: "Support for repairing Aba'ala primary school".

97 The disbursement to rebuild the school was reported, but it was easy to argue on the relevance of supporting this since NORAD promoted education initiatives.

98 In his argument, Nustad refers to the work of Robertson (1995).